diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,95227 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 13599,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 7.353481873667182e-05,
+      "grad_norm": 33.42438888549805,
+      "learning_rate": 0.0,
+      "loss": 8.5464,
+      "step": 1
+    },
+    {
+      "epoch": 0.00014706963747334364,
+      "grad_norm": 33.27947998046875,
+      "learning_rate": 4.901960784313725e-07,
+      "loss": 8.4756,
+      "step": 2
+    },
+    {
+      "epoch": 0.00022060445621001543,
+      "grad_norm": 33.260406494140625,
+      "learning_rate": 9.80392156862745e-07,
+      "loss": 8.5203,
+      "step": 3
+    },
+    {
+      "epoch": 0.0002941392749466873,
+      "grad_norm": 32.35409164428711,
+      "learning_rate": 1.4705882352941177e-06,
+      "loss": 8.3242,
+      "step": 4
+    },
+    {
+      "epoch": 0.00036767409368335905,
+      "grad_norm": 33.71382522583008,
+      "learning_rate": 1.96078431372549e-06,
+      "loss": 8.4519,
+      "step": 5
+    },
+    {
+      "epoch": 0.00044120891242003087,
+      "grad_norm": 33.300838470458984,
+      "learning_rate": 2.450980392156863e-06,
+      "loss": 8.5344,
+      "step": 6
+    },
+    {
+      "epoch": 0.0005147437311567027,
+      "grad_norm": 33.16044616699219,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 8.3724,
+      "step": 7
+    },
+    {
+      "epoch": 0.0005882785498933746,
+      "grad_norm": 33.9757080078125,
+      "learning_rate": 3.431372549019608e-06,
+      "loss": 8.5652,
+      "step": 8
+    },
+    {
+      "epoch": 0.0006618133686300463,
+      "grad_norm": 33.517520904541016,
+      "learning_rate": 3.92156862745098e-06,
+      "loss": 8.4422,
+      "step": 9
+    },
+    {
+      "epoch": 0.0007353481873667181,
+      "grad_norm": 32.61171340942383,
+      "learning_rate": 4.411764705882353e-06,
+      "loss": 8.2516,
+      "step": 10
+    },
+    {
+      "epoch": 0.00080888300610339,
+      "grad_norm": 33.461524963378906,
+      "learning_rate": 4.901960784313726e-06,
+      "loss": 8.4964,
+      "step": 11
+    },
+    {
+      "epoch": 0.0008824178248400617,
+      "grad_norm": 33.20984649658203,
+      "learning_rate": 5.392156862745099e-06,
+      "loss": 8.2137,
+      "step": 12
+    },
+    {
+      "epoch": 0.0009559526435767336,
+      "grad_norm": 32.431461334228516,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 8.0981,
+      "step": 13
+    },
+    {
+      "epoch": 0.0010294874623134054,
+      "grad_norm": 32.05772399902344,
+      "learning_rate": 6.372549019607843e-06,
+      "loss": 8.0857,
+      "step": 14
+    },
+    {
+      "epoch": 0.0011030222810500773,
+      "grad_norm": 31.95595932006836,
+      "learning_rate": 6.862745098039216e-06,
+      "loss": 8.179,
+      "step": 15
+    },
+    {
+      "epoch": 0.0011765570997867491,
+      "grad_norm": 31.533361434936523,
+      "learning_rate": 7.3529411764705884e-06,
+      "loss": 7.9392,
+      "step": 16
+    },
+    {
+      "epoch": 0.0012500919185234208,
+      "grad_norm": 31.489395141601562,
+      "learning_rate": 7.84313725490196e-06,
+      "loss": 8.07,
+      "step": 17
+    },
+    {
+      "epoch": 0.0013236267372600927,
+      "grad_norm": 23.298419952392578,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 7.3708,
+      "step": 18
+    },
+    {
+      "epoch": 0.0013971615559967645,
+      "grad_norm": 23.4985294342041,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 7.4888,
+      "step": 19
+    },
+    {
+      "epoch": 0.0014706963747334362,
+      "grad_norm": 21.479463577270508,
+      "learning_rate": 9.31372549019608e-06,
+      "loss": 7.2316,
+      "step": 20
+    },
+    {
+      "epoch": 0.001544231193470108,
+      "grad_norm": 20.94013786315918,
+      "learning_rate": 9.803921568627451e-06,
+      "loss": 7.2703,
+      "step": 21
+    },
+    {
+      "epoch": 0.00161776601220678,
+      "grad_norm": 21.09307861328125,
+      "learning_rate": 1.0294117647058824e-05,
+      "loss": 7.4698,
+      "step": 22
+    },
+    {
+      "epoch": 0.0016913008309434518,
+      "grad_norm": 19.971445083618164,
+      "learning_rate": 1.0784313725490197e-05,
+      "loss": 7.336,
+      "step": 23
+    },
+    {
+      "epoch": 0.0017648356496801235,
+      "grad_norm": 14.1094970703125,
+      "learning_rate": 1.1274509803921569e-05,
+      "loss": 6.5375,
+      "step": 24
+    },
+    {
+      "epoch": 0.0018383704684167953,
+      "grad_norm": 12.502970695495605,
+      "learning_rate": 1.1764705882352942e-05,
+      "loss": 6.647,
+      "step": 25
+    },
+    {
+      "epoch": 0.0019119052871534672,
+      "grad_norm": 11.730005264282227,
+      "learning_rate": 1.2254901960784313e-05,
+      "loss": 6.4809,
+      "step": 26
+    },
+    {
+      "epoch": 0.001985440105890139,
+      "grad_norm": 11.933045387268066,
+      "learning_rate": 1.2745098039215686e-05,
+      "loss": 6.522,
+      "step": 27
+    },
+    {
+      "epoch": 0.0020589749246268108,
+      "grad_norm": 11.171988487243652,
+      "learning_rate": 1.323529411764706e-05,
+      "loss": 6.3035,
+      "step": 28
+    },
+    {
+      "epoch": 0.0021325097433634826,
+      "grad_norm": 11.34299373626709,
+      "learning_rate": 1.3725490196078432e-05,
+      "loss": 6.4642,
+      "step": 29
+    },
+    {
+      "epoch": 0.0022060445621001545,
+      "grad_norm": 10.365833282470703,
+      "learning_rate": 1.4215686274509804e-05,
+      "loss": 6.3085,
+      "step": 30
+    },
+    {
+      "epoch": 0.0022795793808368264,
+      "grad_norm": 8.946077346801758,
+      "learning_rate": 1.4705882352941177e-05,
+      "loss": 6.0133,
+      "step": 31
+    },
+    {
+      "epoch": 0.0023531141995734983,
+      "grad_norm": 6.775783538818359,
+      "learning_rate": 1.5196078431372548e-05,
+      "loss": 5.8244,
+      "step": 32
+    },
+    {
+      "epoch": 0.0024266490183101697,
+      "grad_norm": 5.776180744171143,
+      "learning_rate": 1.568627450980392e-05,
+      "loss": 5.736,
+      "step": 33
+    },
+    {
+      "epoch": 0.0025001838370468416,
+      "grad_norm": 5.557833194732666,
+      "learning_rate": 1.6176470588235296e-05,
+      "loss": 5.8184,
+      "step": 34
+    },
+    {
+      "epoch": 0.0025737186557835134,
+      "grad_norm": 5.11709451675415,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 5.563,
+      "step": 35
+    },
+    {
+      "epoch": 0.0026472534745201853,
+      "grad_norm": 5.017343044281006,
+      "learning_rate": 1.715686274509804e-05,
+      "loss": 5.538,
+      "step": 36
+    },
+    {
+      "epoch": 0.002720788293256857,
+      "grad_norm": 5.129957675933838,
+      "learning_rate": 1.7647058823529414e-05,
+      "loss": 5.5012,
+      "step": 37
+    },
+    {
+      "epoch": 0.002794323111993529,
+      "grad_norm": 4.650245666503906,
+      "learning_rate": 1.8137254901960785e-05,
+      "loss": 5.3628,
+      "step": 38
+    },
+    {
+      "epoch": 0.002867857930730201,
+      "grad_norm": 4.289092540740967,
+      "learning_rate": 1.862745098039216e-05,
+      "loss": 5.4213,
+      "step": 39
+    },
+    {
+      "epoch": 0.0029413927494668724,
+      "grad_norm": 3.927570343017578,
+      "learning_rate": 1.9117647058823528e-05,
+      "loss": 5.2669,
+      "step": 40
+    },
+    {
+      "epoch": 0.0030149275682035443,
+      "grad_norm": 3.940927505493164,
+      "learning_rate": 1.9607843137254903e-05,
+      "loss": 5.3125,
+      "step": 41
+    },
+    {
+      "epoch": 0.003088462386940216,
+      "grad_norm": 3.8628530502319336,
+      "learning_rate": 2.0098039215686277e-05,
+      "loss": 5.193,
+      "step": 42
+    },
+    {
+      "epoch": 0.003161997205676888,
+      "grad_norm": 3.990081787109375,
+      "learning_rate": 2.058823529411765e-05,
+      "loss": 5.0477,
+      "step": 43
+    },
+    {
+      "epoch": 0.00323553202441356,
+      "grad_norm": 3.586876392364502,
+      "learning_rate": 2.107843137254902e-05,
+      "loss": 4.986,
+      "step": 44
+    },
+    {
+      "epoch": 0.0033090668431502318,
+      "grad_norm": 3.356776475906372,
+      "learning_rate": 2.1568627450980395e-05,
+      "loss": 4.9963,
+      "step": 45
+    },
+    {
+      "epoch": 0.0033826016618869036,
+      "grad_norm": 3.476984977722168,
+      "learning_rate": 2.2058823529411766e-05,
+      "loss": 4.8855,
+      "step": 46
+    },
+    {
+      "epoch": 0.003456136480623575,
+      "grad_norm": 3.3652610778808594,
+      "learning_rate": 2.2549019607843138e-05,
+      "loss": 4.7406,
+      "step": 47
+    },
+    {
+      "epoch": 0.003529671299360247,
+      "grad_norm": 3.228240966796875,
+      "learning_rate": 2.303921568627451e-05,
+      "loss": 4.6926,
+      "step": 48
+    },
+    {
+      "epoch": 0.003603206118096919,
+      "grad_norm": 2.766308069229126,
+      "learning_rate": 2.3529411764705884e-05,
+      "loss": 4.6151,
+      "step": 49
+    },
+    {
+      "epoch": 0.0036767409368335907,
+      "grad_norm": 2.8335180282592773,
+      "learning_rate": 2.401960784313726e-05,
+      "loss": 4.6808,
+      "step": 50
+    },
+    {
+      "epoch": 0.0037502757555702626,
+      "grad_norm": 2.5543577671051025,
+      "learning_rate": 2.4509803921568626e-05,
+      "loss": 4.6245,
+      "step": 51
+    },
+    {
+      "epoch": 0.0038238105743069344,
+      "grad_norm": 2.4093687534332275,
+      "learning_rate": 2.5e-05,
+      "loss": 4.6069,
+      "step": 52
+    },
+    {
+      "epoch": 0.0038973453930436063,
+      "grad_norm": 2.233090400695801,
+      "learning_rate": 2.5490196078431373e-05,
+      "loss": 4.4216,
+      "step": 53
+    },
+    {
+      "epoch": 0.003970880211780278,
+      "grad_norm": 2.0228404998779297,
+      "learning_rate": 2.5980392156862747e-05,
+      "loss": 4.3531,
+      "step": 54
+    },
+    {
+      "epoch": 0.00404441503051695,
+      "grad_norm": 1.8846389055252075,
+      "learning_rate": 2.647058823529412e-05,
+      "loss": 4.2662,
+      "step": 55
+    },
+    {
+      "epoch": 0.0041179498492536215,
+      "grad_norm": 1.7868231534957886,
+      "learning_rate": 2.696078431372549e-05,
+      "loss": 4.2191,
+      "step": 56
+    },
+    {
+      "epoch": 0.004191484667990294,
+      "grad_norm": 1.7782880067825317,
+      "learning_rate": 2.7450980392156865e-05,
+      "loss": 4.2185,
+      "step": 57
+    },
+    {
+      "epoch": 0.004265019486726965,
+      "grad_norm": 1.7905795574188232,
+      "learning_rate": 2.7941176470588236e-05,
+      "loss": 4.1929,
+      "step": 58
+    },
+    {
+      "epoch": 0.004338554305463637,
+      "grad_norm": 1.6845123767852783,
+      "learning_rate": 2.8431372549019608e-05,
+      "loss": 4.1337,
+      "step": 59
+    },
+    {
+      "epoch": 0.004412089124200309,
+      "grad_norm": 1.5980770587921143,
+      "learning_rate": 2.8921568627450986e-05,
+      "loss": 4.0985,
+      "step": 60
+    },
+    {
+      "epoch": 0.0044856239429369804,
+      "grad_norm": 1.5408473014831543,
+      "learning_rate": 2.9411764705882354e-05,
+      "loss": 4.0492,
+      "step": 61
+    },
+    {
+      "epoch": 0.004559158761673653,
+      "grad_norm": 1.4808381795883179,
+      "learning_rate": 2.9901960784313725e-05,
+      "loss": 3.9439,
+      "step": 62
+    },
+    {
+      "epoch": 0.004632693580410324,
+      "grad_norm": 1.4135316610336304,
+      "learning_rate": 3.0392156862745097e-05,
+      "loss": 3.9034,
+      "step": 63
+    },
+    {
+      "epoch": 0.0047062283991469965,
+      "grad_norm": 1.3860609531402588,
+      "learning_rate": 3.0882352941176475e-05,
+      "loss": 3.8921,
+      "step": 64
+    },
+    {
+      "epoch": 0.004779763217883668,
+      "grad_norm": 1.3163044452667236,
+      "learning_rate": 3.137254901960784e-05,
+      "loss": 3.8668,
+      "step": 65
+    },
+    {
+      "epoch": 0.004853298036620339,
+      "grad_norm": 1.3111575841903687,
+      "learning_rate": 3.186274509803922e-05,
+      "loss": 3.7666,
+      "step": 66
+    },
+    {
+      "epoch": 0.004926832855357012,
+      "grad_norm": 1.127004623413086,
+      "learning_rate": 3.235294117647059e-05,
+      "loss": 3.7912,
+      "step": 67
+    },
+    {
+      "epoch": 0.005000367674093683,
+      "grad_norm": 1.4024866819381714,
+      "learning_rate": 3.284313725490196e-05,
+      "loss": 3.8148,
+      "step": 68
+    },
+    {
+      "epoch": 0.0050739024928303554,
+      "grad_norm": 1.038182020187378,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 3.7591,
+      "step": 69
+    },
+    {
+      "epoch": 0.005147437311567027,
+      "grad_norm": 0.9515771269798279,
+      "learning_rate": 3.382352941176471e-05,
+      "loss": 3.6804,
+      "step": 70
+    },
+    {
+      "epoch": 0.005220972130303699,
+      "grad_norm": 0.9373134970664978,
+      "learning_rate": 3.431372549019608e-05,
+      "loss": 3.6873,
+      "step": 71
+    },
+    {
+      "epoch": 0.005294506949040371,
+      "grad_norm": 0.8617609143257141,
+      "learning_rate": 3.480392156862745e-05,
+      "loss": 3.6905,
+      "step": 72
+    },
+    {
+      "epoch": 0.005368041767777042,
+      "grad_norm": 0.9113644361495972,
+      "learning_rate": 3.529411764705883e-05,
+      "loss": 3.6142,
+      "step": 73
+    },
+    {
+      "epoch": 0.005441576586513714,
+      "grad_norm": 0.9359166622161865,
+      "learning_rate": 3.5784313725490195e-05,
+      "loss": 3.6282,
+      "step": 74
+    },
+    {
+      "epoch": 0.005515111405250386,
+      "grad_norm": 0.8804528713226318,
+      "learning_rate": 3.627450980392157e-05,
+      "loss": 3.5343,
+      "step": 75
+    },
+    {
+      "epoch": 0.005588646223987058,
+      "grad_norm": 0.9673545956611633,
+      "learning_rate": 3.6764705882352945e-05,
+      "loss": 3.597,
+      "step": 76
+    },
+    {
+      "epoch": 0.00566218104272373,
+      "grad_norm": 0.8104327321052551,
+      "learning_rate": 3.725490196078432e-05,
+      "loss": 3.486,
+      "step": 77
+    },
+    {
+      "epoch": 0.005735715861460402,
+      "grad_norm": 0.7928115725517273,
+      "learning_rate": 3.774509803921569e-05,
+      "loss": 3.5484,
+      "step": 78
+    },
+    {
+      "epoch": 0.005809250680197073,
+      "grad_norm": 0.8209432363510132,
+      "learning_rate": 3.8235294117647055e-05,
+      "loss": 3.5169,
+      "step": 79
+    },
+    {
+      "epoch": 0.005882785498933745,
+      "grad_norm": 0.6916273236274719,
+      "learning_rate": 3.872549019607844e-05,
+      "loss": 3.4468,
+      "step": 80
+    },
+    {
+      "epoch": 0.005956320317670417,
+      "grad_norm": 0.762651801109314,
+      "learning_rate": 3.9215686274509805e-05,
+      "loss": 3.3797,
+      "step": 81
+    },
+    {
+      "epoch": 0.0060298551364070885,
+      "grad_norm": 0.717921793460846,
+      "learning_rate": 3.970588235294117e-05,
+      "loss": 3.3165,
+      "step": 82
+    },
+    {
+      "epoch": 0.006103389955143761,
+      "grad_norm": 0.8078845143318176,
+      "learning_rate": 4.0196078431372555e-05,
+      "loss": 3.3439,
+      "step": 83
+    },
+    {
+      "epoch": 0.006176924773880432,
+      "grad_norm": 0.7795608043670654,
+      "learning_rate": 4.068627450980392e-05,
+      "loss": 3.3876,
+      "step": 84
+    },
+    {
+      "epoch": 0.006250459592617105,
+      "grad_norm": 1.0330259799957275,
+      "learning_rate": 4.11764705882353e-05,
+      "loss": 3.2992,
+      "step": 85
+    },
+    {
+      "epoch": 0.006323994411353776,
+      "grad_norm": 0.7333192229270935,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 3.2663,
+      "step": 86
+    },
+    {
+      "epoch": 0.0063975292300904475,
+      "grad_norm": 1.1012221574783325,
+      "learning_rate": 4.215686274509804e-05,
+      "loss": 3.2605,
+      "step": 87
+    },
+    {
+      "epoch": 0.00647106404882712,
+      "grad_norm": 0.696049690246582,
+      "learning_rate": 4.2647058823529415e-05,
+      "loss": 3.1883,
+      "step": 88
+    },
+    {
+      "epoch": 0.006544598867563791,
+      "grad_norm": 0.8208140134811401,
+      "learning_rate": 4.313725490196079e-05,
+      "loss": 3.2478,
+      "step": 89
+    },
+    {
+      "epoch": 0.0066181336863004635,
+      "grad_norm": 0.9138851761817932,
+      "learning_rate": 4.362745098039216e-05,
+      "loss": 3.2191,
+      "step": 90
+    },
+    {
+      "epoch": 0.006691668505037135,
+      "grad_norm": 0.7101198434829712,
+      "learning_rate": 4.411764705882353e-05,
+      "loss": 3.1605,
+      "step": 91
+    },
+    {
+      "epoch": 0.006765203323773807,
+      "grad_norm": 0.6540734767913818,
+      "learning_rate": 4.460784313725491e-05,
+      "loss": 3.1853,
+      "step": 92
+    },
+    {
+      "epoch": 0.006838738142510479,
+      "grad_norm": 0.6084160208702087,
+      "learning_rate": 4.5098039215686275e-05,
+      "loss": 3.1701,
+      "step": 93
+    },
+    {
+      "epoch": 0.00691227296124715,
+      "grad_norm": 0.6793096661567688,
+      "learning_rate": 4.558823529411765e-05,
+      "loss": 3.104,
+      "step": 94
+    },
+    {
+      "epoch": 0.0069858077799838224,
+      "grad_norm": 0.6092313528060913,
+      "learning_rate": 4.607843137254902e-05,
+      "loss": 3.0892,
+      "step": 95
+    },
+    {
+      "epoch": 0.007059342598720494,
+      "grad_norm": 0.581619381904602,
+      "learning_rate": 4.656862745098039e-05,
+      "loss": 3.1345,
+      "step": 96
+    },
+    {
+      "epoch": 0.007132877417457166,
+      "grad_norm": 0.6190534234046936,
+      "learning_rate": 4.705882352941177e-05,
+      "loss": 3.1498,
+      "step": 97
+    },
+    {
+      "epoch": 0.007206412236193838,
+      "grad_norm": 0.5374878644943237,
+      "learning_rate": 4.7549019607843135e-05,
+      "loss": 3.0334,
+      "step": 98
+    },
+    {
+      "epoch": 0.00727994705493051,
+      "grad_norm": 0.818023681640625,
+      "learning_rate": 4.803921568627452e-05,
+      "loss": 3.0788,
+      "step": 99
+    },
+    {
+      "epoch": 0.007353481873667181,
+      "grad_norm": 0.975313127040863,
+      "learning_rate": 4.8529411764705885e-05,
+      "loss": 3.057,
+      "step": 100
+    },
+    {
+      "epoch": 0.007427016692403853,
+      "grad_norm": 0.8088687658309937,
+      "learning_rate": 4.901960784313725e-05,
+      "loss": 2.9335,
+      "step": 101
+    },
+    {
+      "epoch": 0.007500551511140525,
+      "grad_norm": 0.5294756889343262,
+      "learning_rate": 4.9509803921568634e-05,
+      "loss": 2.9822,
+      "step": 102
+    },
+    {
+      "epoch": 0.007574086329877197,
+      "grad_norm": 0.7083865404129028,
+      "learning_rate": 5e-05,
+      "loss": 2.9954,
+      "step": 103
+    },
+    {
+      "epoch": 0.007647621148613869,
+      "grad_norm": 0.6039468050003052,
+      "learning_rate": 5.049019607843137e-05,
+      "loss": 2.9165,
+      "step": 104
+    },
+    {
+      "epoch": 0.00772115596735054,
+      "grad_norm": 0.6230583786964417,
+      "learning_rate": 5.0980392156862745e-05,
+      "loss": 3.025,
+      "step": 105
+    },
+    {
+      "epoch": 0.007794690786087213,
+      "grad_norm": 0.5654554963111877,
+      "learning_rate": 5.147058823529411e-05,
+      "loss": 2.9435,
+      "step": 106
+    },
+    {
+      "epoch": 0.007868225604823885,
+      "grad_norm": 0.820525050163269,
+      "learning_rate": 5.1960784313725495e-05,
+      "loss": 2.918,
+      "step": 107
+    },
+    {
+      "epoch": 0.007941760423560556,
+      "grad_norm": 1.1012895107269287,
+      "learning_rate": 5.245098039215687e-05,
+      "loss": 2.9456,
+      "step": 108
+    },
+    {
+      "epoch": 0.008015295242297228,
+      "grad_norm": 0.5024237632751465,
+      "learning_rate": 5.294117647058824e-05,
+      "loss": 2.8843,
+      "step": 109
+    },
+    {
+      "epoch": 0.0080888300610339,
+      "grad_norm": 0.7928162813186646,
+      "learning_rate": 5.343137254901961e-05,
+      "loss": 2.883,
+      "step": 110
+    },
+    {
+      "epoch": 0.00816236487977057,
+      "grad_norm": 0.542624831199646,
+      "learning_rate": 5.392156862745098e-05,
+      "loss": 2.8545,
+      "step": 111
+    },
+    {
+      "epoch": 0.008235899698507243,
+      "grad_norm": 0.6043968200683594,
+      "learning_rate": 5.441176470588235e-05,
+      "loss": 2.8252,
+      "step": 112
+    },
+    {
+      "epoch": 0.008309434517243915,
+      "grad_norm": 0.6008192896842957,
+      "learning_rate": 5.490196078431373e-05,
+      "loss": 2.8056,
+      "step": 113
+    },
+    {
+      "epoch": 0.008382969335980588,
+      "grad_norm": 0.750988245010376,
+      "learning_rate": 5.5392156862745104e-05,
+      "loss": 2.8196,
+      "step": 114
+    },
+    {
+      "epoch": 0.008456504154717258,
+      "grad_norm": 0.5485605001449585,
+      "learning_rate": 5.588235294117647e-05,
+      "loss": 2.7808,
+      "step": 115
+    },
+    {
+      "epoch": 0.00853003897345393,
+      "grad_norm": 1.0856144428253174,
+      "learning_rate": 5.637254901960785e-05,
+      "loss": 2.7626,
+      "step": 116
+    },
+    {
+      "epoch": 0.008603573792190603,
+      "grad_norm": 0.891465425491333,
+      "learning_rate": 5.6862745098039215e-05,
+      "loss": 2.7837,
+      "step": 117
+    },
+    {
+      "epoch": 0.008677108610927273,
+      "grad_norm": 0.5742112398147583,
+      "learning_rate": 5.735294117647059e-05,
+      "loss": 2.7846,
+      "step": 118
+    },
+    {
+      "epoch": 0.008750643429663946,
+      "grad_norm": 0.5832791328430176,
+      "learning_rate": 5.784313725490197e-05,
+      "loss": 2.7258,
+      "step": 119
+    },
+    {
+      "epoch": 0.008824178248400618,
+      "grad_norm": 0.7935657501220703,
+      "learning_rate": 5.833333333333334e-05,
+      "loss": 2.7539,
+      "step": 120
+    },
+    {
+      "epoch": 0.00889771306713729,
+      "grad_norm": 1.4480737447738647,
+      "learning_rate": 5.882352941176471e-05,
+      "loss": 2.6833,
+      "step": 121
+    },
+    {
+      "epoch": 0.008971247885873961,
+      "grad_norm": 0.6163828372955322,
+      "learning_rate": 5.931372549019608e-05,
+      "loss": 2.6938,
+      "step": 122
+    },
+    {
+      "epoch": 0.009044782704610633,
+      "grad_norm": 0.6668141484260559,
+      "learning_rate": 5.980392156862745e-05,
+      "loss": 2.725,
+      "step": 123
+    },
+    {
+      "epoch": 0.009118317523347306,
+      "grad_norm": 0.7240054607391357,
+      "learning_rate": 6.0294117647058825e-05,
+      "loss": 2.6212,
+      "step": 124
+    },
+    {
+      "epoch": 0.009191852342083976,
+      "grad_norm": 0.42524638772010803,
+      "learning_rate": 6.078431372549019e-05,
+      "loss": 2.6904,
+      "step": 125
+    },
+    {
+      "epoch": 0.009265387160820648,
+      "grad_norm": 0.8101039528846741,
+      "learning_rate": 6.127450980392158e-05,
+      "loss": 2.6704,
+      "step": 126
+    },
+    {
+      "epoch": 0.00933892197955732,
+      "grad_norm": 0.5232560038566589,
+      "learning_rate": 6.176470588235295e-05,
+      "loss": 2.7004,
+      "step": 127
+    },
+    {
+      "epoch": 0.009412456798293993,
+      "grad_norm": 0.41690272092819214,
+      "learning_rate": 6.225490196078432e-05,
+      "loss": 2.6232,
+      "step": 128
+    },
+    {
+      "epoch": 0.009485991617030664,
+      "grad_norm": 0.5982264876365662,
+      "learning_rate": 6.274509803921569e-05,
+      "loss": 2.536,
+      "step": 129
+    },
+    {
+      "epoch": 0.009559526435767336,
+      "grad_norm": 0.44974029064178467,
+      "learning_rate": 6.323529411764705e-05,
+      "loss": 2.5842,
+      "step": 130
+    },
+    {
+      "epoch": 0.009633061254504008,
+      "grad_norm": 0.41083669662475586,
+      "learning_rate": 6.372549019607843e-05,
+      "loss": 2.5606,
+      "step": 131
+    },
+    {
+      "epoch": 0.009706596073240679,
+      "grad_norm": 0.51421719789505,
+      "learning_rate": 6.421568627450982e-05,
+      "loss": 2.5744,
+      "step": 132
+    },
+    {
+      "epoch": 0.009780130891977351,
+      "grad_norm": 0.4501510262489319,
+      "learning_rate": 6.470588235294118e-05,
+      "loss": 2.6038,
+      "step": 133
+    },
+    {
+      "epoch": 0.009853665710714023,
+      "grad_norm": 0.4612300992012024,
+      "learning_rate": 6.519607843137255e-05,
+      "loss": 2.579,
+      "step": 134
+    },
+    {
+      "epoch": 0.009927200529450696,
+      "grad_norm": 0.4287264347076416,
+      "learning_rate": 6.568627450980392e-05,
+      "loss": 2.535,
+      "step": 135
+    },
+    {
+      "epoch": 0.010000735348187366,
+      "grad_norm": 0.6688863039016724,
+      "learning_rate": 6.61764705882353e-05,
+      "loss": 2.546,
+      "step": 136
+    },
+    {
+      "epoch": 0.010074270166924039,
+      "grad_norm": 0.37818267941474915,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 2.5349,
+      "step": 137
+    },
+    {
+      "epoch": 0.010147804985660711,
+      "grad_norm": 0.4656243920326233,
+      "learning_rate": 6.715686274509804e-05,
+      "loss": 2.4943,
+      "step": 138
+    },
+    {
+      "epoch": 0.010221339804397381,
+      "grad_norm": 0.32601550221443176,
+      "learning_rate": 6.764705882352942e-05,
+      "loss": 2.5056,
+      "step": 139
+    },
+    {
+      "epoch": 0.010294874623134054,
+      "grad_norm": 0.5192354917526245,
+      "learning_rate": 6.813725490196079e-05,
+      "loss": 2.4995,
+      "step": 140
+    },
+    {
+      "epoch": 0.010368409441870726,
+      "grad_norm": 0.7013667225837708,
+      "learning_rate": 6.862745098039216e-05,
+      "loss": 2.4951,
+      "step": 141
+    },
+    {
+      "epoch": 0.010441944260607398,
+      "grad_norm": 0.34450435638427734,
+      "learning_rate": 6.911764705882354e-05,
+      "loss": 2.453,
+      "step": 142
+    },
+    {
+      "epoch": 0.010515479079344069,
+      "grad_norm": 0.42529910802841187,
+      "learning_rate": 6.96078431372549e-05,
+      "loss": 2.4945,
+      "step": 143
+    },
+    {
+      "epoch": 0.010589013898080741,
+      "grad_norm": 0.3829617202281952,
+      "learning_rate": 7.009803921568627e-05,
+      "loss": 2.4705,
+      "step": 144
+    },
+    {
+      "epoch": 0.010662548716817414,
+      "grad_norm": 0.374667763710022,
+      "learning_rate": 7.058823529411765e-05,
+      "loss": 2.4509,
+      "step": 145
+    },
+    {
+      "epoch": 0.010736083535554084,
+      "grad_norm": 0.3648475706577301,
+      "learning_rate": 7.107843137254902e-05,
+      "loss": 2.4296,
+      "step": 146
+    },
+    {
+      "epoch": 0.010809618354290756,
+      "grad_norm": 0.585560142993927,
+      "learning_rate": 7.156862745098039e-05,
+      "loss": 2.4374,
+      "step": 147
+    },
+    {
+      "epoch": 0.010883153173027429,
+      "grad_norm": 0.43878328800201416,
+      "learning_rate": 7.205882352941177e-05,
+      "loss": 2.4446,
+      "step": 148
+    },
+    {
+      "epoch": 0.010956687991764101,
+      "grad_norm": 0.4675384759902954,
+      "learning_rate": 7.254901960784314e-05,
+      "loss": 2.3798,
+      "step": 149
+    },
+    {
+      "epoch": 0.011030222810500772,
+      "grad_norm": 0.5181499719619751,
+      "learning_rate": 7.303921568627451e-05,
+      "loss": 2.3857,
+      "step": 150
+    },
+    {
+      "epoch": 0.011103757629237444,
+      "grad_norm": 0.5040928721427917,
+      "learning_rate": 7.352941176470589e-05,
+      "loss": 2.3475,
+      "step": 151
+    },
+    {
+      "epoch": 0.011177292447974116,
+      "grad_norm": 0.44567710161209106,
+      "learning_rate": 7.401960784313726e-05,
+      "loss": 2.3078,
+      "step": 152
+    },
+    {
+      "epoch": 0.011250827266710787,
+      "grad_norm": 0.460347443819046,
+      "learning_rate": 7.450980392156864e-05,
+      "loss": 2.3069,
+      "step": 153
+    },
+    {
+      "epoch": 0.01132436208544746,
+      "grad_norm": 0.4324535131454468,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 2.2817,
+      "step": 154
+    },
+    {
+      "epoch": 0.011397896904184131,
+      "grad_norm": 0.43691062927246094,
+      "learning_rate": 7.549019607843137e-05,
+      "loss": 2.3179,
+      "step": 155
+    },
+    {
+      "epoch": 0.011471431722920804,
+      "grad_norm": 0.3476044237613678,
+      "learning_rate": 7.598039215686274e-05,
+      "loss": 2.3033,
+      "step": 156
+    },
+    {
+      "epoch": 0.011544966541657474,
+      "grad_norm": 0.35938164591789246,
+      "learning_rate": 7.647058823529411e-05,
+      "loss": 2.2947,
+      "step": 157
+    },
+    {
+      "epoch": 0.011618501360394147,
+      "grad_norm": 0.3413945734500885,
+      "learning_rate": 7.696078431372549e-05,
+      "loss": 2.2253,
+      "step": 158
+    },
+    {
+      "epoch": 0.011692036179130819,
+      "grad_norm": 0.3611355423927307,
+      "learning_rate": 7.745098039215687e-05,
+      "loss": 2.1561,
+      "step": 159
+    },
+    {
+      "epoch": 0.01176557099786749,
+      "grad_norm": 0.28506579995155334,
+      "learning_rate": 7.794117647058824e-05,
+      "loss": 2.1241,
+      "step": 160
+    },
+    {
+      "epoch": 0.011839105816604162,
+      "grad_norm": 0.39731264114379883,
+      "learning_rate": 7.843137254901961e-05,
+      "loss": 2.1294,
+      "step": 161
+    },
+    {
+      "epoch": 0.011912640635340834,
+      "grad_norm": 0.3432599902153015,
+      "learning_rate": 7.892156862745098e-05,
+      "loss": 2.1016,
+      "step": 162
+    },
+    {
+      "epoch": 0.011986175454077506,
+      "grad_norm": 0.4184660017490387,
+      "learning_rate": 7.941176470588235e-05,
+      "loss": 2.1039,
+      "step": 163
+    },
+    {
+      "epoch": 0.012059710272814177,
+      "grad_norm": 0.4879499077796936,
+      "learning_rate": 7.990196078431373e-05,
+      "loss": 2.1649,
+      "step": 164
+    },
+    {
+      "epoch": 0.01213324509155085,
+      "grad_norm": 0.3443756401538849,
+      "learning_rate": 8.039215686274511e-05,
+      "loss": 2.0982,
+      "step": 165
+    },
+    {
+      "epoch": 0.012206779910287522,
+      "grad_norm": 0.2792808413505554,
+      "learning_rate": 8.088235294117648e-05,
+      "loss": 2.0655,
+      "step": 166
+    },
+    {
+      "epoch": 0.012280314729024192,
+      "grad_norm": 0.3022691607475281,
+      "learning_rate": 8.137254901960785e-05,
+      "loss": 2.0588,
+      "step": 167
+    },
+    {
+      "epoch": 0.012353849547760865,
+      "grad_norm": 0.2800402045249939,
+      "learning_rate": 8.186274509803921e-05,
+      "loss": 2.0664,
+      "step": 168
+    },
+    {
+      "epoch": 0.012427384366497537,
+      "grad_norm": 0.2736533582210541,
+      "learning_rate": 8.23529411764706e-05,
+      "loss": 2.0248,
+      "step": 169
+    },
+    {
+      "epoch": 0.01250091918523421,
+      "grad_norm": 0.2967151701450348,
+      "learning_rate": 8.284313725490198e-05,
+      "loss": 2.0596,
+      "step": 170
+    },
+    {
+      "epoch": 0.01257445400397088,
+      "grad_norm": 0.2552861273288727,
+      "learning_rate": 8.333333333333334e-05,
+      "loss": 2.051,
+      "step": 171
+    },
+    {
+      "epoch": 0.012647988822707552,
+      "grad_norm": 0.32275858521461487,
+      "learning_rate": 8.382352941176471e-05,
+      "loss": 2.0326,
+      "step": 172
+    },
+    {
+      "epoch": 0.012721523641444224,
+      "grad_norm": 0.26918187737464905,
+      "learning_rate": 8.431372549019608e-05,
+      "loss": 2.0183,
+      "step": 173
+    },
+    {
+      "epoch": 0.012795058460180895,
+      "grad_norm": 0.3527700901031494,
+      "learning_rate": 8.480392156862745e-05,
+      "loss": 2.003,
+      "step": 174
+    },
+    {
+      "epoch": 0.012868593278917567,
+      "grad_norm": 0.5925028324127197,
+      "learning_rate": 8.529411764705883e-05,
+      "loss": 2.054,
+      "step": 175
+    },
+    {
+      "epoch": 0.01294212809765424,
+      "grad_norm": 0.44507408142089844,
+      "learning_rate": 8.57843137254902e-05,
+      "loss": 1.9904,
+      "step": 176
+    },
+    {
+      "epoch": 0.013015662916390912,
+      "grad_norm": 0.2849460244178772,
+      "learning_rate": 8.627450980392158e-05,
+      "loss": 1.9717,
+      "step": 177
+    },
+    {
+      "epoch": 0.013089197735127582,
+      "grad_norm": 0.36799147725105286,
+      "learning_rate": 8.676470588235295e-05,
+      "loss": 1.9667,
+      "step": 178
+    },
+    {
+      "epoch": 0.013162732553864255,
+      "grad_norm": 0.2681286036968231,
+      "learning_rate": 8.725490196078432e-05,
+      "loss": 1.933,
+      "step": 179
+    },
+    {
+      "epoch": 0.013236267372600927,
+      "grad_norm": 0.477252334356308,
+      "learning_rate": 8.774509803921568e-05,
+      "loss": 1.9772,
+      "step": 180
+    },
+    {
+      "epoch": 0.013309802191337598,
+      "grad_norm": 0.3011188209056854,
+      "learning_rate": 8.823529411764706e-05,
+      "loss": 1.9324,
+      "step": 181
+    },
+    {
+      "epoch": 0.01338333701007427,
+      "grad_norm": 0.38431316614151,
+      "learning_rate": 8.872549019607843e-05,
+      "loss": 2.0002,
+      "step": 182
+    },
+    {
+      "epoch": 0.013456871828810942,
+      "grad_norm": 0.29670026898384094,
+      "learning_rate": 8.921568627450981e-05,
+      "loss": 1.9656,
+      "step": 183
+    },
+    {
+      "epoch": 0.013530406647547615,
+      "grad_norm": 0.31347739696502686,
+      "learning_rate": 8.970588235294118e-05,
+      "loss": 1.9514,
+      "step": 184
+    },
+    {
+      "epoch": 0.013603941466284285,
+      "grad_norm": 0.2673075199127197,
+      "learning_rate": 9.019607843137255e-05,
+      "loss": 1.9498,
+      "step": 185
+    },
+    {
+      "epoch": 0.013677476285020957,
+      "grad_norm": 0.3136028051376343,
+      "learning_rate": 9.068627450980393e-05,
+      "loss": 1.9721,
+      "step": 186
+    },
+    {
+      "epoch": 0.01375101110375763,
+      "grad_norm": 0.5579624772071838,
+      "learning_rate": 9.11764705882353e-05,
+      "loss": 1.9239,
+      "step": 187
+    },
+    {
+      "epoch": 0.0138245459224943,
+      "grad_norm": 0.3856685757637024,
+      "learning_rate": 9.166666666666667e-05,
+      "loss": 1.9332,
+      "step": 188
+    },
+    {
+      "epoch": 0.013898080741230973,
+      "grad_norm": 0.28228023648262024,
+      "learning_rate": 9.215686274509804e-05,
+      "loss": 1.9108,
+      "step": 189
+    },
+    {
+      "epoch": 0.013971615559967645,
+      "grad_norm": 0.4346674680709839,
+      "learning_rate": 9.264705882352942e-05,
+      "loss": 1.92,
+      "step": 190
+    },
+    {
+      "epoch": 0.014045150378704317,
+      "grad_norm": 0.351190447807312,
+      "learning_rate": 9.313725490196079e-05,
+      "loss": 1.9285,
+      "step": 191
+    },
+    {
+      "epoch": 0.014118685197440988,
+      "grad_norm": 0.21171818673610687,
+      "learning_rate": 9.362745098039217e-05,
+      "loss": 1.9293,
+      "step": 192
+    },
+    {
+      "epoch": 0.01419222001617766,
+      "grad_norm": 0.3983679413795471,
+      "learning_rate": 9.411764705882353e-05,
+      "loss": 1.9897,
+      "step": 193
+    },
+    {
+      "epoch": 0.014265754834914332,
+      "grad_norm": 0.3011804223060608,
+      "learning_rate": 9.46078431372549e-05,
+      "loss": 1.9293,
+      "step": 194
+    },
+    {
+      "epoch": 0.014339289653651003,
+      "grad_norm": 0.2575826346874237,
+      "learning_rate": 9.509803921568627e-05,
+      "loss": 1.908,
+      "step": 195
+    },
+    {
+      "epoch": 0.014412824472387675,
+      "grad_norm": 0.47639793157577515,
+      "learning_rate": 9.558823529411765e-05,
+      "loss": 1.881,
+      "step": 196
+    },
+    {
+      "epoch": 0.014486359291124348,
+      "grad_norm": 0.37741199135780334,
+      "learning_rate": 9.607843137254903e-05,
+      "loss": 1.9267,
+      "step": 197
+    },
+    {
+      "epoch": 0.01455989410986102,
+      "grad_norm": 0.3432460129261017,
+      "learning_rate": 9.65686274509804e-05,
+      "loss": 1.9186,
+      "step": 198
+    },
+    {
+      "epoch": 0.01463342892859769,
+      "grad_norm": 0.9100985527038574,
+      "learning_rate": 9.705882352941177e-05,
+      "loss": 1.8975,
+      "step": 199
+    },
+    {
+      "epoch": 0.014706963747334363,
+      "grad_norm": 0.7420558333396912,
+      "learning_rate": 9.754901960784314e-05,
+      "loss": 1.891,
+      "step": 200
+    },
+    {
+      "epoch": 0.014780498566071035,
+      "grad_norm": 0.3264927864074707,
+      "learning_rate": 9.80392156862745e-05,
+      "loss": 1.9087,
+      "step": 201
+    },
+    {
+      "epoch": 0.014854033384807706,
+      "grad_norm": 0.7239006161689758,
+      "learning_rate": 9.852941176470589e-05,
+      "loss": 1.9489,
+      "step": 202
+    },
+    {
+      "epoch": 0.014927568203544378,
+      "grad_norm": 0.18545539677143097,
+      "learning_rate": 9.901960784313727e-05,
+      "loss": 1.8945,
+      "step": 203
+    },
+    {
+      "epoch": 0.01500110302228105,
+      "grad_norm": 0.6481702923774719,
+      "learning_rate": 9.950980392156864e-05,
+      "loss": 1.8781,
+      "step": 204
+    },
+    {
+      "epoch": 0.015074637841017723,
+      "grad_norm": 0.32571813464164734,
+      "learning_rate": 0.0001,
+      "loss": 1.8735,
+      "step": 205
+    },
+    {
+      "epoch": 0.015148172659754393,
+      "grad_norm": 0.41916728019714355,
+      "learning_rate": 0.00010049019607843137,
+      "loss": 1.9214,
+      "step": 206
+    },
+    {
+      "epoch": 0.015221707478491065,
+      "grad_norm": 0.3117308020591736,
+      "learning_rate": 0.00010098039215686274,
+      "loss": 1.8516,
+      "step": 207
+    },
+    {
+      "epoch": 0.015295242297227738,
+      "grad_norm": 0.3670138418674469,
+      "learning_rate": 0.00010147058823529412,
+      "loss": 1.8843,
+      "step": 208
+    },
+    {
+      "epoch": 0.015368777115964408,
+      "grad_norm": 0.4449373781681061,
+      "learning_rate": 0.00010196078431372549,
+      "loss": 1.8576,
+      "step": 209
+    },
+    {
+      "epoch": 0.01544231193470108,
+      "grad_norm": 0.1805167943239212,
+      "learning_rate": 0.00010245098039215686,
+      "loss": 1.8781,
+      "step": 210
+    },
+    {
+      "epoch": 0.015515846753437753,
+      "grad_norm": 0.3232683837413788,
+      "learning_rate": 0.00010294117647058823,
+      "loss": 1.8588,
+      "step": 211
+    },
+    {
+      "epoch": 0.015589381572174425,
+      "grad_norm": 0.19792911410331726,
+      "learning_rate": 0.00010343137254901962,
+      "loss": 1.8691,
+      "step": 212
+    },
+    {
+      "epoch": 0.015662916390911098,
+      "grad_norm": 0.21309353411197662,
+      "learning_rate": 0.00010392156862745099,
+      "loss": 1.8859,
+      "step": 213
+    },
+    {
+      "epoch": 0.01573645120964777,
+      "grad_norm": 0.20114189386367798,
+      "learning_rate": 0.00010441176470588237,
+      "loss": 1.8676,
+      "step": 214
+    },
+    {
+      "epoch": 0.01580998602838444,
+      "grad_norm": 0.2236410677433014,
+      "learning_rate": 0.00010490196078431374,
+      "loss": 1.8386,
+      "step": 215
+    },
+    {
+      "epoch": 0.01588352084712111,
+      "grad_norm": 0.3259761929512024,
+      "learning_rate": 0.00010539215686274511,
+      "loss": 1.807,
+      "step": 216
+    },
+    {
+      "epoch": 0.015957055665857783,
+      "grad_norm": 0.4042021930217743,
+      "learning_rate": 0.00010588235294117647,
+      "loss": 1.7895,
+      "step": 217
+    },
+    {
+      "epoch": 0.016030590484594456,
+      "grad_norm": 0.20548312366008759,
+      "learning_rate": 0.00010637254901960784,
+      "loss": 1.7854,
+      "step": 218
+    },
+    {
+      "epoch": 0.016104125303331128,
+      "grad_norm": 0.30719080567359924,
+      "learning_rate": 0.00010686274509803922,
+      "loss": 1.8355,
+      "step": 219
+    },
+    {
+      "epoch": 0.0161776601220678,
+      "grad_norm": 0.272167831659317,
+      "learning_rate": 0.00010735294117647059,
+      "loss": 1.7918,
+      "step": 220
+    },
+    {
+      "epoch": 0.016251194940804473,
+      "grad_norm": 0.24975953996181488,
+      "learning_rate": 0.00010784313725490196,
+      "loss": 1.8217,
+      "step": 221
+    },
+    {
+      "epoch": 0.01632472975954114,
+      "grad_norm": 0.2422274649143219,
+      "learning_rate": 0.00010833333333333333,
+      "loss": 1.8331,
+      "step": 222
+    },
+    {
+      "epoch": 0.016398264578277814,
+      "grad_norm": 0.2918996512889862,
+      "learning_rate": 0.0001088235294117647,
+      "loss": 1.7626,
+      "step": 223
+    },
+    {
+      "epoch": 0.016471799397014486,
+      "grad_norm": 0.19592070579528809,
+      "learning_rate": 0.00010931372549019608,
+      "loss": 1.7791,
+      "step": 224
+    },
+    {
+      "epoch": 0.01654533421575116,
+      "grad_norm": 0.21093954145908356,
+      "learning_rate": 0.00010980392156862746,
+      "loss": 1.7452,
+      "step": 225
+    },
+    {
+      "epoch": 0.01661886903448783,
+      "grad_norm": 0.23876960575580597,
+      "learning_rate": 0.00011029411764705884,
+      "loss": 1.7807,
+      "step": 226
+    },
+    {
+      "epoch": 0.016692403853224503,
+      "grad_norm": 0.21197018027305603,
+      "learning_rate": 0.00011078431372549021,
+      "loss": 1.759,
+      "step": 227
+    },
+    {
+      "epoch": 0.016765938671961175,
+      "grad_norm": 0.22616982460021973,
+      "learning_rate": 0.00011127450980392158,
+      "loss": 1.7346,
+      "step": 228
+    },
+    {
+      "epoch": 0.016839473490697844,
+      "grad_norm": 0.2368088811635971,
+      "learning_rate": 0.00011176470588235294,
+      "loss": 1.7787,
+      "step": 229
+    },
+    {
+      "epoch": 0.016913008309434516,
+      "grad_norm": 0.35544857382774353,
+      "learning_rate": 0.00011225490196078433,
+      "loss": 1.7711,
+      "step": 230
+    },
+    {
+      "epoch": 0.01698654312817119,
+      "grad_norm": 0.37238553166389465,
+      "learning_rate": 0.0001127450980392157,
+      "loss": 1.7567,
+      "step": 231
+    },
+    {
+      "epoch": 0.01706007794690786,
+      "grad_norm": 0.19599607586860657,
+      "learning_rate": 0.00011323529411764706,
+      "loss": 1.7711,
+      "step": 232
+    },
+    {
+      "epoch": 0.017133612765644533,
+      "grad_norm": 0.20085231959819794,
+      "learning_rate": 0.00011372549019607843,
+      "loss": 1.7525,
+      "step": 233
+    },
+    {
+      "epoch": 0.017207147584381206,
+      "grad_norm": 0.21313263475894928,
+      "learning_rate": 0.0001142156862745098,
+      "loss": 1.7579,
+      "step": 234
+    },
+    {
+      "epoch": 0.017280682403117878,
+      "grad_norm": 0.17997272312641144,
+      "learning_rate": 0.00011470588235294118,
+      "loss": 1.7486,
+      "step": 235
+    },
+    {
+      "epoch": 0.017354217221854547,
+      "grad_norm": 0.20306064188480377,
+      "learning_rate": 0.00011519607843137255,
+      "loss": 1.7951,
+      "step": 236
+    },
+    {
+      "epoch": 0.01742775204059122,
+      "grad_norm": 0.26720717549324036,
+      "learning_rate": 0.00011568627450980394,
+      "loss": 1.7214,
+      "step": 237
+    },
+    {
+      "epoch": 0.01750128685932789,
+      "grad_norm": 0.21869398653507233,
+      "learning_rate": 0.00011617647058823531,
+      "loss": 1.7304,
+      "step": 238
+    },
+    {
+      "epoch": 0.017574821678064564,
+      "grad_norm": 0.38471007347106934,
+      "learning_rate": 0.00011666666666666668,
+      "loss": 1.7241,
+      "step": 239
+    },
+    {
+      "epoch": 0.017648356496801236,
+      "grad_norm": 0.334971159696579,
+      "learning_rate": 0.00011715686274509805,
+      "loss": 1.7114,
+      "step": 240
+    },
+    {
+      "epoch": 0.01772189131553791,
+      "grad_norm": 0.1807050257921219,
+      "learning_rate": 0.00011764705882352942,
+      "loss": 1.7521,
+      "step": 241
+    },
+    {
+      "epoch": 0.01779542613427458,
+      "grad_norm": 0.3371660113334656,
+      "learning_rate": 0.0001181372549019608,
+      "loss": 1.7181,
+      "step": 242
+    },
+    {
+      "epoch": 0.01786896095301125,
+      "grad_norm": 0.4160744249820709,
+      "learning_rate": 0.00011862745098039216,
+      "loss": 1.728,
+      "step": 243
+    },
+    {
+      "epoch": 0.017942495771747922,
+      "grad_norm": 0.3168802857398987,
+      "learning_rate": 0.00011911764705882353,
+      "loss": 1.7,
+      "step": 244
+    },
+    {
+      "epoch": 0.018016030590484594,
+      "grad_norm": 0.18611112236976624,
+      "learning_rate": 0.0001196078431372549,
+      "loss": 1.729,
+      "step": 245
+    },
+    {
+      "epoch": 0.018089565409221266,
+      "grad_norm": 0.37810763716697693,
+      "learning_rate": 0.00012009803921568628,
+      "loss": 1.7224,
+      "step": 246
+    },
+    {
+      "epoch": 0.01816310022795794,
+      "grad_norm": 0.24156223237514496,
+      "learning_rate": 0.00012058823529411765,
+      "loss": 1.7735,
+      "step": 247
+    },
+    {
+      "epoch": 0.01823663504669461,
+      "grad_norm": 0.20902138948440552,
+      "learning_rate": 0.00012107843137254902,
+      "loss": 1.7236,
+      "step": 248
+    },
+    {
+      "epoch": 0.018310169865431283,
+      "grad_norm": 0.20123736560344696,
+      "learning_rate": 0.00012156862745098039,
+      "loss": 1.6916,
+      "step": 249
+    },
+    {
+      "epoch": 0.018383704684167952,
+      "grad_norm": 0.18057729303836823,
+      "learning_rate": 0.00012205882352941178,
+      "loss": 1.7111,
+      "step": 250
+    },
+    {
+      "epoch": 0.018457239502904624,
+      "grad_norm": 0.2886579930782318,
+      "learning_rate": 0.00012254901960784316,
+      "loss": 1.7338,
+      "step": 251
+    },
+    {
+      "epoch": 0.018530774321641297,
+      "grad_norm": 0.25238513946533203,
+      "learning_rate": 0.00012303921568627453,
+      "loss": 1.6763,
+      "step": 252
+    },
+    {
+      "epoch": 0.01860430914037797,
+      "grad_norm": 0.186184823513031,
+      "learning_rate": 0.0001235294117647059,
+      "loss": 1.7009,
+      "step": 253
+    },
+    {
+      "epoch": 0.01867784395911464,
+      "grad_norm": 0.23223787546157837,
+      "learning_rate": 0.00012401960784313727,
+      "loss": 1.7085,
+      "step": 254
+    },
+    {
+      "epoch": 0.018751378777851314,
+      "grad_norm": 0.3164445161819458,
+      "learning_rate": 0.00012450980392156863,
+      "loss": 1.7234,
+      "step": 255
+    },
+    {
+      "epoch": 0.018824913596587986,
+      "grad_norm": 0.23090463876724243,
+      "learning_rate": 0.000125,
+      "loss": 1.7042,
+      "step": 256
+    },
+    {
+      "epoch": 0.018898448415324655,
+      "grad_norm": 0.23832933604717255,
+      "learning_rate": 0.00012549019607843137,
+      "loss": 1.7639,
+      "step": 257
+    },
+    {
+      "epoch": 0.018971983234061327,
+      "grad_norm": 0.22112390398979187,
+      "learning_rate": 0.00012598039215686274,
+      "loss": 1.6888,
+      "step": 258
+    },
+    {
+      "epoch": 0.019045518052798,
+      "grad_norm": 0.1827714890241623,
+      "learning_rate": 0.0001264705882352941,
+      "loss": 1.7097,
+      "step": 259
+    },
+    {
+      "epoch": 0.019119052871534672,
+      "grad_norm": 0.3825037479400635,
+      "learning_rate": 0.0001269607843137255,
+      "loss": 1.6861,
+      "step": 260
+    },
+    {
+      "epoch": 0.019192587690271344,
+      "grad_norm": 0.5042731165885925,
+      "learning_rate": 0.00012745098039215687,
+      "loss": 1.6917,
+      "step": 261
+    },
+    {
+      "epoch": 0.019266122509008016,
+      "grad_norm": 0.29187434911727905,
+      "learning_rate": 0.00012794117647058824,
+      "loss": 1.7048,
+      "step": 262
+    },
+    {
+      "epoch": 0.01933965732774469,
+      "grad_norm": 0.2352616786956787,
+      "learning_rate": 0.00012843137254901963,
+      "loss": 1.6692,
+      "step": 263
+    },
+    {
+      "epoch": 0.019413192146481358,
+      "grad_norm": 0.4212090075016022,
+      "learning_rate": 0.000128921568627451,
+      "loss": 1.6902,
+      "step": 264
+    },
+    {
+      "epoch": 0.01948672696521803,
+      "grad_norm": 0.2832290232181549,
+      "learning_rate": 0.00012941176470588237,
+      "loss": 1.6816,
+      "step": 265
+    },
+    {
+      "epoch": 0.019560261783954702,
+      "grad_norm": 0.3578559160232544,
+      "learning_rate": 0.00012990196078431374,
+      "loss": 1.67,
+      "step": 266
+    },
+    {
+      "epoch": 0.019633796602691374,
+      "grad_norm": 0.6371394395828247,
+      "learning_rate": 0.0001303921568627451,
+      "loss": 1.7581,
+      "step": 267
+    },
+    {
+      "epoch": 0.019707331421428047,
+      "grad_norm": 0.5351107716560364,
+      "learning_rate": 0.00013088235294117647,
+      "loss": 1.6584,
+      "step": 268
+    },
+    {
+      "epoch": 0.01978086624016472,
+      "grad_norm": 0.24974121153354645,
+      "learning_rate": 0.00013137254901960784,
+      "loss": 1.722,
+      "step": 269
+    },
+    {
+      "epoch": 0.01985440105890139,
+      "grad_norm": 0.2302001714706421,
+      "learning_rate": 0.0001318627450980392,
+      "loss": 1.6977,
+      "step": 270
+    },
+    {
+      "epoch": 0.01992793587763806,
+      "grad_norm": 0.23371398448944092,
+      "learning_rate": 0.0001323529411764706,
+      "loss": 1.6909,
+      "step": 271
+    },
+    {
+      "epoch": 0.020001470696374733,
+      "grad_norm": 0.20866428315639496,
+      "learning_rate": 0.00013284313725490197,
+      "loss": 1.6578,
+      "step": 272
+    },
+    {
+      "epoch": 0.020075005515111405,
+      "grad_norm": 0.23985245823860168,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 1.6954,
+      "step": 273
+    },
+    {
+      "epoch": 0.020148540333848077,
+      "grad_norm": 0.2035207897424698,
+      "learning_rate": 0.0001338235294117647,
+      "loss": 1.7014,
+      "step": 274
+    },
+    {
+      "epoch": 0.02022207515258475,
+      "grad_norm": 0.26359865069389343,
+      "learning_rate": 0.00013431372549019608,
+      "loss": 1.6818,
+      "step": 275
+    },
+    {
+      "epoch": 0.020295609971321422,
+      "grad_norm": 0.4266003370285034,
+      "learning_rate": 0.00013480392156862747,
+      "loss": 1.6606,
+      "step": 276
+    },
+    {
+      "epoch": 0.020369144790058094,
+      "grad_norm": 0.35862836241722107,
+      "learning_rate": 0.00013529411764705884,
+      "loss": 1.7039,
+      "step": 277
+    },
+    {
+      "epoch": 0.020442679608794763,
+      "grad_norm": 0.22850853204727173,
+      "learning_rate": 0.0001357843137254902,
+      "loss": 1.6648,
+      "step": 278
+    },
+    {
+      "epoch": 0.020516214427531435,
+      "grad_norm": 0.16235023736953735,
+      "learning_rate": 0.00013627450980392157,
+      "loss": 1.651,
+      "step": 279
+    },
+    {
+      "epoch": 0.020589749246268108,
+      "grad_norm": 0.1857730895280838,
+      "learning_rate": 0.00013676470588235294,
+      "loss": 1.6604,
+      "step": 280
+    },
+    {
+      "epoch": 0.02066328406500478,
+      "grad_norm": 0.14600954949855804,
+      "learning_rate": 0.0001372549019607843,
+      "loss": 1.6596,
+      "step": 281
+    },
+    {
+      "epoch": 0.020736818883741452,
+      "grad_norm": 0.2198246568441391,
+      "learning_rate": 0.00013774509803921568,
+      "loss": 1.637,
+      "step": 282
+    },
+    {
+      "epoch": 0.020810353702478124,
+      "grad_norm": 0.25800272822380066,
+      "learning_rate": 0.00013823529411764707,
+      "loss": 1.6902,
+      "step": 283
+    },
+    {
+      "epoch": 0.020883888521214797,
+      "grad_norm": 0.2584097385406494,
+      "learning_rate": 0.00013872549019607844,
+      "loss": 1.6626,
+      "step": 284
+    },
+    {
+      "epoch": 0.020957423339951466,
+      "grad_norm": 0.3224467635154724,
+      "learning_rate": 0.0001392156862745098,
+      "loss": 1.6434,
+      "step": 285
+    },
+    {
+      "epoch": 0.021030958158688138,
+      "grad_norm": 0.33233165740966797,
+      "learning_rate": 0.00013970588235294118,
+      "loss": 1.6344,
+      "step": 286
+    },
+    {
+      "epoch": 0.02110449297742481,
+      "grad_norm": 0.19228670001029968,
+      "learning_rate": 0.00014019607843137255,
+      "loss": 1.6853,
+      "step": 287
+    },
+    {
+      "epoch": 0.021178027796161483,
+      "grad_norm": 0.18349063396453857,
+      "learning_rate": 0.00014068627450980394,
+      "loss": 1.6289,
+      "step": 288
+    },
+    {
+      "epoch": 0.021251562614898155,
+      "grad_norm": 0.3070968687534332,
+      "learning_rate": 0.0001411764705882353,
+      "loss": 1.5537,
+      "step": 289
+    },
+    {
+      "epoch": 0.021325097433634827,
+      "grad_norm": 0.3327500522136688,
+      "learning_rate": 0.00014166666666666668,
+      "loss": 1.5885,
+      "step": 290
+    },
+    {
+      "epoch": 0.0213986322523715,
+      "grad_norm": 0.19584834575653076,
+      "learning_rate": 0.00014215686274509804,
+      "loss": 1.6608,
+      "step": 291
+    },
+    {
+      "epoch": 0.02147216707110817,
+      "grad_norm": 0.37559205293655396,
+      "learning_rate": 0.0001426470588235294,
+      "loss": 1.5469,
+      "step": 292
+    },
+    {
+      "epoch": 0.02154570188984484,
+      "grad_norm": 0.31330201029777527,
+      "learning_rate": 0.00014313725490196078,
+      "loss": 1.6038,
+      "step": 293
+    },
+    {
+      "epoch": 0.021619236708581513,
+      "grad_norm": 0.3181653618812561,
+      "learning_rate": 0.00014362745098039218,
+      "loss": 1.5838,
+      "step": 294
+    },
+    {
+      "epoch": 0.021692771527318185,
+      "grad_norm": 0.5094121694564819,
+      "learning_rate": 0.00014411764705882354,
+      "loss": 1.5903,
+      "step": 295
+    },
+    {
+      "epoch": 0.021766306346054858,
+      "grad_norm": 0.2156895399093628,
+      "learning_rate": 0.0001446078431372549,
+      "loss": 1.5424,
+      "step": 296
+    },
+    {
+      "epoch": 0.02183984116479153,
+      "grad_norm": 0.4831033945083618,
+      "learning_rate": 0.00014509803921568628,
+      "loss": 1.5654,
+      "step": 297
+    },
+    {
+      "epoch": 0.021913375983528202,
+      "grad_norm": 0.45851844549179077,
+      "learning_rate": 0.00014558823529411765,
+      "loss": 1.5726,
+      "step": 298
+    },
+    {
+      "epoch": 0.02198691080226487,
+      "grad_norm": 0.23476281762123108,
+      "learning_rate": 0.00014607843137254902,
+      "loss": 1.5436,
+      "step": 299
+    },
+    {
+      "epoch": 0.022060445621001543,
+      "grad_norm": 0.37024497985839844,
+      "learning_rate": 0.00014656862745098038,
+      "loss": 1.5522,
+      "step": 300
+    },
+    {
+      "epoch": 0.022133980439738216,
+      "grad_norm": 0.17856507003307343,
+      "learning_rate": 0.00014705882352941178,
+      "loss": 1.5615,
+      "step": 301
+    },
+    {
+      "epoch": 0.022207515258474888,
+      "grad_norm": 0.3484886884689331,
+      "learning_rate": 0.00014754901960784315,
+      "loss": 1.55,
+      "step": 302
+    },
+    {
+      "epoch": 0.02228105007721156,
+      "grad_norm": 0.2221526801586151,
+      "learning_rate": 0.00014803921568627451,
+      "loss": 1.5898,
+      "step": 303
+    },
+    {
+      "epoch": 0.022354584895948233,
+      "grad_norm": 0.37504568696022034,
+      "learning_rate": 0.00014852941176470588,
+      "loss": 1.564,
+      "step": 304
+    },
+    {
+      "epoch": 0.022428119714684905,
+      "grad_norm": 0.40676963329315186,
+      "learning_rate": 0.00014901960784313728,
+      "loss": 1.5973,
+      "step": 305
+    },
+    {
+      "epoch": 0.022501654533421574,
+      "grad_norm": 0.23638203740119934,
+      "learning_rate": 0.00014950980392156865,
+      "loss": 1.5726,
+      "step": 306
+    },
+    {
+      "epoch": 0.022575189352158246,
+      "grad_norm": 0.3754679560661316,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 1.594,
+      "step": 307
+    },
+    {
+      "epoch": 0.02264872417089492,
+      "grad_norm": 0.20535136759281158,
+      "learning_rate": 0.00015049019607843138,
+      "loss": 1.5739,
+      "step": 308
+    },
+    {
+      "epoch": 0.02272225898963159,
+      "grad_norm": 0.3234655559062958,
+      "learning_rate": 0.00015098039215686275,
+      "loss": 1.5717,
+      "step": 309
+    },
+    {
+      "epoch": 0.022795793808368263,
+      "grad_norm": 0.16701354086399078,
+      "learning_rate": 0.00015147058823529412,
+      "loss": 1.5752,
+      "step": 310
+    },
+    {
+      "epoch": 0.022869328627104935,
+      "grad_norm": 0.34804749488830566,
+      "learning_rate": 0.00015196078431372549,
+      "loss": 1.5459,
+      "step": 311
+    },
+    {
+      "epoch": 0.022942863445841608,
+      "grad_norm": 0.1486961990594864,
+      "learning_rate": 0.00015245098039215685,
+      "loss": 1.5686,
+      "step": 312
+    },
+    {
+      "epoch": 0.023016398264578276,
+      "grad_norm": 0.3194875121116638,
+      "learning_rate": 0.00015294117647058822,
+      "loss": 1.5884,
+      "step": 313
+    },
+    {
+      "epoch": 0.02308993308331495,
+      "grad_norm": 0.22025567293167114,
+      "learning_rate": 0.00015343137254901962,
+      "loss": 1.5444,
+      "step": 314
+    },
+    {
+      "epoch": 0.02316346790205162,
+      "grad_norm": 0.2141379714012146,
+      "learning_rate": 0.00015392156862745098,
+      "loss": 1.5575,
+      "step": 315
+    },
+    {
+      "epoch": 0.023237002720788293,
+      "grad_norm": 0.3076874315738678,
+      "learning_rate": 0.00015441176470588238,
+      "loss": 1.5905,
+      "step": 316
+    },
+    {
+      "epoch": 0.023310537539524966,
+      "grad_norm": 0.12764014303684235,
+      "learning_rate": 0.00015490196078431375,
+      "loss": 1.5249,
+      "step": 317
+    },
+    {
+      "epoch": 0.023384072358261638,
+      "grad_norm": 0.2084774374961853,
+      "learning_rate": 0.00015539215686274512,
+      "loss": 1.5579,
+      "step": 318
+    },
+    {
+      "epoch": 0.02345760717699831,
+      "grad_norm": 0.1560492366552353,
+      "learning_rate": 0.00015588235294117648,
+      "loss": 1.537,
+      "step": 319
+    },
+    {
+      "epoch": 0.02353114199573498,
+      "grad_norm": 0.1532682478427887,
+      "learning_rate": 0.00015637254901960785,
+      "loss": 1.5265,
+      "step": 320
+    },
+    {
+      "epoch": 0.02360467681447165,
+      "grad_norm": 0.17342878878116608,
+      "learning_rate": 0.00015686274509803922,
+      "loss": 1.5575,
+      "step": 321
+    },
+    {
+      "epoch": 0.023678211633208324,
+      "grad_norm": 0.16476188600063324,
+      "learning_rate": 0.0001573529411764706,
+      "loss": 1.5366,
+      "step": 322
+    },
+    {
+      "epoch": 0.023751746451944996,
+      "grad_norm": 0.13635559380054474,
+      "learning_rate": 0.00015784313725490196,
+      "loss": 1.5353,
+      "step": 323
+    },
+    {
+      "epoch": 0.02382528127068167,
+      "grad_norm": 0.13971006870269775,
+      "learning_rate": 0.00015833333333333332,
+      "loss": 1.5243,
+      "step": 324
+    },
+    {
+      "epoch": 0.02389881608941834,
+      "grad_norm": 0.16141103208065033,
+      "learning_rate": 0.0001588235294117647,
+      "loss": 1.5178,
+      "step": 325
+    },
+    {
+      "epoch": 0.023972350908155013,
+      "grad_norm": 0.15082556009292603,
+      "learning_rate": 0.0001593137254901961,
+      "loss": 1.5411,
+      "step": 326
+    },
+    {
+      "epoch": 0.024045885726891682,
+      "grad_norm": 0.152690589427948,
+      "learning_rate": 0.00015980392156862746,
+      "loss": 1.489,
+      "step": 327
+    },
+    {
+      "epoch": 0.024119420545628354,
+      "grad_norm": 0.14512793719768524,
+      "learning_rate": 0.00016029411764705885,
+      "loss": 1.5177,
+      "step": 328
+    },
+    {
+      "epoch": 0.024192955364365026,
+      "grad_norm": 0.13109825551509857,
+      "learning_rate": 0.00016078431372549022,
+      "loss": 1.5237,
+      "step": 329
+    },
+    {
+      "epoch": 0.0242664901831017,
+      "grad_norm": 0.18076738715171814,
+      "learning_rate": 0.00016127450980392159,
+      "loss": 1.4895,
+      "step": 330
+    },
+    {
+      "epoch": 0.02434002500183837,
+      "grad_norm": 0.17187830805778503,
+      "learning_rate": 0.00016176470588235295,
+      "loss": 1.5078,
+      "step": 331
+    },
+    {
+      "epoch": 0.024413559820575043,
+      "grad_norm": 0.14935794472694397,
+      "learning_rate": 0.00016225490196078432,
+      "loss": 1.5805,
+      "step": 332
+    },
+    {
+      "epoch": 0.024487094639311716,
+      "grad_norm": 0.18950457870960236,
+      "learning_rate": 0.0001627450980392157,
+      "loss": 1.4867,
+      "step": 333
+    },
+    {
+      "epoch": 0.024560629458048384,
+      "grad_norm": 0.17502963542938232,
+      "learning_rate": 0.00016323529411764706,
+      "loss": 1.4814,
+      "step": 334
+    },
+    {
+      "epoch": 0.024634164276785057,
+      "grad_norm": 0.1747254729270935,
+      "learning_rate": 0.00016372549019607843,
+      "loss": 1.5031,
+      "step": 335
+    },
+    {
+      "epoch": 0.02470769909552173,
+      "grad_norm": 0.230714350938797,
+      "learning_rate": 0.0001642156862745098,
+      "loss": 1.5377,
+      "step": 336
+    },
+    {
+      "epoch": 0.0247812339142584,
+      "grad_norm": 0.2713477611541748,
+      "learning_rate": 0.0001647058823529412,
+      "loss": 1.5268,
+      "step": 337
+    },
+    {
+      "epoch": 0.024854768732995074,
+      "grad_norm": 0.16902808845043182,
+      "learning_rate": 0.00016519607843137256,
+      "loss": 1.4728,
+      "step": 338
+    },
+    {
+      "epoch": 0.024928303551731746,
+      "grad_norm": 0.22565551102161407,
+      "learning_rate": 0.00016568627450980395,
+      "loss": 1.488,
+      "step": 339
+    },
+    {
+      "epoch": 0.02500183837046842,
+      "grad_norm": 0.1785699427127838,
+      "learning_rate": 0.00016617647058823532,
+      "loss": 1.5297,
+      "step": 340
+    },
+    {
+      "epoch": 0.025075373189205087,
+      "grad_norm": 0.21787653863430023,
+      "learning_rate": 0.0001666666666666667,
+      "loss": 1.5291,
+      "step": 341
+    },
+    {
+      "epoch": 0.02514890800794176,
+      "grad_norm": 0.12997353076934814,
+      "learning_rate": 0.00016715686274509806,
+      "loss": 1.539,
+      "step": 342
+    },
+    {
+      "epoch": 0.025222442826678432,
+      "grad_norm": 0.1611165702342987,
+      "learning_rate": 0.00016764705882352942,
+      "loss": 1.45,
+      "step": 343
+    },
+    {
+      "epoch": 0.025295977645415104,
+      "grad_norm": 0.18920809030532837,
+      "learning_rate": 0.0001681372549019608,
+      "loss": 1.5225,
+      "step": 344
+    },
+    {
+      "epoch": 0.025369512464151776,
+      "grad_norm": 0.1225525364279747,
+      "learning_rate": 0.00016862745098039216,
+      "loss": 1.4613,
+      "step": 345
+    },
+    {
+      "epoch": 0.02544304728288845,
+      "grad_norm": 0.18548697233200073,
+      "learning_rate": 0.00016911764705882353,
+      "loss": 1.428,
+      "step": 346
+    },
+    {
+      "epoch": 0.02551658210162512,
+      "grad_norm": 0.21388238668441772,
+      "learning_rate": 0.0001696078431372549,
+      "loss": 1.4236,
+      "step": 347
+    },
+    {
+      "epoch": 0.02559011692036179,
+      "grad_norm": 0.2182207703590393,
+      "learning_rate": 0.0001700980392156863,
+      "loss": 1.4385,
+      "step": 348
+    },
+    {
+      "epoch": 0.025663651739098462,
+      "grad_norm": 0.17997369170188904,
+      "learning_rate": 0.00017058823529411766,
+      "loss": 1.4298,
+      "step": 349
+    },
+    {
+      "epoch": 0.025737186557835134,
+      "grad_norm": 0.22576956450939178,
+      "learning_rate": 0.00017107843137254903,
+      "loss": 1.4373,
+      "step": 350
+    },
+    {
+      "epoch": 0.025810721376571807,
+      "grad_norm": 0.15359701216220856,
+      "learning_rate": 0.0001715686274509804,
+      "loss": 1.4173,
+      "step": 351
+    },
+    {
+      "epoch": 0.02588425619530848,
+      "grad_norm": 0.20151320099830627,
+      "learning_rate": 0.0001720588235294118,
+      "loss": 1.432,
+      "step": 352
+    },
+    {
+      "epoch": 0.02595779101404515,
+      "grad_norm": 0.1420849859714508,
+      "learning_rate": 0.00017254901960784316,
+      "loss": 1.4377,
+      "step": 353
+    },
+    {
+      "epoch": 0.026031325832781824,
+      "grad_norm": 0.1588618904352188,
+      "learning_rate": 0.00017303921568627453,
+      "loss": 1.441,
+      "step": 354
+    },
+    {
+      "epoch": 0.026104860651518492,
+      "grad_norm": 0.13913603127002716,
+      "learning_rate": 0.0001735294117647059,
+      "loss": 1.4348,
+      "step": 355
+    },
+    {
+      "epoch": 0.026178395470255165,
+      "grad_norm": 0.22640866041183472,
+      "learning_rate": 0.00017401960784313726,
+      "loss": 1.4287,
+      "step": 356
+    },
+    {
+      "epoch": 0.026251930288991837,
+      "grad_norm": 0.1877104490995407,
+      "learning_rate": 0.00017450980392156863,
+      "loss": 1.4357,
+      "step": 357
+    },
+    {
+      "epoch": 0.02632546510772851,
+      "grad_norm": 0.1879999190568924,
+      "learning_rate": 0.000175,
+      "loss": 1.4689,
+      "step": 358
+    },
+    {
+      "epoch": 0.02639899992646518,
+      "grad_norm": 0.23328062891960144,
+      "learning_rate": 0.00017549019607843137,
+      "loss": 1.4431,
+      "step": 359
+    },
+    {
+      "epoch": 0.026472534745201854,
+      "grad_norm": 0.1980171948671341,
+      "learning_rate": 0.00017598039215686276,
+      "loss": 1.4262,
+      "step": 360
+    },
+    {
+      "epoch": 0.026546069563938526,
+      "grad_norm": 0.26641011238098145,
+      "learning_rate": 0.00017647058823529413,
+      "loss": 1.427,
+      "step": 361
+    },
+    {
+      "epoch": 0.026619604382675195,
+      "grad_norm": 0.257282555103302,
+      "learning_rate": 0.0001769607843137255,
+      "loss": 1.4269,
+      "step": 362
+    },
+    {
+      "epoch": 0.026693139201411867,
+      "grad_norm": 0.28722676634788513,
+      "learning_rate": 0.00017745098039215687,
+      "loss": 1.4256,
+      "step": 363
+    },
+    {
+      "epoch": 0.02676667402014854,
+      "grad_norm": 0.22197920083999634,
+      "learning_rate": 0.00017794117647058823,
+      "loss": 1.4366,
+      "step": 364
+    },
+    {
+      "epoch": 0.026840208838885212,
+      "grad_norm": 0.3489218056201935,
+      "learning_rate": 0.00017843137254901963,
+      "loss": 1.4008,
+      "step": 365
+    },
+    {
+      "epoch": 0.026913743657621884,
+      "grad_norm": 0.20068013668060303,
+      "learning_rate": 0.000178921568627451,
+      "loss": 1.4637,
+      "step": 366
+    },
+    {
+      "epoch": 0.026987278476358557,
+      "grad_norm": 0.35178759694099426,
+      "learning_rate": 0.00017941176470588236,
+      "loss": 1.4216,
+      "step": 367
+    },
+    {
+      "epoch": 0.02706081329509523,
+      "grad_norm": 0.21447311341762543,
+      "learning_rate": 0.00017990196078431373,
+      "loss": 1.4417,
+      "step": 368
+    },
+    {
+      "epoch": 0.027134348113831898,
+      "grad_norm": 0.3598118722438812,
+      "learning_rate": 0.0001803921568627451,
+      "loss": 1.4549,
+      "step": 369
+    },
+    {
+      "epoch": 0.02720788293256857,
+      "grad_norm": 0.266422837972641,
+      "learning_rate": 0.00018088235294117647,
+      "loss": 1.4373,
+      "step": 370
+    },
+    {
+      "epoch": 0.027281417751305242,
+      "grad_norm": 0.35622239112854004,
+      "learning_rate": 0.00018137254901960786,
+      "loss": 1.3928,
+      "step": 371
+    },
+    {
+      "epoch": 0.027354952570041915,
+      "grad_norm": 0.3308359682559967,
+      "learning_rate": 0.00018186274509803923,
+      "loss": 1.4208,
+      "step": 372
+    },
+    {
+      "epoch": 0.027428487388778587,
+      "grad_norm": 0.31302496790885925,
+      "learning_rate": 0.0001823529411764706,
+      "loss": 1.4059,
+      "step": 373
+    },
+    {
+      "epoch": 0.02750202220751526,
+      "grad_norm": 0.19252116978168488,
+      "learning_rate": 0.00018284313725490197,
+      "loss": 1.4172,
+      "step": 374
+    },
+    {
+      "epoch": 0.02757555702625193,
+      "grad_norm": 0.4443724751472473,
+      "learning_rate": 0.00018333333333333334,
+      "loss": 1.4589,
+      "step": 375
+    },
+    {
+      "epoch": 0.0276490918449886,
+      "grad_norm": 0.2017659842967987,
+      "learning_rate": 0.0001838235294117647,
+      "loss": 1.4371,
+      "step": 376
+    },
+    {
+      "epoch": 0.027722626663725273,
+      "grad_norm": 0.37223950028419495,
+      "learning_rate": 0.00018431372549019607,
+      "loss": 1.3979,
+      "step": 377
+    },
+    {
+      "epoch": 0.027796161482461945,
+      "grad_norm": 0.13605129718780518,
+      "learning_rate": 0.00018480392156862747,
+      "loss": 1.4087,
+      "step": 378
+    },
+    {
+      "epoch": 0.027869696301198617,
+      "grad_norm": 0.3101280927658081,
+      "learning_rate": 0.00018529411764705883,
+      "loss": 1.4371,
+      "step": 379
+    },
+    {
+      "epoch": 0.02794323111993529,
+      "grad_norm": 0.13537485897541046,
+      "learning_rate": 0.0001857843137254902,
+      "loss": 1.4199,
+      "step": 380
+    },
+    {
+      "epoch": 0.028016765938671962,
+      "grad_norm": 0.2042521983385086,
+      "learning_rate": 0.00018627450980392157,
+      "loss": 1.4261,
+      "step": 381
+    },
+    {
+      "epoch": 0.028090300757408634,
+      "grad_norm": 0.2524612545967102,
+      "learning_rate": 0.00018676470588235297,
+      "loss": 1.4561,
+      "step": 382
+    },
+    {
+      "epoch": 0.028163835576145303,
+      "grad_norm": 0.2053963541984558,
+      "learning_rate": 0.00018725490196078433,
+      "loss": 1.4123,
+      "step": 383
+    },
+    {
+      "epoch": 0.028237370394881976,
+      "grad_norm": 0.24226069450378418,
+      "learning_rate": 0.0001877450980392157,
+      "loss": 1.4326,
+      "step": 384
+    },
+    {
+      "epoch": 0.028310905213618648,
+      "grad_norm": 0.1278466135263443,
+      "learning_rate": 0.00018823529411764707,
+      "loss": 1.4275,
+      "step": 385
+    },
+    {
+      "epoch": 0.02838444003235532,
+      "grad_norm": 0.30572548508644104,
+      "learning_rate": 0.00018872549019607844,
+      "loss": 1.3827,
+      "step": 386
+    },
+    {
+      "epoch": 0.028457974851091992,
+      "grad_norm": 0.12484510987997055,
+      "learning_rate": 0.0001892156862745098,
+      "loss": 1.4277,
+      "step": 387
+    },
+    {
+      "epoch": 0.028531509669828665,
+      "grad_norm": 0.3359142243862152,
+      "learning_rate": 0.00018970588235294117,
+      "loss": 1.3851,
+      "step": 388
+    },
+    {
+      "epoch": 0.028605044488565337,
+      "grad_norm": 0.11149458587169647,
+      "learning_rate": 0.00019019607843137254,
+      "loss": 1.4104,
+      "step": 389
+    },
+    {
+      "epoch": 0.028678579307302006,
+      "grad_norm": 0.26489025354385376,
+      "learning_rate": 0.00019068627450980394,
+      "loss": 1.4432,
+      "step": 390
+    },
+    {
+      "epoch": 0.028752114126038678,
+      "grad_norm": 0.13289614021778107,
+      "learning_rate": 0.0001911764705882353,
+      "loss": 1.4088,
+      "step": 391
+    },
+    {
+      "epoch": 0.02882564894477535,
+      "grad_norm": 0.2556290924549103,
+      "learning_rate": 0.00019166666666666667,
+      "loss": 1.4209,
+      "step": 392
+    },
+    {
+      "epoch": 0.028899183763512023,
+      "grad_norm": 0.1379719227552414,
+      "learning_rate": 0.00019215686274509807,
+      "loss": 1.3965,
+      "step": 393
+    },
+    {
+      "epoch": 0.028972718582248695,
+      "grad_norm": 0.17995013296604156,
+      "learning_rate": 0.00019264705882352944,
+      "loss": 1.4223,
+      "step": 394
+    },
+    {
+      "epoch": 0.029046253400985367,
+      "grad_norm": 0.20030899345874786,
+      "learning_rate": 0.0001931372549019608,
+      "loss": 1.4069,
+      "step": 395
+    },
+    {
+      "epoch": 0.02911978821972204,
+      "grad_norm": 0.2008051723241806,
+      "learning_rate": 0.00019362745098039217,
+      "loss": 1.4349,
+      "step": 396
+    },
+    {
+      "epoch": 0.02919332303845871,
+      "grad_norm": 0.17532874643802643,
+      "learning_rate": 0.00019411764705882354,
+      "loss": 1.4215,
+      "step": 397
+    },
+    {
+      "epoch": 0.02926685785719538,
+      "grad_norm": 0.16829505562782288,
+      "learning_rate": 0.0001946078431372549,
+      "loss": 1.3564,
+      "step": 398
+    },
+    {
+      "epoch": 0.029340392675932053,
+      "grad_norm": 0.16335788369178772,
+      "learning_rate": 0.00019509803921568628,
+      "loss": 1.3969,
+      "step": 399
+    },
+    {
+      "epoch": 0.029413927494668726,
+      "grad_norm": 0.2881438136100769,
+      "learning_rate": 0.00019558823529411764,
+      "loss": 1.407,
+      "step": 400
+    },
+    {
+      "epoch": 0.029487462313405398,
+      "grad_norm": 0.12539958953857422,
+      "learning_rate": 0.000196078431372549,
+      "loss": 1.4409,
+      "step": 401
+    },
+    {
+      "epoch": 0.02956099713214207,
+      "grad_norm": 0.2703852951526642,
+      "learning_rate": 0.00019656862745098038,
+      "loss": 1.3932,
+      "step": 402
+    },
+    {
+      "epoch": 0.029634531950878742,
+      "grad_norm": 0.1419435292482376,
+      "learning_rate": 0.00019705882352941177,
+      "loss": 1.3809,
+      "step": 403
+    },
+    {
+      "epoch": 0.02970806676961541,
+      "grad_norm": 0.2732055187225342,
+      "learning_rate": 0.00019754901960784314,
+      "loss": 1.3993,
+      "step": 404
+    },
+    {
+      "epoch": 0.029781601588352084,
+      "grad_norm": 0.15631918609142303,
+      "learning_rate": 0.00019803921568627454,
+      "loss": 1.4352,
+      "step": 405
+    },
+    {
+      "epoch": 0.029855136407088756,
+      "grad_norm": 0.17677758634090424,
+      "learning_rate": 0.0001985294117647059,
+      "loss": 1.4115,
+      "step": 406
+    },
+    {
+      "epoch": 0.029928671225825428,
+      "grad_norm": 0.1566472053527832,
+      "learning_rate": 0.00019901960784313727,
+      "loss": 1.4188,
+      "step": 407
+    },
+    {
+      "epoch": 0.0300022060445621,
+      "grad_norm": 0.11110509932041168,
+      "learning_rate": 0.00019950980392156864,
+      "loss": 1.3464,
+      "step": 408
+    },
+    {
+      "epoch": 0.030075740863298773,
+      "grad_norm": 0.17791087925434113,
+      "learning_rate": 0.0002,
+      "loss": 1.4138,
+      "step": 409
+    },
+    {
+      "epoch": 0.030149275682035445,
+      "grad_norm": 0.1319073736667633,
+      "learning_rate": 0.00019999999716394725,
+      "loss": 1.3568,
+      "step": 410
+    },
+    {
+      "epoch": 0.030222810500772114,
+      "grad_norm": 0.18880265951156616,
+      "learning_rate": 0.0001999999886557891,
+      "loss": 1.3662,
+      "step": 411
+    },
+    {
+      "epoch": 0.030296345319508786,
+      "grad_norm": 0.1976701021194458,
+      "learning_rate": 0.00019999997447552606,
+      "loss": 1.3586,
+      "step": 412
+    },
+    {
+      "epoch": 0.03036988013824546,
+      "grad_norm": 0.19535310566425323,
+      "learning_rate": 0.00019999995462315897,
+      "loss": 1.3736,
+      "step": 413
+    },
+    {
+      "epoch": 0.03044341495698213,
+      "grad_norm": 0.17990554869174957,
+      "learning_rate": 0.00019999992909868887,
+      "loss": 1.392,
+      "step": 414
+    },
+    {
+      "epoch": 0.030516949775718803,
+      "grad_norm": 0.26179641485214233,
+      "learning_rate": 0.0001999998979021173,
+      "loss": 1.413,
+      "step": 415
+    },
+    {
+      "epoch": 0.030590484594455476,
+      "grad_norm": 0.14347496628761292,
+      "learning_rate": 0.00019999986103344598,
+      "loss": 1.3506,
+      "step": 416
+    },
+    {
+      "epoch": 0.030664019413192148,
+      "grad_norm": 0.20818090438842773,
+      "learning_rate": 0.00019999981849267698,
+      "loss": 1.3386,
+      "step": 417
+    },
+    {
+      "epoch": 0.030737554231928817,
+      "grad_norm": 0.17982816696166992,
+      "learning_rate": 0.00019999977027981275,
+      "loss": 1.352,
+      "step": 418
+    },
+    {
+      "epoch": 0.03081108905066549,
+      "grad_norm": 0.19797399640083313,
+      "learning_rate": 0.00019999971639485603,
+      "loss": 1.434,
+      "step": 419
+    },
+    {
+      "epoch": 0.03088462386940216,
+      "grad_norm": 0.15006445348262787,
+      "learning_rate": 0.00019999965683780983,
+      "loss": 1.3689,
+      "step": 420
+    },
+    {
+      "epoch": 0.030958158688138834,
+      "grad_norm": 0.15265107154846191,
+      "learning_rate": 0.0001999995916086776,
+      "loss": 1.372,
+      "step": 421
+    },
+    {
+      "epoch": 0.031031693506875506,
+      "grad_norm": 0.1910703331232071,
+      "learning_rate": 0.00019999952070746298,
+      "loss": 1.3874,
+      "step": 422
+    },
+    {
+      "epoch": 0.031105228325612178,
+      "grad_norm": 0.20779851078987122,
+      "learning_rate": 0.00019999944413417003,
+      "loss": 1.3788,
+      "step": 423
+    },
+    {
+      "epoch": 0.03117876314434885,
+      "grad_norm": 0.20299364626407623,
+      "learning_rate": 0.00019999936188880304,
+      "loss": 1.3877,
+      "step": 424
+    },
+    {
+      "epoch": 0.03125229796308552,
+      "grad_norm": 0.1946035921573639,
+      "learning_rate": 0.00019999927397136672,
+      "loss": 1.4031,
+      "step": 425
+    },
+    {
+      "epoch": 0.031325832781822195,
+      "grad_norm": 0.17915979027748108,
+      "learning_rate": 0.00019999918038186607,
+      "loss": 1.3881,
+      "step": 426
+    },
+    {
+      "epoch": 0.031399367600558864,
+      "grad_norm": 0.18923264741897583,
+      "learning_rate": 0.00019999908112030638,
+      "loss": 1.3469,
+      "step": 427
+    },
+    {
+      "epoch": 0.03147290241929554,
+      "grad_norm": 0.13932152092456818,
+      "learning_rate": 0.00019999897618669325,
+      "loss": 1.3408,
+      "step": 428
+    },
+    {
+      "epoch": 0.03154643723803221,
+      "grad_norm": 0.1577254682779312,
+      "learning_rate": 0.00019999886558103266,
+      "loss": 1.3372,
+      "step": 429
+    },
+    {
+      "epoch": 0.03161997205676888,
+      "grad_norm": 0.15099476277828217,
+      "learning_rate": 0.0001999987493033309,
+      "loss": 1.3688,
+      "step": 430
+    },
+    {
+      "epoch": 0.03169350687550555,
+      "grad_norm": 0.13525256514549255,
+      "learning_rate": 0.00019999862735359453,
+      "loss": 1.3588,
+      "step": 431
+    },
+    {
+      "epoch": 0.03176704169424222,
+      "grad_norm": 0.12422113120555878,
+      "learning_rate": 0.0001999984997318305,
+      "loss": 1.3835,
+      "step": 432
+    },
+    {
+      "epoch": 0.0318405765129789,
+      "grad_norm": 0.1376965045928955,
+      "learning_rate": 0.00019999836643804598,
+      "loss": 1.3696,
+      "step": 433
+    },
+    {
+      "epoch": 0.03191411133171557,
+      "grad_norm": 0.12303666770458221,
+      "learning_rate": 0.00019999822747224864,
+      "loss": 1.3302,
+      "step": 434
+    },
+    {
+      "epoch": 0.03198764615045224,
+      "grad_norm": 0.12676239013671875,
+      "learning_rate": 0.0001999980828344463,
+      "loss": 1.3464,
+      "step": 435
+    },
+    {
+      "epoch": 0.03206118096918891,
+      "grad_norm": 0.12209544330835342,
+      "learning_rate": 0.00019999793252464714,
+      "loss": 1.3528,
+      "step": 436
+    },
+    {
+      "epoch": 0.03213471578792558,
+      "grad_norm": 0.14016193151474,
+      "learning_rate": 0.00019999777654285974,
+      "loss": 1.3398,
+      "step": 437
+    },
+    {
+      "epoch": 0.032208250606662256,
+      "grad_norm": 0.1419256627559662,
+      "learning_rate": 0.00019999761488909292,
+      "loss": 1.3632,
+      "step": 438
+    },
+    {
+      "epoch": 0.032281785425398925,
+      "grad_norm": 0.12737326323986053,
+      "learning_rate": 0.00019999744756335586,
+      "loss": 1.3605,
+      "step": 439
+    },
+    {
+      "epoch": 0.0323553202441356,
+      "grad_norm": 0.19141319394111633,
+      "learning_rate": 0.000199997274565658,
+      "loss": 1.3628,
+      "step": 440
+    },
+    {
+      "epoch": 0.03242885506287227,
+      "grad_norm": 0.12736383080482483,
+      "learning_rate": 0.00019999709589600925,
+      "loss": 1.3524,
+      "step": 441
+    },
+    {
+      "epoch": 0.032502389881608945,
+      "grad_norm": 0.12150347977876663,
+      "learning_rate": 0.00019999691155441968,
+      "loss": 1.3606,
+      "step": 442
+    },
+    {
+      "epoch": 0.032575924700345614,
+      "grad_norm": 0.17770102620124817,
+      "learning_rate": 0.00019999672154089973,
+      "loss": 1.3881,
+      "step": 443
+    },
+    {
+      "epoch": 0.03264945951908228,
+      "grad_norm": 0.10384153574705124,
+      "learning_rate": 0.00019999652585546022,
+      "loss": 1.3973,
+      "step": 444
+    },
+    {
+      "epoch": 0.03272299433781896,
+      "grad_norm": 0.14906629920005798,
+      "learning_rate": 0.00019999632449811222,
+      "loss": 1.3373,
+      "step": 445
+    },
+    {
+      "epoch": 0.03279652915655563,
+      "grad_norm": 0.123542420566082,
+      "learning_rate": 0.00019999611746886718,
+      "loss": 1.3362,
+      "step": 446
+    },
+    {
+      "epoch": 0.0328700639752923,
+      "grad_norm": 0.12174753844738007,
+      "learning_rate": 0.00019999590476773682,
+      "loss": 1.3198,
+      "step": 447
+    },
+    {
+      "epoch": 0.03294359879402897,
+      "grad_norm": 0.11821915209293365,
+      "learning_rate": 0.00019999568639473322,
+      "loss": 1.3167,
+      "step": 448
+    },
+    {
+      "epoch": 0.03301713361276565,
+      "grad_norm": 0.12207061052322388,
+      "learning_rate": 0.00019999546234986872,
+      "loss": 1.2931,
+      "step": 449
+    },
+    {
+      "epoch": 0.03309066843150232,
+      "grad_norm": 0.11416268348693848,
+      "learning_rate": 0.0001999952326331561,
+      "loss": 1.2682,
+      "step": 450
+    },
+    {
+      "epoch": 0.033164203250238986,
+      "grad_norm": 0.11748199909925461,
+      "learning_rate": 0.00019999499724460837,
+      "loss": 1.3031,
+      "step": 451
+    },
+    {
+      "epoch": 0.03323773806897566,
+      "grad_norm": 0.12995539605617523,
+      "learning_rate": 0.00019999475618423884,
+      "loss": 1.2729,
+      "step": 452
+    },
+    {
+      "epoch": 0.03331127288771233,
+      "grad_norm": 0.12546691298484802,
+      "learning_rate": 0.0001999945094520612,
+      "loss": 1.3418,
+      "step": 453
+    },
+    {
+      "epoch": 0.033384807706449006,
+      "grad_norm": 0.1396404504776001,
+      "learning_rate": 0.0001999942570480895,
+      "loss": 1.2904,
+      "step": 454
+    },
+    {
+      "epoch": 0.033458342525185675,
+      "grad_norm": 0.13924859464168549,
+      "learning_rate": 0.00019999399897233796,
+      "loss": 1.272,
+      "step": 455
+    },
+    {
+      "epoch": 0.03353187734392235,
+      "grad_norm": 0.14166858792304993,
+      "learning_rate": 0.00019999373522482133,
+      "loss": 1.2949,
+      "step": 456
+    },
+    {
+      "epoch": 0.03360541216265902,
+      "grad_norm": 0.12089133262634277,
+      "learning_rate": 0.00019999346580555448,
+      "loss": 1.2844,
+      "step": 457
+    },
+    {
+      "epoch": 0.03367894698139569,
+      "grad_norm": 0.11832351982593536,
+      "learning_rate": 0.0001999931907145527,
+      "loss": 1.3145,
+      "step": 458
+    },
+    {
+      "epoch": 0.033752481800132364,
+      "grad_norm": 0.11356719583272934,
+      "learning_rate": 0.00019999290995183165,
+      "loss": 1.3137,
+      "step": 459
+    },
+    {
+      "epoch": 0.03382601661886903,
+      "grad_norm": 0.1521197110414505,
+      "learning_rate": 0.00019999262351740719,
+      "loss": 1.3118,
+      "step": 460
+    },
+    {
+      "epoch": 0.03389955143760571,
+      "grad_norm": 0.1212448924779892,
+      "learning_rate": 0.00019999233141129562,
+      "loss": 1.3018,
+      "step": 461
+    },
+    {
+      "epoch": 0.03397308625634238,
+      "grad_norm": 0.16737763583660126,
+      "learning_rate": 0.00019999203363351347,
+      "loss": 1.3104,
+      "step": 462
+    },
+    {
+      "epoch": 0.03404662107507905,
+      "grad_norm": 0.15121692419052124,
+      "learning_rate": 0.00019999173018407767,
+      "loss": 1.2898,
+      "step": 463
+    },
+    {
+      "epoch": 0.03412015589381572,
+      "grad_norm": 0.12433341145515442,
+      "learning_rate": 0.0001999914210630054,
+      "loss": 1.2994,
+      "step": 464
+    },
+    {
+      "epoch": 0.03419369071255239,
+      "grad_norm": 0.13267605006694794,
+      "learning_rate": 0.0001999911062703142,
+      "loss": 1.2841,
+      "step": 465
+    },
+    {
+      "epoch": 0.03426722553128907,
+      "grad_norm": 0.14349037408828735,
+      "learning_rate": 0.00019999078580602194,
+      "loss": 1.2968,
+      "step": 466
+    },
+    {
+      "epoch": 0.034340760350025736,
+      "grad_norm": 0.14032845199108124,
+      "learning_rate": 0.00019999045967014674,
+      "loss": 1.2889,
+      "step": 467
+    },
+    {
+      "epoch": 0.03441429516876241,
+      "grad_norm": 0.12519770860671997,
+      "learning_rate": 0.00019999012786270717,
+      "loss": 1.3273,
+      "step": 468
+    },
+    {
+      "epoch": 0.03448782998749908,
+      "grad_norm": 0.13742230832576752,
+      "learning_rate": 0.00019998979038372204,
+      "loss": 1.2695,
+      "step": 469
+    },
+    {
+      "epoch": 0.034561364806235756,
+      "grad_norm": 0.1406245231628418,
+      "learning_rate": 0.00019998944723321045,
+      "loss": 1.2946,
+      "step": 470
+    },
+    {
+      "epoch": 0.034634899624972425,
+      "grad_norm": 0.11969190835952759,
+      "learning_rate": 0.0001999890984111919,
+      "loss": 1.2979,
+      "step": 471
+    },
+    {
+      "epoch": 0.034708434443709094,
+      "grad_norm": 0.1397286206483841,
+      "learning_rate": 0.00019998874391768617,
+      "loss": 1.304,
+      "step": 472
+    },
+    {
+      "epoch": 0.03478196926244577,
+      "grad_norm": 0.14878150820732117,
+      "learning_rate": 0.00019998838375271335,
+      "loss": 1.2989,
+      "step": 473
+    },
+    {
+      "epoch": 0.03485550408118244,
+      "grad_norm": 0.14968284964561462,
+      "learning_rate": 0.00019998801791629392,
+      "loss": 1.2752,
+      "step": 474
+    },
+    {
+      "epoch": 0.034929038899919114,
+      "grad_norm": 0.1841110736131668,
+      "learning_rate": 0.00019998764640844857,
+      "loss": 1.3086,
+      "step": 475
+    },
+    {
+      "epoch": 0.03500257371865578,
+      "grad_norm": 0.13120782375335693,
+      "learning_rate": 0.00019998726922919836,
+      "loss": 1.2989,
+      "step": 476
+    },
+    {
+      "epoch": 0.03507610853739246,
+      "grad_norm": 0.17819121479988098,
+      "learning_rate": 0.00019998688637856476,
+      "loss": 1.3269,
+      "step": 477
+    },
+    {
+      "epoch": 0.03514964335612913,
+      "grad_norm": 0.16018305718898773,
+      "learning_rate": 0.00019998649785656945,
+      "loss": 1.2997,
+      "step": 478
+    },
+    {
+      "epoch": 0.035223178174865796,
+      "grad_norm": 0.2175554782152176,
+      "learning_rate": 0.00019998610366323444,
+      "loss": 1.2789,
+      "step": 479
+    },
+    {
+      "epoch": 0.03529671299360247,
+      "grad_norm": 0.14388512074947357,
+      "learning_rate": 0.0001999857037985821,
+      "loss": 1.3082,
+      "step": 480
+    },
+    {
+      "epoch": 0.03537024781233914,
+      "grad_norm": 0.1513884961605072,
+      "learning_rate": 0.00019998529826263515,
+      "loss": 1.3129,
+      "step": 481
+    },
+    {
+      "epoch": 0.03544378263107582,
+      "grad_norm": 0.13684803247451782,
+      "learning_rate": 0.00019998488705541654,
+      "loss": 1.261,
+      "step": 482
+    },
+    {
+      "epoch": 0.035517317449812486,
+      "grad_norm": 0.12327020615339279,
+      "learning_rate": 0.00019998447017694962,
+      "loss": 1.2967,
+      "step": 483
+    },
+    {
+      "epoch": 0.03559085226854916,
+      "grad_norm": 0.13387431204319,
+      "learning_rate": 0.00019998404762725804,
+      "loss": 1.3061,
+      "step": 484
+    },
+    {
+      "epoch": 0.03566438708728583,
+      "grad_norm": 0.11352293938398361,
+      "learning_rate": 0.00019998361940636578,
+      "loss": 1.2998,
+      "step": 485
+    },
+    {
+      "epoch": 0.0357379219060225,
+      "grad_norm": 0.11300481855869293,
+      "learning_rate": 0.00019998318551429708,
+      "loss": 1.308,
+      "step": 486
+    },
+    {
+      "epoch": 0.035811456724759175,
+      "grad_norm": 0.11518463492393494,
+      "learning_rate": 0.0001999827459510766,
+      "loss": 1.2856,
+      "step": 487
+    },
+    {
+      "epoch": 0.035884991543495844,
+      "grad_norm": 0.11767704784870148,
+      "learning_rate": 0.00019998230071672924,
+      "loss": 1.2816,
+      "step": 488
+    },
+    {
+      "epoch": 0.03595852636223252,
+      "grad_norm": 0.11238985508680344,
+      "learning_rate": 0.0001999818498112803,
+      "loss": 1.2986,
+      "step": 489
+    },
+    {
+      "epoch": 0.03603206118096919,
+      "grad_norm": 0.11698058992624283,
+      "learning_rate": 0.0001999813932347553,
+      "loss": 1.35,
+      "step": 490
+    },
+    {
+      "epoch": 0.036105595999705864,
+      "grad_norm": 0.1335151493549347,
+      "learning_rate": 0.00019998093098718016,
+      "loss": 1.3028,
+      "step": 491
+    },
+    {
+      "epoch": 0.03617913081844253,
+      "grad_norm": 0.13023127615451813,
+      "learning_rate": 0.00019998046306858109,
+      "loss": 1.285,
+      "step": 492
+    },
+    {
+      "epoch": 0.0362526656371792,
+      "grad_norm": 0.12120972573757172,
+      "learning_rate": 0.00019997998947898465,
+      "loss": 1.3183,
+      "step": 493
+    },
+    {
+      "epoch": 0.03632620045591588,
+      "grad_norm": 0.11683407425880432,
+      "learning_rate": 0.0001999795102184177,
+      "loss": 1.3164,
+      "step": 494
+    },
+    {
+      "epoch": 0.036399735274652546,
+      "grad_norm": 0.12935255467891693,
+      "learning_rate": 0.00019997902528690742,
+      "loss": 1.2892,
+      "step": 495
+    },
+    {
+      "epoch": 0.03647327009338922,
+      "grad_norm": 0.1460026502609253,
+      "learning_rate": 0.00019997853468448128,
+      "loss": 1.2714,
+      "step": 496
+    },
+    {
+      "epoch": 0.03654680491212589,
+      "grad_norm": 0.12446979433298111,
+      "learning_rate": 0.00019997803841116717,
+      "loss": 1.2669,
+      "step": 497
+    },
+    {
+      "epoch": 0.03662033973086257,
+      "grad_norm": 0.11518928408622742,
+      "learning_rate": 0.00019997753646699322,
+      "loss": 1.2719,
+      "step": 498
+    },
+    {
+      "epoch": 0.036693874549599236,
+      "grad_norm": 0.14306655526161194,
+      "learning_rate": 0.00019997702885198788,
+      "loss": 1.3437,
+      "step": 499
+    },
+    {
+      "epoch": 0.036767409368335904,
+      "grad_norm": 0.1200539767742157,
+      "learning_rate": 0.00019997651556617996,
+      "loss": 1.2875,
+      "step": 500
+    },
+    {
+      "epoch": 0.03684094418707258,
+      "grad_norm": 0.1254974603652954,
+      "learning_rate": 0.00019997599660959855,
+      "loss": 1.317,
+      "step": 501
+    },
+    {
+      "epoch": 0.03691447900580925,
+      "grad_norm": 0.14290323853492737,
+      "learning_rate": 0.0001999754719822731,
+      "loss": 1.3014,
+      "step": 502
+    },
+    {
+      "epoch": 0.036988013824545925,
+      "grad_norm": 0.20829331874847412,
+      "learning_rate": 0.00019997494168423342,
+      "loss": 1.2934,
+      "step": 503
+    },
+    {
+      "epoch": 0.037061548643282594,
+      "grad_norm": 0.13102738559246063,
+      "learning_rate": 0.00019997440571550948,
+      "loss": 1.2898,
+      "step": 504
+    },
+    {
+      "epoch": 0.03713508346201927,
+      "grad_norm": 0.15092189610004425,
+      "learning_rate": 0.00019997386407613176,
+      "loss": 1.3164,
+      "step": 505
+    },
+    {
+      "epoch": 0.03720861828075594,
+      "grad_norm": 0.1561775505542755,
+      "learning_rate": 0.00019997331676613098,
+      "loss": 1.3042,
+      "step": 506
+    },
+    {
+      "epoch": 0.03728215309949261,
+      "grad_norm": 0.2192060500383377,
+      "learning_rate": 0.00019997276378553815,
+      "loss": 1.3157,
+      "step": 507
+    },
+    {
+      "epoch": 0.03735568791822928,
+      "grad_norm": 0.12174142897129059,
+      "learning_rate": 0.0001999722051343847,
+      "loss": 1.337,
+      "step": 508
+    },
+    {
+      "epoch": 0.03742922273696595,
+      "grad_norm": 0.25044193863868713,
+      "learning_rate": 0.00019997164081270218,
+      "loss": 1.3122,
+      "step": 509
+    },
+    {
+      "epoch": 0.03750275755570263,
+      "grad_norm": 0.1560681015253067,
+      "learning_rate": 0.00019997107082052275,
+      "loss": 1.2552,
+      "step": 510
+    },
+    {
+      "epoch": 0.037576292374439296,
+      "grad_norm": 0.1631559133529663,
+      "learning_rate": 0.00019997049515787867,
+      "loss": 1.3033,
+      "step": 511
+    },
+    {
+      "epoch": 0.03764982719317597,
+      "grad_norm": 0.17368929088115692,
+      "learning_rate": 0.00019996991382480262,
+      "loss": 1.2969,
+      "step": 512
+    },
+    {
+      "epoch": 0.03772336201191264,
+      "grad_norm": 0.130717933177948,
+      "learning_rate": 0.0001999693268213275,
+      "loss": 1.2863,
+      "step": 513
+    },
+    {
+      "epoch": 0.03779689683064931,
+      "grad_norm": 0.133191779255867,
+      "learning_rate": 0.0001999687341474867,
+      "loss": 1.269,
+      "step": 514
+    },
+    {
+      "epoch": 0.037870431649385985,
+      "grad_norm": 0.12593477964401245,
+      "learning_rate": 0.0001999681358033138,
+      "loss": 1.2525,
+      "step": 515
+    },
+    {
+      "epoch": 0.037943966468122654,
+      "grad_norm": 0.12342957407236099,
+      "learning_rate": 0.00019996753178884273,
+      "loss": 1.2616,
+      "step": 516
+    },
+    {
+      "epoch": 0.03801750128685933,
+      "grad_norm": 0.12695534527301788,
+      "learning_rate": 0.00019996692210410774,
+      "loss": 1.2837,
+      "step": 517
+    },
+    {
+      "epoch": 0.038091036105596,
+      "grad_norm": 0.15527509152889252,
+      "learning_rate": 0.00019996630674914346,
+      "loss": 1.269,
+      "step": 518
+    },
+    {
+      "epoch": 0.038164570924332675,
+      "grad_norm": 0.14306145906448364,
+      "learning_rate": 0.00019996568572398476,
+      "loss": 1.2691,
+      "step": 519
+    },
+    {
+      "epoch": 0.038238105743069344,
+      "grad_norm": 0.13390520215034485,
+      "learning_rate": 0.00019996505902866686,
+      "loss": 1.2904,
+      "step": 520
+    },
+    {
+      "epoch": 0.03831164056180601,
+      "grad_norm": 0.13560260832309723,
+      "learning_rate": 0.00019996442666322527,
+      "loss": 1.2643,
+      "step": 521
+    },
+    {
+      "epoch": 0.03838517538054269,
+      "grad_norm": 0.14034047722816467,
+      "learning_rate": 0.000199963788627696,
+      "loss": 1.2738,
+      "step": 522
+    },
+    {
+      "epoch": 0.03845871019927936,
+      "grad_norm": 0.17238809168338776,
+      "learning_rate": 0.00019996314492211505,
+      "loss": 1.2776,
+      "step": 523
+    },
+    {
+      "epoch": 0.03853224501801603,
+      "grad_norm": 0.15615062415599823,
+      "learning_rate": 0.00019996249554651906,
+      "loss": 1.2708,
+      "step": 524
+    },
+    {
+      "epoch": 0.0386057798367527,
+      "grad_norm": 0.15405699610710144,
+      "learning_rate": 0.00019996184050094482,
+      "loss": 1.2648,
+      "step": 525
+    },
+    {
+      "epoch": 0.03867931465548938,
+      "grad_norm": 0.126106858253479,
+      "learning_rate": 0.0001999611797854295,
+      "loss": 1.2545,
+      "step": 526
+    },
+    {
+      "epoch": 0.038752849474226046,
+      "grad_norm": 0.15363343060016632,
+      "learning_rate": 0.00019996051340001058,
+      "loss": 1.256,
+      "step": 527
+    },
+    {
+      "epoch": 0.038826384292962715,
+      "grad_norm": 0.1156841292977333,
+      "learning_rate": 0.00019995984134472586,
+      "loss": 1.257,
+      "step": 528
+    },
+    {
+      "epoch": 0.03889991911169939,
+      "grad_norm": 0.17104220390319824,
+      "learning_rate": 0.00019995916361961342,
+      "loss": 1.2627,
+      "step": 529
+    },
+    {
+      "epoch": 0.03897345393043606,
+      "grad_norm": 0.14539583027362823,
+      "learning_rate": 0.00019995848022471172,
+      "loss": 1.271,
+      "step": 530
+    },
+    {
+      "epoch": 0.039046988749172735,
+      "grad_norm": 0.15870751440525055,
+      "learning_rate": 0.00019995779116005954,
+      "loss": 1.2965,
+      "step": 531
+    },
+    {
+      "epoch": 0.039120523567909404,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00019995709642569594,
+      "loss": 1.2536,
+      "step": 532
+    },
+    {
+      "epoch": 0.03919405838664608,
+      "grad_norm": 0.14622120559215546,
+      "learning_rate": 0.0001999563960216604,
+      "loss": 1.2649,
+      "step": 533
+    },
+    {
+      "epoch": 0.03926759320538275,
+      "grad_norm": 0.12924382090568542,
+      "learning_rate": 0.00019995568994799254,
+      "loss": 1.2818,
+      "step": 534
+    },
+    {
+      "epoch": 0.03934112802411942,
+      "grad_norm": 0.1417049765586853,
+      "learning_rate": 0.00019995497820473248,
+      "loss": 1.2765,
+      "step": 535
+    },
+    {
+      "epoch": 0.039414662842856094,
+      "grad_norm": 0.11714164912700653,
+      "learning_rate": 0.00019995426079192057,
+      "loss": 1.2419,
+      "step": 536
+    },
+    {
+      "epoch": 0.03948819766159276,
+      "grad_norm": 0.11202321946620941,
+      "learning_rate": 0.00019995353770959747,
+      "loss": 1.2671,
+      "step": 537
+    },
+    {
+      "epoch": 0.03956173248032944,
+      "grad_norm": 0.12602698802947998,
+      "learning_rate": 0.00019995280895780427,
+      "loss": 1.2489,
+      "step": 538
+    },
+    {
+      "epoch": 0.03963526729906611,
+      "grad_norm": 0.12062248587608337,
+      "learning_rate": 0.00019995207453658221,
+      "loss": 1.3094,
+      "step": 539
+    },
+    {
+      "epoch": 0.03970880211780278,
+      "grad_norm": 0.11796469241380692,
+      "learning_rate": 0.00019995133444597306,
+      "loss": 1.2317,
+      "step": 540
+    },
+    {
+      "epoch": 0.03978233693653945,
+      "grad_norm": 0.12278874218463898,
+      "learning_rate": 0.00019995058868601867,
+      "loss": 1.2571,
+      "step": 541
+    },
+    {
+      "epoch": 0.03985587175527612,
+      "grad_norm": 0.1197562888264656,
+      "learning_rate": 0.00019994983725676147,
+      "loss": 1.2588,
+      "step": 542
+    },
+    {
+      "epoch": 0.039929406574012796,
+      "grad_norm": 0.13917088508605957,
+      "learning_rate": 0.000199949080158244,
+      "loss": 1.2752,
+      "step": 543
+    },
+    {
+      "epoch": 0.040002941392749465,
+      "grad_norm": 0.1304134726524353,
+      "learning_rate": 0.0001999483173905092,
+      "loss": 1.2533,
+      "step": 544
+    },
+    {
+      "epoch": 0.04007647621148614,
+      "grad_norm": 0.13553845882415771,
+      "learning_rate": 0.00019994754895360037,
+      "loss": 1.2778,
+      "step": 545
+    },
+    {
+      "epoch": 0.04015001103022281,
+      "grad_norm": 0.137639582157135,
+      "learning_rate": 0.0001999467748475611,
+      "loss": 1.2579,
+      "step": 546
+    },
+    {
+      "epoch": 0.040223545848959485,
+      "grad_norm": 0.160695418715477,
+      "learning_rate": 0.00019994599507243524,
+      "loss": 1.2692,
+      "step": 547
+    },
+    {
+      "epoch": 0.040297080667696154,
+      "grad_norm": 0.12828147411346436,
+      "learning_rate": 0.0001999452096282671,
+      "loss": 1.2641,
+      "step": 548
+    },
+    {
+      "epoch": 0.04037061548643282,
+      "grad_norm": 0.12474434822797775,
+      "learning_rate": 0.0001999444185151012,
+      "loss": 1.2936,
+      "step": 549
+    },
+    {
+      "epoch": 0.0404441503051695,
+      "grad_norm": 0.12575843930244446,
+      "learning_rate": 0.0001999436217329824,
+      "loss": 1.2907,
+      "step": 550
+    },
+    {
+      "epoch": 0.04051768512390617,
+      "grad_norm": 0.13603703677654266,
+      "learning_rate": 0.0001999428192819559,
+      "loss": 1.2706,
+      "step": 551
+    },
+    {
+      "epoch": 0.040591219942642844,
+      "grad_norm": 0.12826284766197205,
+      "learning_rate": 0.00019994201116206722,
+      "loss": 1.2701,
+      "step": 552
+    },
+    {
+      "epoch": 0.04066475476137951,
+      "grad_norm": 0.12777231633663177,
+      "learning_rate": 0.00019994119737336218,
+      "loss": 1.249,
+      "step": 553
+    },
+    {
+      "epoch": 0.04073828958011619,
+      "grad_norm": 0.12681232392787933,
+      "learning_rate": 0.00019994037791588697,
+      "loss": 1.2401,
+      "step": 554
+    },
+    {
+      "epoch": 0.04081182439885286,
+      "grad_norm": 0.1858748346567154,
+      "learning_rate": 0.00019993955278968802,
+      "loss": 1.2769,
+      "step": 555
+    },
+    {
+      "epoch": 0.040885359217589526,
+      "grad_norm": 0.1558796465396881,
+      "learning_rate": 0.00019993872199481217,
+      "loss": 1.2448,
+      "step": 556
+    },
+    {
+      "epoch": 0.0409588940363262,
+      "grad_norm": 0.1965806931257248,
+      "learning_rate": 0.00019993788553130656,
+      "loss": 1.2623,
+      "step": 557
+    },
+    {
+      "epoch": 0.04103242885506287,
+      "grad_norm": 0.17347517609596252,
+      "learning_rate": 0.00019993704339921862,
+      "loss": 1.2806,
+      "step": 558
+    },
+    {
+      "epoch": 0.041105963673799546,
+      "grad_norm": 0.21769075095653534,
+      "learning_rate": 0.0001999361955985961,
+      "loss": 1.2335,
+      "step": 559
+    },
+    {
+      "epoch": 0.041179498492536215,
+      "grad_norm": 0.16181844472885132,
+      "learning_rate": 0.00019993534212948708,
+      "loss": 1.2214,
+      "step": 560
+    },
+    {
+      "epoch": 0.04125303331127289,
+      "grad_norm": 0.16399718821048737,
+      "learning_rate": 0.00019993448299194,
+      "loss": 1.2211,
+      "step": 561
+    },
+    {
+      "epoch": 0.04132656813000956,
+      "grad_norm": 0.18589700758457184,
+      "learning_rate": 0.0001999336181860036,
+      "loss": 1.2971,
+      "step": 562
+    },
+    {
+      "epoch": 0.04140010294874623,
+      "grad_norm": 0.14839410781860352,
+      "learning_rate": 0.0001999327477117269,
+      "loss": 1.2412,
+      "step": 563
+    },
+    {
+      "epoch": 0.041473637767482904,
+      "grad_norm": 0.16526144742965698,
+      "learning_rate": 0.0001999318715691593,
+      "loss": 1.2256,
+      "step": 564
+    },
+    {
+      "epoch": 0.04154717258621957,
+      "grad_norm": 0.1342661827802658,
+      "learning_rate": 0.00019993098975835047,
+      "loss": 1.2625,
+      "step": 565
+    },
+    {
+      "epoch": 0.04162070740495625,
+      "grad_norm": 0.18630363047122955,
+      "learning_rate": 0.00019993010227935043,
+      "loss": 1.2083,
+      "step": 566
+    },
+    {
+      "epoch": 0.04169424222369292,
+      "grad_norm": 0.12657314538955688,
+      "learning_rate": 0.00019992920913220955,
+      "loss": 1.2538,
+      "step": 567
+    },
+    {
+      "epoch": 0.041767777042429594,
+      "grad_norm": 0.16338984668254852,
+      "learning_rate": 0.00019992831031697843,
+      "loss": 1.2744,
+      "step": 568
+    },
+    {
+      "epoch": 0.04184131186116626,
+      "grad_norm": 0.10938669741153717,
+      "learning_rate": 0.00019992740583370814,
+      "loss": 1.2416,
+      "step": 569
+    },
+    {
+      "epoch": 0.04191484667990293,
+      "grad_norm": 0.1782827526330948,
+      "learning_rate": 0.0001999264956824499,
+      "loss": 1.2746,
+      "step": 570
+    },
+    {
+      "epoch": 0.04198838149863961,
+      "grad_norm": 0.13834355771541595,
+      "learning_rate": 0.00019992557986325536,
+      "loss": 1.2193,
+      "step": 571
+    },
+    {
+      "epoch": 0.042061916317376276,
+      "grad_norm": 0.1679130494594574,
+      "learning_rate": 0.0001999246583761765,
+      "loss": 1.2464,
+      "step": 572
+    },
+    {
+      "epoch": 0.04213545113611295,
+      "grad_norm": 0.15923476219177246,
+      "learning_rate": 0.00019992373122126552,
+      "loss": 1.2583,
+      "step": 573
+    },
+    {
+      "epoch": 0.04220898595484962,
+      "grad_norm": 0.1477038562297821,
+      "learning_rate": 0.00019992279839857512,
+      "loss": 1.2647,
+      "step": 574
+    },
+    {
+      "epoch": 0.042282520773586296,
+      "grad_norm": 0.12338787317276001,
+      "learning_rate": 0.0001999218599081581,
+      "loss": 1.2839,
+      "step": 575
+    },
+    {
+      "epoch": 0.042356055592322965,
+      "grad_norm": 0.1304710954427719,
+      "learning_rate": 0.00019992091575006776,
+      "loss": 1.2215,
+      "step": 576
+    },
+    {
+      "epoch": 0.042429590411059634,
+      "grad_norm": 0.15191765129566193,
+      "learning_rate": 0.00019991996592435757,
+      "loss": 1.2154,
+      "step": 577
+    },
+    {
+      "epoch": 0.04250312522979631,
+      "grad_norm": 0.12123032659292221,
+      "learning_rate": 0.00019991901043108152,
+      "loss": 1.2115,
+      "step": 578
+    },
+    {
+      "epoch": 0.04257666004853298,
+      "grad_norm": 0.1599038690328598,
+      "learning_rate": 0.00019991804927029373,
+      "loss": 1.2215,
+      "step": 579
+    },
+    {
+      "epoch": 0.042650194867269654,
+      "grad_norm": 0.13235723972320557,
+      "learning_rate": 0.00019991708244204874,
+      "loss": 1.1829,
+      "step": 580
+    },
+    {
+      "epoch": 0.04272372968600632,
+      "grad_norm": 0.18952064216136932,
+      "learning_rate": 0.0001999161099464014,
+      "loss": 1.2448,
+      "step": 581
+    },
+    {
+      "epoch": 0.042797264504743,
+      "grad_norm": 0.12207041680812836,
+      "learning_rate": 0.00019991513178340683,
+      "loss": 1.1943,
+      "step": 582
+    },
+    {
+      "epoch": 0.04287079932347967,
+      "grad_norm": 0.21172550320625305,
+      "learning_rate": 0.00019991414795312052,
+      "loss": 1.2157,
+      "step": 583
+    },
+    {
+      "epoch": 0.04294433414221634,
+      "grad_norm": 0.13774508237838745,
+      "learning_rate": 0.00019991315845559836,
+      "loss": 1.2022,
+      "step": 584
+    },
+    {
+      "epoch": 0.04301786896095301,
+      "grad_norm": 0.15768995881080627,
+      "learning_rate": 0.00019991216329089638,
+      "loss": 1.2233,
+      "step": 585
+    },
+    {
+      "epoch": 0.04309140377968968,
+      "grad_norm": 0.1420089453458786,
+      "learning_rate": 0.00019991116245907103,
+      "loss": 1.1996,
+      "step": 586
+    },
+    {
+      "epoch": 0.04316493859842636,
+      "grad_norm": 0.16509822010993958,
+      "learning_rate": 0.00019991015596017918,
+      "loss": 1.186,
+      "step": 587
+    },
+    {
+      "epoch": 0.043238473417163026,
+      "grad_norm": 0.14053432643413544,
+      "learning_rate": 0.00019990914379427777,
+      "loss": 1.2143,
+      "step": 588
+    },
+    {
+      "epoch": 0.0433120082358997,
+      "grad_norm": 0.17106956243515015,
+      "learning_rate": 0.00019990812596142429,
+      "loss": 1.1833,
+      "step": 589
+    },
+    {
+      "epoch": 0.04338554305463637,
+      "grad_norm": 0.17074918746948242,
+      "learning_rate": 0.0001999071024616765,
+      "loss": 1.2297,
+      "step": 590
+    },
+    {
+      "epoch": 0.04345907787337304,
+      "grad_norm": 0.12714025378227234,
+      "learning_rate": 0.0001999060732950924,
+      "loss": 1.1859,
+      "step": 591
+    },
+    {
+      "epoch": 0.043532612692109715,
+      "grad_norm": 0.17084473371505737,
+      "learning_rate": 0.00019990503846173044,
+      "loss": 1.1917,
+      "step": 592
+    },
+    {
+      "epoch": 0.043606147510846384,
+      "grad_norm": 0.13978450000286102,
+      "learning_rate": 0.0001999039979616492,
+      "loss": 1.1997,
+      "step": 593
+    },
+    {
+      "epoch": 0.04367968232958306,
+      "grad_norm": 0.17776277661323547,
+      "learning_rate": 0.0001999029517949078,
+      "loss": 1.2435,
+      "step": 594
+    },
+    {
+      "epoch": 0.04375321714831973,
+      "grad_norm": 0.11459385603666306,
+      "learning_rate": 0.00019990189996156553,
+      "loss": 1.2023,
+      "step": 595
+    },
+    {
+      "epoch": 0.043826751967056404,
+      "grad_norm": 0.12233222275972366,
+      "learning_rate": 0.00019990084246168205,
+      "loss": 1.2289,
+      "step": 596
+    },
+    {
+      "epoch": 0.04390028678579307,
+      "grad_norm": 0.1477910876274109,
+      "learning_rate": 0.00019989977929531737,
+      "loss": 1.188,
+      "step": 597
+    },
+    {
+      "epoch": 0.04397382160452974,
+      "grad_norm": 0.13479216396808624,
+      "learning_rate": 0.0001998987104625318,
+      "loss": 1.2156,
+      "step": 598
+    },
+    {
+      "epoch": 0.04404735642326642,
+      "grad_norm": 0.13216571509838104,
+      "learning_rate": 0.00019989763596338592,
+      "loss": 1.225,
+      "step": 599
+    },
+    {
+      "epoch": 0.04412089124200309,
+      "grad_norm": 0.16002953052520752,
+      "learning_rate": 0.00019989655579794068,
+      "loss": 1.1973,
+      "step": 600
+    },
+    {
+      "epoch": 0.04419442606073976,
+      "grad_norm": 0.16152030229568481,
+      "learning_rate": 0.0001998954699662574,
+      "loss": 1.1896,
+      "step": 601
+    },
+    {
+      "epoch": 0.04426796087947643,
+      "grad_norm": 0.15902768075466156,
+      "learning_rate": 0.00019989437846839762,
+      "loss": 1.1537,
+      "step": 602
+    },
+    {
+      "epoch": 0.04434149569821311,
+      "grad_norm": 0.14405213296413422,
+      "learning_rate": 0.00019989328130442328,
+      "loss": 1.1791,
+      "step": 603
+    },
+    {
+      "epoch": 0.044415030516949776,
+      "grad_norm": 0.17151133716106415,
+      "learning_rate": 0.00019989217847439663,
+      "loss": 1.1769,
+      "step": 604
+    },
+    {
+      "epoch": 0.044488565335686445,
+      "grad_norm": 0.11376479268074036,
+      "learning_rate": 0.00019989106997838016,
+      "loss": 1.1686,
+      "step": 605
+    },
+    {
+      "epoch": 0.04456210015442312,
+      "grad_norm": 0.17980825901031494,
+      "learning_rate": 0.00019988995581643682,
+      "loss": 1.1863,
+      "step": 606
+    },
+    {
+      "epoch": 0.04463563497315979,
+      "grad_norm": 0.10117840766906738,
+      "learning_rate": 0.00019988883598862973,
+      "loss": 1.1875,
+      "step": 607
+    },
+    {
+      "epoch": 0.044709169791896465,
+      "grad_norm": 0.14440101385116577,
+      "learning_rate": 0.00019988771049502245,
+      "loss": 1.1542,
+      "step": 608
+    },
+    {
+      "epoch": 0.044782704610633134,
+      "grad_norm": 0.1276853084564209,
+      "learning_rate": 0.00019988657933567885,
+      "loss": 1.2036,
+      "step": 609
+    },
+    {
+      "epoch": 0.04485623942936981,
+      "grad_norm": 0.1428319811820984,
+      "learning_rate": 0.000199885442510663,
+      "loss": 1.1556,
+      "step": 610
+    },
+    {
+      "epoch": 0.04492977424810648,
+      "grad_norm": 0.13773562014102936,
+      "learning_rate": 0.0001998843000200395,
+      "loss": 1.1695,
+      "step": 611
+    },
+    {
+      "epoch": 0.04500330906684315,
+      "grad_norm": 0.16761422157287598,
+      "learning_rate": 0.00019988315186387305,
+      "loss": 1.2018,
+      "step": 612
+    },
+    {
+      "epoch": 0.04507684388557982,
+      "grad_norm": 0.1521965116262436,
+      "learning_rate": 0.00019988199804222884,
+      "loss": 1.1558,
+      "step": 613
+    },
+    {
+      "epoch": 0.04515037870431649,
+      "grad_norm": 0.19961011409759521,
+      "learning_rate": 0.00019988083855517226,
+      "loss": 1.2084,
+      "step": 614
+    },
+    {
+      "epoch": 0.04522391352305317,
+      "grad_norm": 0.19364510476589203,
+      "learning_rate": 0.00019987967340276912,
+      "loss": 1.1599,
+      "step": 615
+    },
+    {
+      "epoch": 0.04529744834178984,
+      "grad_norm": 0.14366336166858673,
+      "learning_rate": 0.00019987850258508551,
+      "loss": 1.1793,
+      "step": 616
+    },
+    {
+      "epoch": 0.04537098316052651,
+      "grad_norm": 0.2933100461959839,
+      "learning_rate": 0.00019987732610218784,
+      "loss": 1.1764,
+      "step": 617
+    },
+    {
+      "epoch": 0.04544451797926318,
+      "grad_norm": 0.21911568939685822,
+      "learning_rate": 0.00019987614395414278,
+      "loss": 1.1572,
+      "step": 618
+    },
+    {
+      "epoch": 0.04551805279799985,
+      "grad_norm": 0.14135406911373138,
+      "learning_rate": 0.00019987495614101747,
+      "loss": 1.172,
+      "step": 619
+    },
+    {
+      "epoch": 0.045591587616736526,
+      "grad_norm": 0.2571685314178467,
+      "learning_rate": 0.00019987376266287924,
+      "loss": 1.1679,
+      "step": 620
+    },
+    {
+      "epoch": 0.045665122435473195,
+      "grad_norm": 0.1510409712791443,
+      "learning_rate": 0.00019987256351979577,
+      "loss": 1.1738,
+      "step": 621
+    },
+    {
+      "epoch": 0.04573865725420987,
+      "grad_norm": 0.17767217755317688,
+      "learning_rate": 0.00019987135871183512,
+      "loss": 1.1561,
+      "step": 622
+    },
+    {
+      "epoch": 0.04581219207294654,
+      "grad_norm": 0.14005716145038605,
+      "learning_rate": 0.0001998701482390656,
+      "loss": 1.1997,
+      "step": 623
+    },
+    {
+      "epoch": 0.045885726891683215,
+      "grad_norm": 0.18734516203403473,
+      "learning_rate": 0.0001998689321015559,
+      "loss": 1.1583,
+      "step": 624
+    },
+    {
+      "epoch": 0.045959261710419884,
+      "grad_norm": 0.1571679711341858,
+      "learning_rate": 0.00019986771029937493,
+      "loss": 1.1682,
+      "step": 625
+    },
+    {
+      "epoch": 0.04603279652915655,
+      "grad_norm": 0.15152455866336823,
+      "learning_rate": 0.00019986648283259206,
+      "loss": 1.1748,
+      "step": 626
+    },
+    {
+      "epoch": 0.04610633134789323,
+      "grad_norm": 0.15373283624649048,
+      "learning_rate": 0.00019986524970127688,
+      "loss": 1.1842,
+      "step": 627
+    },
+    {
+      "epoch": 0.0461798661666299,
+      "grad_norm": 0.17395956814289093,
+      "learning_rate": 0.00019986401090549936,
+      "loss": 1.1861,
+      "step": 628
+    },
+    {
+      "epoch": 0.04625340098536657,
+      "grad_norm": 0.15966609120368958,
+      "learning_rate": 0.00019986276644532973,
+      "loss": 1.1703,
+      "step": 629
+    },
+    {
+      "epoch": 0.04632693580410324,
+      "grad_norm": 0.15236185491085052,
+      "learning_rate": 0.0001998615163208386,
+      "loss": 1.1623,
+      "step": 630
+    },
+    {
+      "epoch": 0.04640047062283992,
+      "grad_norm": 0.1609046310186386,
+      "learning_rate": 0.0001998602605320969,
+      "loss": 1.1715,
+      "step": 631
+    },
+    {
+      "epoch": 0.04647400544157659,
+      "grad_norm": 0.1398119181394577,
+      "learning_rate": 0.00019985899907917583,
+      "loss": 1.191,
+      "step": 632
+    },
+    {
+      "epoch": 0.046547540260313255,
+      "grad_norm": 0.1180499866604805,
+      "learning_rate": 0.00019985773196214693,
+      "loss": 1.1566,
+      "step": 633
+    },
+    {
+      "epoch": 0.04662107507904993,
+      "grad_norm": 0.13246628642082214,
+      "learning_rate": 0.0001998564591810821,
+      "loss": 1.1862,
+      "step": 634
+    },
+    {
+      "epoch": 0.0466946098977866,
+      "grad_norm": 0.14615857601165771,
+      "learning_rate": 0.00019985518073605351,
+      "loss": 1.1701,
+      "step": 635
+    },
+    {
+      "epoch": 0.046768144716523276,
+      "grad_norm": 0.12159596383571625,
+      "learning_rate": 0.00019985389662713373,
+      "loss": 1.1368,
+      "step": 636
+    },
+    {
+      "epoch": 0.046841679535259945,
+      "grad_norm": 0.13723962008953094,
+      "learning_rate": 0.0001998526068543955,
+      "loss": 1.1439,
+      "step": 637
+    },
+    {
+      "epoch": 0.04691521435399662,
+      "grad_norm": 0.12192691117525101,
+      "learning_rate": 0.00019985131141791204,
+      "loss": 1.1631,
+      "step": 638
+    },
+    {
+      "epoch": 0.04698874917273329,
+      "grad_norm": 0.15148544311523438,
+      "learning_rate": 0.00019985001031775683,
+      "loss": 1.166,
+      "step": 639
+    },
+    {
+      "epoch": 0.04706228399146996,
+      "grad_norm": 0.12088466435670853,
+      "learning_rate": 0.00019984870355400366,
+      "loss": 1.1779,
+      "step": 640
+    },
+    {
+      "epoch": 0.047135818810206634,
+      "grad_norm": 0.14709223806858063,
+      "learning_rate": 0.00019984739112672667,
+      "loss": 1.1741,
+      "step": 641
+    },
+    {
+      "epoch": 0.0472093536289433,
+      "grad_norm": 0.12241767346858978,
+      "learning_rate": 0.00019984607303600026,
+      "loss": 1.1674,
+      "step": 642
+    },
+    {
+      "epoch": 0.04728288844767998,
+      "grad_norm": 0.1415087729692459,
+      "learning_rate": 0.00019984474928189921,
+      "loss": 1.167,
+      "step": 643
+    },
+    {
+      "epoch": 0.04735642326641665,
+      "grad_norm": 0.10452502220869064,
+      "learning_rate": 0.00019984341986449861,
+      "loss": 1.2414,
+      "step": 644
+    },
+    {
+      "epoch": 0.04742995808515332,
+      "grad_norm": 0.13781610131263733,
+      "learning_rate": 0.00019984208478387387,
+      "loss": 1.1479,
+      "step": 645
+    },
+    {
+      "epoch": 0.04750349290388999,
+      "grad_norm": 0.17079167068004608,
+      "learning_rate": 0.0001998407440401007,
+      "loss": 1.1772,
+      "step": 646
+    },
+    {
+      "epoch": 0.04757702772262666,
+      "grad_norm": 0.12031634896993637,
+      "learning_rate": 0.0001998393976332552,
+      "loss": 1.1762,
+      "step": 647
+    },
+    {
+      "epoch": 0.04765056254136334,
+      "grad_norm": 0.1425374299287796,
+      "learning_rate": 0.00019983804556341367,
+      "loss": 1.1527,
+      "step": 648
+    },
+    {
+      "epoch": 0.047724097360100005,
+      "grad_norm": 0.10987982898950577,
+      "learning_rate": 0.00019983668783065287,
+      "loss": 1.1408,
+      "step": 649
+    },
+    {
+      "epoch": 0.04779763217883668,
+      "grad_norm": 0.12323692440986633,
+      "learning_rate": 0.00019983532443504975,
+      "loss": 1.1715,
+      "step": 650
+    },
+    {
+      "epoch": 0.04787116699757335,
+      "grad_norm": 0.11918117851018906,
+      "learning_rate": 0.00019983395537668167,
+      "loss": 1.1838,
+      "step": 651
+    },
+    {
+      "epoch": 0.047944701816310026,
+      "grad_norm": 0.13060833513736725,
+      "learning_rate": 0.00019983258065562628,
+      "loss": 1.197,
+      "step": 652
+    },
+    {
+      "epoch": 0.048018236635046695,
+      "grad_norm": 0.13385479152202606,
+      "learning_rate": 0.00019983120027196154,
+      "loss": 1.1359,
+      "step": 653
+    },
+    {
+      "epoch": 0.048091771453783363,
+      "grad_norm": 0.1583932489156723,
+      "learning_rate": 0.00019982981422576578,
+      "loss": 1.1596,
+      "step": 654
+    },
+    {
+      "epoch": 0.04816530627252004,
+      "grad_norm": 0.1265403926372528,
+      "learning_rate": 0.0001998284225171176,
+      "loss": 1.1788,
+      "step": 655
+    },
+    {
+      "epoch": 0.04823884109125671,
+      "grad_norm": 0.1402433216571808,
+      "learning_rate": 0.00019982702514609595,
+      "loss": 1.1865,
+      "step": 656
+    },
+    {
+      "epoch": 0.048312375909993384,
+      "grad_norm": 0.2232215255498886,
+      "learning_rate": 0.00019982562211278007,
+      "loss": 1.1787,
+      "step": 657
+    },
+    {
+      "epoch": 0.04838591072873005,
+      "grad_norm": 0.2733139991760254,
+      "learning_rate": 0.00019982421341724956,
+      "loss": 1.1358,
+      "step": 658
+    },
+    {
+      "epoch": 0.04845944554746673,
+      "grad_norm": 0.19811305403709412,
+      "learning_rate": 0.00019982279905958431,
+      "loss": 1.1808,
+      "step": 659
+    },
+    {
+      "epoch": 0.0485329803662034,
+      "grad_norm": 0.175303652882576,
+      "learning_rate": 0.00019982137903986456,
+      "loss": 1.1379,
+      "step": 660
+    },
+    {
+      "epoch": 0.048606515184940066,
+      "grad_norm": 0.21475018560886383,
+      "learning_rate": 0.00019981995335817084,
+      "loss": 1.1259,
+      "step": 661
+    },
+    {
+      "epoch": 0.04868005000367674,
+      "grad_norm": 0.14543622732162476,
+      "learning_rate": 0.00019981852201458403,
+      "loss": 1.1489,
+      "step": 662
+    },
+    {
+      "epoch": 0.04875358482241341,
+      "grad_norm": 0.2561386227607727,
+      "learning_rate": 0.0001998170850091853,
+      "loss": 1.1557,
+      "step": 663
+    },
+    {
+      "epoch": 0.04882711964115009,
+      "grad_norm": 0.15503980219364166,
+      "learning_rate": 0.00019981564234205617,
+      "loss": 1.1836,
+      "step": 664
+    },
+    {
+      "epoch": 0.048900654459886755,
+      "grad_norm": 0.1714758425951004,
+      "learning_rate": 0.00019981419401327846,
+      "loss": 1.1895,
+      "step": 665
+    },
+    {
+      "epoch": 0.04897418927862343,
+      "grad_norm": 0.14249791204929352,
+      "learning_rate": 0.00019981274002293432,
+      "loss": 1.1554,
+      "step": 666
+    },
+    {
+      "epoch": 0.0490477240973601,
+      "grad_norm": 0.1494312882423401,
+      "learning_rate": 0.00019981128037110626,
+      "loss": 1.1991,
+      "step": 667
+    },
+    {
+      "epoch": 0.04912125891609677,
+      "grad_norm": 0.1556432843208313,
+      "learning_rate": 0.00019980981505787702,
+      "loss": 1.1513,
+      "step": 668
+    },
+    {
+      "epoch": 0.049194793734833445,
+      "grad_norm": 0.13401061296463013,
+      "learning_rate": 0.00019980834408332972,
+      "loss": 1.1407,
+      "step": 669
+    },
+    {
+      "epoch": 0.049268328553570113,
+      "grad_norm": 0.12537458539009094,
+      "learning_rate": 0.00019980686744754786,
+      "loss": 1.1457,
+      "step": 670
+    },
+    {
+      "epoch": 0.04934186337230679,
+      "grad_norm": 0.17435936629772186,
+      "learning_rate": 0.0001998053851506151,
+      "loss": 1.1417,
+      "step": 671
+    },
+    {
+      "epoch": 0.04941539819104346,
+      "grad_norm": 0.13490881025791168,
+      "learning_rate": 0.0001998038971926156,
+      "loss": 1.1426,
+      "step": 672
+    },
+    {
+      "epoch": 0.049488933009780134,
+      "grad_norm": 0.132877916097641,
+      "learning_rate": 0.00019980240357363372,
+      "loss": 1.1868,
+      "step": 673
+    },
+    {
+      "epoch": 0.0495624678285168,
+      "grad_norm": 0.17138533294200897,
+      "learning_rate": 0.00019980090429375417,
+      "loss": 1.152,
+      "step": 674
+    },
+    {
+      "epoch": 0.04963600264725347,
+      "grad_norm": 0.11379387229681015,
+      "learning_rate": 0.00019979939935306203,
+      "loss": 1.1694,
+      "step": 675
+    },
+    {
+      "epoch": 0.04970953746599015,
+      "grad_norm": 0.13884910941123962,
+      "learning_rate": 0.0001997978887516426,
+      "loss": 1.1601,
+      "step": 676
+    },
+    {
+      "epoch": 0.049783072284726816,
+      "grad_norm": 0.12439777702093124,
+      "learning_rate": 0.0001997963724895816,
+      "loss": 1.1346,
+      "step": 677
+    },
+    {
+      "epoch": 0.04985660710346349,
+      "grad_norm": 0.13561555743217468,
+      "learning_rate": 0.00019979485056696505,
+      "loss": 1.1626,
+      "step": 678
+    },
+    {
+      "epoch": 0.04993014192220016,
+      "grad_norm": 0.14887495338916779,
+      "learning_rate": 0.00019979332298387926,
+      "loss": 1.1659,
+      "step": 679
+    },
+    {
+      "epoch": 0.05000367674093684,
+      "grad_norm": 0.16662698984146118,
+      "learning_rate": 0.00019979178974041087,
+      "loss": 1.1642,
+      "step": 680
+    },
+    {
+      "epoch": 0.050077211559673505,
+      "grad_norm": 0.13602270185947418,
+      "learning_rate": 0.00019979025083664685,
+      "loss": 1.1624,
+      "step": 681
+    },
+    {
+      "epoch": 0.050150746378410174,
+      "grad_norm": 0.1344245821237564,
+      "learning_rate": 0.00019978870627267446,
+      "loss": 1.1656,
+      "step": 682
+    },
+    {
+      "epoch": 0.05022428119714685,
+      "grad_norm": 0.1382242739200592,
+      "learning_rate": 0.00019978715604858138,
+      "loss": 1.1424,
+      "step": 683
+    },
+    {
+      "epoch": 0.05029781601588352,
+      "grad_norm": 0.16264671087265015,
+      "learning_rate": 0.0001997856001644555,
+      "loss": 1.1978,
+      "step": 684
+    },
+    {
+      "epoch": 0.050371350834620195,
+      "grad_norm": 0.14416548609733582,
+      "learning_rate": 0.00019978403862038507,
+      "loss": 1.184,
+      "step": 685
+    },
+    {
+      "epoch": 0.050444885653356863,
+      "grad_norm": 0.15558549761772156,
+      "learning_rate": 0.00019978247141645862,
+      "loss": 1.1624,
+      "step": 686
+    },
+    {
+      "epoch": 0.05051842047209354,
+      "grad_norm": 0.16009987890720367,
+      "learning_rate": 0.00019978089855276514,
+      "loss": 1.1259,
+      "step": 687
+    },
+    {
+      "epoch": 0.05059195529083021,
+      "grad_norm": 0.14287689328193665,
+      "learning_rate": 0.00019977932002939377,
+      "loss": 1.1637,
+      "step": 688
+    },
+    {
+      "epoch": 0.05066549010956688,
+      "grad_norm": 0.16049784421920776,
+      "learning_rate": 0.00019977773584643407,
+      "loss": 1.1628,
+      "step": 689
+    },
+    {
+      "epoch": 0.05073902492830355,
+      "grad_norm": 0.13123364746570587,
+      "learning_rate": 0.0001997761460039759,
+      "loss": 1.163,
+      "step": 690
+    },
+    {
+      "epoch": 0.05081255974704022,
+      "grad_norm": 0.1405520737171173,
+      "learning_rate": 0.0001997745505021094,
+      "loss": 1.1938,
+      "step": 691
+    },
+    {
+      "epoch": 0.0508860945657769,
+      "grad_norm": 0.12127561867237091,
+      "learning_rate": 0.00019977294934092515,
+      "loss": 1.1568,
+      "step": 692
+    },
+    {
+      "epoch": 0.050959629384513566,
+      "grad_norm": 0.14394094049930573,
+      "learning_rate": 0.0001997713425205139,
+      "loss": 1.1582,
+      "step": 693
+    },
+    {
+      "epoch": 0.05103316420325024,
+      "grad_norm": 0.17823369801044464,
+      "learning_rate": 0.00019976973004096676,
+      "loss": 1.171,
+      "step": 694
+    },
+    {
+      "epoch": 0.05110669902198691,
+      "grad_norm": 0.13323648273944855,
+      "learning_rate": 0.00019976811190237527,
+      "loss": 1.153,
+      "step": 695
+    },
+    {
+      "epoch": 0.05118023384072358,
+      "grad_norm": 0.14252136647701263,
+      "learning_rate": 0.0001997664881048312,
+      "loss": 1.1581,
+      "step": 696
+    },
+    {
+      "epoch": 0.051253768659460255,
+      "grad_norm": 0.15850652754306793,
+      "learning_rate": 0.00019976485864842665,
+      "loss": 1.1594,
+      "step": 697
+    },
+    {
+      "epoch": 0.051327303478196924,
+      "grad_norm": 0.16661126911640167,
+      "learning_rate": 0.00019976322353325402,
+      "loss": 1.14,
+      "step": 698
+    },
+    {
+      "epoch": 0.0514008382969336,
+      "grad_norm": 0.16577155888080597,
+      "learning_rate": 0.000199761582759406,
+      "loss": 1.1513,
+      "step": 699
+    },
+    {
+      "epoch": 0.05147437311567027,
+      "grad_norm": 0.2150808870792389,
+      "learning_rate": 0.00019975993632697582,
+      "loss": 1.1307,
+      "step": 700
+    },
+    {
+      "epoch": 0.051547907934406945,
+      "grad_norm": 0.164160817861557,
+      "learning_rate": 0.00019975828423605672,
+      "loss": 1.155,
+      "step": 701
+    },
+    {
+      "epoch": 0.05162144275314361,
+      "grad_norm": 0.24409367144107819,
+      "learning_rate": 0.00019975662648674246,
+      "loss": 1.1666,
+      "step": 702
+    },
+    {
+      "epoch": 0.05169497757188028,
+      "grad_norm": 0.14832551777362823,
+      "learning_rate": 0.00019975496307912712,
+      "loss": 1.1669,
+      "step": 703
+    },
+    {
+      "epoch": 0.05176851239061696,
+      "grad_norm": 0.16623827815055847,
+      "learning_rate": 0.00019975329401330494,
+      "loss": 1.1265,
+      "step": 704
+    },
+    {
+      "epoch": 0.05184204720935363,
+      "grad_norm": 0.1744905710220337,
+      "learning_rate": 0.00019975161928937064,
+      "loss": 1.1616,
+      "step": 705
+    },
+    {
+      "epoch": 0.0519155820280903,
+      "grad_norm": 0.16726204752922058,
+      "learning_rate": 0.00019974993890741928,
+      "loss": 1.1061,
+      "step": 706
+    },
+    {
+      "epoch": 0.05198911684682697,
+      "grad_norm": 0.21289455890655518,
+      "learning_rate": 0.00019974825286754605,
+      "loss": 1.1929,
+      "step": 707
+    },
+    {
+      "epoch": 0.05206265166556365,
+      "grad_norm": 0.14401444792747498,
+      "learning_rate": 0.00019974656116984672,
+      "loss": 1.145,
+      "step": 708
+    },
+    {
+      "epoch": 0.052136186484300316,
+      "grad_norm": 0.21082563698291779,
+      "learning_rate": 0.00019974486381441712,
+      "loss": 1.1258,
+      "step": 709
+    },
+    {
+      "epoch": 0.052209721303036985,
+      "grad_norm": 0.1656319797039032,
+      "learning_rate": 0.0001997431608013536,
+      "loss": 1.1668,
+      "step": 710
+    },
+    {
+      "epoch": 0.05228325612177366,
+      "grad_norm": 0.18572795391082764,
+      "learning_rate": 0.00019974145213075273,
+      "loss": 1.1267,
+      "step": 711
+    },
+    {
+      "epoch": 0.05235679094051033,
+      "grad_norm": 0.2058330476284027,
+      "learning_rate": 0.00019973973780271144,
+      "loss": 1.1282,
+      "step": 712
+    },
+    {
+      "epoch": 0.052430325759247005,
+      "grad_norm": 0.15663354098796844,
+      "learning_rate": 0.00019973801781732694,
+      "loss": 1.1432,
+      "step": 713
+    },
+    {
+      "epoch": 0.052503860577983674,
+      "grad_norm": 0.18617095053195953,
+      "learning_rate": 0.00019973629217469684,
+      "loss": 1.1311,
+      "step": 714
+    },
+    {
+      "epoch": 0.05257739539672035,
+      "grad_norm": 0.1889110803604126,
+      "learning_rate": 0.000199734560874919,
+      "loss": 1.1224,
+      "step": 715
+    },
+    {
+      "epoch": 0.05265093021545702,
+      "grad_norm": 0.14481861889362335,
+      "learning_rate": 0.0001997328239180916,
+      "loss": 1.114,
+      "step": 716
+    },
+    {
+      "epoch": 0.05272446503419369,
+      "grad_norm": 0.14674606919288635,
+      "learning_rate": 0.00019973108130431314,
+      "loss": 1.1096,
+      "step": 717
+    },
+    {
+      "epoch": 0.05279799985293036,
+      "grad_norm": 0.2109202891588211,
+      "learning_rate": 0.00019972933303368255,
+      "loss": 1.1489,
+      "step": 718
+    },
+    {
+      "epoch": 0.05287153467166703,
+      "grad_norm": 0.2580674886703491,
+      "learning_rate": 0.0001997275791062989,
+      "loss": 1.1219,
+      "step": 719
+    },
+    {
+      "epoch": 0.05294506949040371,
+      "grad_norm": 0.2616204023361206,
+      "learning_rate": 0.00019972581952226176,
+      "loss": 1.1447,
+      "step": 720
+    },
+    {
+      "epoch": 0.05301860430914038,
+      "grad_norm": 0.163419708609581,
+      "learning_rate": 0.00019972405428167085,
+      "loss": 1.1432,
+      "step": 721
+    },
+    {
+      "epoch": 0.05309213912787705,
+      "grad_norm": 0.24269454181194305,
+      "learning_rate": 0.00019972228338462636,
+      "loss": 1.1253,
+      "step": 722
+    },
+    {
+      "epoch": 0.05316567394661372,
+      "grad_norm": 0.2597217857837677,
+      "learning_rate": 0.00019972050683122872,
+      "loss": 1.0965,
+      "step": 723
+    },
+    {
+      "epoch": 0.05323920876535039,
+      "grad_norm": 0.18102772533893585,
+      "learning_rate": 0.00019971872462157868,
+      "loss": 1.1107,
+      "step": 724
+    },
+    {
+      "epoch": 0.053312743584087066,
+      "grad_norm": 0.24585627019405365,
+      "learning_rate": 0.00019971693675577734,
+      "loss": 1.1349,
+      "step": 725
+    },
+    {
+      "epoch": 0.053386278402823735,
+      "grad_norm": 0.24396342039108276,
+      "learning_rate": 0.0001997151432339261,
+      "loss": 1.1062,
+      "step": 726
+    },
+    {
+      "epoch": 0.05345981322156041,
+      "grad_norm": 0.15085284411907196,
+      "learning_rate": 0.00019971334405612673,
+      "loss": 1.1406,
+      "step": 727
+    },
+    {
+      "epoch": 0.05353334804029708,
+      "grad_norm": 0.20030665397644043,
+      "learning_rate": 0.00019971153922248127,
+      "loss": 1.1186,
+      "step": 728
+    },
+    {
+      "epoch": 0.053606882859033755,
+      "grad_norm": 0.1220228299498558,
+      "learning_rate": 0.00019970972873309203,
+      "loss": 1.129,
+      "step": 729
+    },
+    {
+      "epoch": 0.053680417677770424,
+      "grad_norm": 0.18272370100021362,
+      "learning_rate": 0.0001997079125880618,
+      "loss": 1.1289,
+      "step": 730
+    },
+    {
+      "epoch": 0.05375395249650709,
+      "grad_norm": 0.15365155041217804,
+      "learning_rate": 0.00019970609078749348,
+      "loss": 1.1445,
+      "step": 731
+    },
+    {
+      "epoch": 0.05382748731524377,
+      "grad_norm": 0.15165071189403534,
+      "learning_rate": 0.0001997042633314905,
+      "loss": 1.1416,
+      "step": 732
+    },
+    {
+      "epoch": 0.05390102213398044,
+      "grad_norm": 0.13192003965377808,
+      "learning_rate": 0.0001997024302201565,
+      "loss": 1.1375,
+      "step": 733
+    },
+    {
+      "epoch": 0.05397455695271711,
+      "grad_norm": 0.14392445981502533,
+      "learning_rate": 0.00019970059145359538,
+      "loss": 1.1515,
+      "step": 734
+    },
+    {
+      "epoch": 0.05404809177145378,
+      "grad_norm": 0.13041673600673676,
+      "learning_rate": 0.00019969874703191156,
+      "loss": 1.1315,
+      "step": 735
+    },
+    {
+      "epoch": 0.05412162659019046,
+      "grad_norm": 0.1448061466217041,
+      "learning_rate": 0.00019969689695520954,
+      "loss": 1.1326,
+      "step": 736
+    },
+    {
+      "epoch": 0.05419516140892713,
+      "grad_norm": 0.13795584440231323,
+      "learning_rate": 0.00019969504122359434,
+      "loss": 1.1921,
+      "step": 737
+    },
+    {
+      "epoch": 0.054268696227663796,
+      "grad_norm": 0.1298113614320755,
+      "learning_rate": 0.0001996931798371712,
+      "loss": 1.1322,
+      "step": 738
+    },
+    {
+      "epoch": 0.05434223104640047,
+      "grad_norm": 0.1608419418334961,
+      "learning_rate": 0.00019969131279604568,
+      "loss": 1.1369,
+      "step": 739
+    },
+    {
+      "epoch": 0.05441576586513714,
+      "grad_norm": 0.20988024771213531,
+      "learning_rate": 0.00019968944010032366,
+      "loss": 1.1285,
+      "step": 740
+    },
+    {
+      "epoch": 0.054489300683873816,
+      "grad_norm": 0.14484381675720215,
+      "learning_rate": 0.0001996875617501114,
+      "loss": 1.1207,
+      "step": 741
+    },
+    {
+      "epoch": 0.054562835502610485,
+      "grad_norm": 0.16293908655643463,
+      "learning_rate": 0.00019968567774551545,
+      "loss": 1.1338,
+      "step": 742
+    },
+    {
+      "epoch": 0.05463637032134716,
+      "grad_norm": 0.1738443374633789,
+      "learning_rate": 0.00019968378808664266,
+      "loss": 1.174,
+      "step": 743
+    },
+    {
+      "epoch": 0.05470990514008383,
+      "grad_norm": 0.15462669730186462,
+      "learning_rate": 0.00019968189277360022,
+      "loss": 1.1609,
+      "step": 744
+    },
+    {
+      "epoch": 0.0547834399588205,
+      "grad_norm": 0.20348049700260162,
+      "learning_rate": 0.0001996799918064956,
+      "loss": 1.1476,
+      "step": 745
+    },
+    {
+      "epoch": 0.054856974777557174,
+      "grad_norm": 0.1813180148601532,
+      "learning_rate": 0.00019967808518543665,
+      "loss": 1.1221,
+      "step": 746
+    },
+    {
+      "epoch": 0.05493050959629384,
+      "grad_norm": 0.14869056642055511,
+      "learning_rate": 0.00019967617291053152,
+      "loss": 1.1592,
+      "step": 747
+    },
+    {
+      "epoch": 0.05500404441503052,
+      "grad_norm": 0.1876605898141861,
+      "learning_rate": 0.00019967425498188866,
+      "loss": 1.1929,
+      "step": 748
+    },
+    {
+      "epoch": 0.05507757923376719,
+      "grad_norm": 0.2406010627746582,
+      "learning_rate": 0.00019967233139961686,
+      "loss": 1.1232,
+      "step": 749
+    },
+    {
+      "epoch": 0.05515111405250386,
+      "grad_norm": 0.1598140001296997,
+      "learning_rate": 0.00019967040216382524,
+      "loss": 1.1377,
+      "step": 750
+    },
+    {
+      "epoch": 0.05522464887124053,
+      "grad_norm": 0.23054197430610657,
+      "learning_rate": 0.00019966846727462325,
+      "loss": 1.1398,
+      "step": 751
+    },
+    {
+      "epoch": 0.0552981836899772,
+      "grad_norm": 0.18477433919906616,
+      "learning_rate": 0.00019966652673212058,
+      "loss": 1.1378,
+      "step": 752
+    },
+    {
+      "epoch": 0.05537171850871388,
+      "grad_norm": 0.15528306365013123,
+      "learning_rate": 0.00019966458053642737,
+      "loss": 1.1025,
+      "step": 753
+    },
+    {
+      "epoch": 0.055445253327450546,
+      "grad_norm": 0.18870909512043,
+      "learning_rate": 0.00019966262868765394,
+      "loss": 1.1454,
+      "step": 754
+    },
+    {
+      "epoch": 0.05551878814618722,
+      "grad_norm": 0.14036564528942108,
+      "learning_rate": 0.00019966067118591102,
+      "loss": 1.1339,
+      "step": 755
+    },
+    {
+      "epoch": 0.05559232296492389,
+      "grad_norm": 0.1574016958475113,
+      "learning_rate": 0.00019965870803130968,
+      "loss": 1.1106,
+      "step": 756
+    },
+    {
+      "epoch": 0.055665857783660566,
+      "grad_norm": 0.1965419352054596,
+      "learning_rate": 0.00019965673922396125,
+      "loss": 1.1673,
+      "step": 757
+    },
+    {
+      "epoch": 0.055739392602397235,
+      "grad_norm": 0.1356707215309143,
+      "learning_rate": 0.0001996547647639774,
+      "loss": 1.1473,
+      "step": 758
+    },
+    {
+      "epoch": 0.055812927421133904,
+      "grad_norm": 0.19418513774871826,
+      "learning_rate": 0.0001996527846514701,
+      "loss": 1.1105,
+      "step": 759
+    },
+    {
+      "epoch": 0.05588646223987058,
+      "grad_norm": 0.18403807282447815,
+      "learning_rate": 0.0001996507988865517,
+      "loss": 1.0825,
+      "step": 760
+    },
+    {
+      "epoch": 0.05595999705860725,
+      "grad_norm": 0.15853463113307953,
+      "learning_rate": 0.0001996488074693348,
+      "loss": 1.1191,
+      "step": 761
+    },
+    {
+      "epoch": 0.056033531877343924,
+      "grad_norm": 0.15332169830799103,
+      "learning_rate": 0.0001996468103999324,
+      "loss": 1.1516,
+      "step": 762
+    },
+    {
+      "epoch": 0.05610706669608059,
+      "grad_norm": 0.17458893358707428,
+      "learning_rate": 0.00019964480767845774,
+      "loss": 1.1355,
+      "step": 763
+    },
+    {
+      "epoch": 0.05618060151481727,
+      "grad_norm": 0.15182726085186005,
+      "learning_rate": 0.00019964279930502444,
+      "loss": 1.1387,
+      "step": 764
+    },
+    {
+      "epoch": 0.05625413633355394,
+      "grad_norm": 0.17585308849811554,
+      "learning_rate": 0.00019964078527974643,
+      "loss": 1.1319,
+      "step": 765
+    },
+    {
+      "epoch": 0.056327671152290607,
+      "grad_norm": 0.14050382375717163,
+      "learning_rate": 0.0001996387656027379,
+      "loss": 1.1623,
+      "step": 766
+    },
+    {
+      "epoch": 0.05640120597102728,
+      "grad_norm": 0.18852655589580536,
+      "learning_rate": 0.0001996367402741134,
+      "loss": 1.1202,
+      "step": 767
+    },
+    {
+      "epoch": 0.05647474078976395,
+      "grad_norm": 0.15703269839286804,
+      "learning_rate": 0.00019963470929398788,
+      "loss": 1.1574,
+      "step": 768
+    },
+    {
+      "epoch": 0.05654827560850063,
+      "grad_norm": 0.1775537133216858,
+      "learning_rate": 0.00019963267266247652,
+      "loss": 1.1257,
+      "step": 769
+    },
+    {
+      "epoch": 0.056621810427237296,
+      "grad_norm": 0.1762571632862091,
+      "learning_rate": 0.00019963063037969477,
+      "loss": 1.168,
+      "step": 770
+    },
+    {
+      "epoch": 0.05669534524597397,
+      "grad_norm": 0.17216037213802338,
+      "learning_rate": 0.00019962858244575851,
+      "loss": 1.1577,
+      "step": 771
+    },
+    {
+      "epoch": 0.05676888006471064,
+      "grad_norm": 0.15470843017101288,
+      "learning_rate": 0.00019962652886078397,
+      "loss": 1.1279,
+      "step": 772
+    },
+    {
+      "epoch": 0.05684241488344731,
+      "grad_norm": 0.1754273921251297,
+      "learning_rate": 0.00019962446962488753,
+      "loss": 1.1559,
+      "step": 773
+    },
+    {
+      "epoch": 0.056915949702183985,
+      "grad_norm": 0.16136792302131653,
+      "learning_rate": 0.00019962240473818608,
+      "loss": 1.1212,
+      "step": 774
+    },
+    {
+      "epoch": 0.056989484520920654,
+      "grad_norm": 0.14752142131328583,
+      "learning_rate": 0.00019962033420079666,
+      "loss": 1.1453,
+      "step": 775
+    },
+    {
+      "epoch": 0.05706301933965733,
+      "grad_norm": 0.16730742156505585,
+      "learning_rate": 0.00019961825801283678,
+      "loss": 1.1454,
+      "step": 776
+    },
+    {
+      "epoch": 0.057136554158394,
+      "grad_norm": 0.1539158970117569,
+      "learning_rate": 0.00019961617617442415,
+      "loss": 1.1197,
+      "step": 777
+    },
+    {
+      "epoch": 0.057210088977130674,
+      "grad_norm": 0.1935870498418808,
+      "learning_rate": 0.00019961408868567686,
+      "loss": 1.1412,
+      "step": 778
+    },
+    {
+      "epoch": 0.05728362379586734,
+      "grad_norm": 0.155315101146698,
+      "learning_rate": 0.00019961199554671338,
+      "loss": 1.1499,
+      "step": 779
+    },
+    {
+      "epoch": 0.05735715861460401,
+      "grad_norm": 0.2634023427963257,
+      "learning_rate": 0.00019960989675765233,
+      "loss": 1.1642,
+      "step": 780
+    },
+    {
+      "epoch": 0.05743069343334069,
+      "grad_norm": 0.16195780038833618,
+      "learning_rate": 0.00019960779231861287,
+      "loss": 1.1376,
+      "step": 781
+    },
+    {
+      "epoch": 0.057504228252077356,
+      "grad_norm": 0.22846443951129913,
+      "learning_rate": 0.0001996056822297143,
+      "loss": 1.1278,
+      "step": 782
+    },
+    {
+      "epoch": 0.05757776307081403,
+      "grad_norm": 0.18559294939041138,
+      "learning_rate": 0.0001996035664910763,
+      "loss": 1.1659,
+      "step": 783
+    },
+    {
+      "epoch": 0.0576512978895507,
+      "grad_norm": 0.21610809862613678,
+      "learning_rate": 0.00019960144510281888,
+      "loss": 1.1509,
+      "step": 784
+    },
+    {
+      "epoch": 0.05772483270828738,
+      "grad_norm": 0.2257750928401947,
+      "learning_rate": 0.0001995993180650624,
+      "loss": 1.0951,
+      "step": 785
+    },
+    {
+      "epoch": 0.057798367527024046,
+      "grad_norm": 0.24561654031276703,
+      "learning_rate": 0.00019959718537792748,
+      "loss": 1.1064,
+      "step": 786
+    },
+    {
+      "epoch": 0.057871902345760715,
+      "grad_norm": 0.16686856746673584,
+      "learning_rate": 0.00019959504704153508,
+      "loss": 1.1271,
+      "step": 787
+    },
+    {
+      "epoch": 0.05794543716449739,
+      "grad_norm": 0.22467289865016937,
+      "learning_rate": 0.00019959290305600653,
+      "loss": 1.1332,
+      "step": 788
+    },
+    {
+      "epoch": 0.05801897198323406,
+      "grad_norm": 0.2947998344898224,
+      "learning_rate": 0.00019959075342146343,
+      "loss": 1.1208,
+      "step": 789
+    },
+    {
+      "epoch": 0.058092506801970735,
+      "grad_norm": 0.3257547914981842,
+      "learning_rate": 0.00019958859813802766,
+      "loss": 1.1261,
+      "step": 790
+    },
+    {
+      "epoch": 0.058166041620707404,
+      "grad_norm": 0.24946001172065735,
+      "learning_rate": 0.0001995864372058215,
+      "loss": 1.0948,
+      "step": 791
+    },
+    {
+      "epoch": 0.05823957643944408,
+      "grad_norm": 0.2296852469444275,
+      "learning_rate": 0.00019958427062496755,
+      "loss": 1.11,
+      "step": 792
+    },
+    {
+      "epoch": 0.05831311125818075,
+      "grad_norm": 0.23633046448230743,
+      "learning_rate": 0.00019958209839558865,
+      "loss": 1.1018,
+      "step": 793
+    },
+    {
+      "epoch": 0.05838664607691742,
+      "grad_norm": 0.21587896347045898,
+      "learning_rate": 0.00019957992051780808,
+      "loss": 1.0973,
+      "step": 794
+    },
+    {
+      "epoch": 0.05846018089565409,
+      "grad_norm": 0.20049883425235748,
+      "learning_rate": 0.00019957773699174928,
+      "loss": 1.126,
+      "step": 795
+    },
+    {
+      "epoch": 0.05853371571439076,
+      "grad_norm": 0.27811989188194275,
+      "learning_rate": 0.00019957554781753617,
+      "loss": 1.1112,
+      "step": 796
+    },
+    {
+      "epoch": 0.05860725053312744,
+      "grad_norm": 0.30196577310562134,
+      "learning_rate": 0.0001995733529952929,
+      "loss": 1.138,
+      "step": 797
+    },
+    {
+      "epoch": 0.058680785351864106,
+      "grad_norm": 0.17211753129959106,
+      "learning_rate": 0.00019957115252514397,
+      "loss": 1.1096,
+      "step": 798
+    },
+    {
+      "epoch": 0.05875432017060078,
+      "grad_norm": 0.24512962996959686,
+      "learning_rate": 0.00019956894640721416,
+      "loss": 1.0995,
+      "step": 799
+    },
+    {
+      "epoch": 0.05882785498933745,
+      "grad_norm": 0.22058527171611786,
+      "learning_rate": 0.00019956673464162865,
+      "loss": 1.109,
+      "step": 800
+    },
+    {
+      "epoch": 0.05890138980807412,
+      "grad_norm": 0.19565853476524353,
+      "learning_rate": 0.00019956451722851287,
+      "loss": 1.1031,
+      "step": 801
+    },
+    {
+      "epoch": 0.058974924626810796,
+      "grad_norm": 0.1726284623146057,
+      "learning_rate": 0.00019956229416799262,
+      "loss": 1.1202,
+      "step": 802
+    },
+    {
+      "epoch": 0.059048459445547465,
+      "grad_norm": 0.16436973214149475,
+      "learning_rate": 0.00019956006546019395,
+      "loss": 1.0853,
+      "step": 803
+    },
+    {
+      "epoch": 0.05912199426428414,
+      "grad_norm": 0.19401706755161285,
+      "learning_rate": 0.00019955783110524327,
+      "loss": 1.0959,
+      "step": 804
+    },
+    {
+      "epoch": 0.05919552908302081,
+      "grad_norm": 0.1678544133901596,
+      "learning_rate": 0.00019955559110326738,
+      "loss": 1.1444,
+      "step": 805
+    },
+    {
+      "epoch": 0.059269063901757485,
+      "grad_norm": 0.16545085608959198,
+      "learning_rate": 0.0001995533454543933,
+      "loss": 1.1146,
+      "step": 806
+    },
+    {
+      "epoch": 0.059342598720494154,
+      "grad_norm": 0.17924515902996063,
+      "learning_rate": 0.00019955109415874839,
+      "loss": 1.1107,
+      "step": 807
+    },
+    {
+      "epoch": 0.05941613353923082,
+      "grad_norm": 0.16909149289131165,
+      "learning_rate": 0.00019954883721646035,
+      "loss": 1.1347,
+      "step": 808
+    },
+    {
+      "epoch": 0.0594896683579675,
+      "grad_norm": 0.21218372881412506,
+      "learning_rate": 0.00019954657462765724,
+      "loss": 1.0965,
+      "step": 809
+    },
+    {
+      "epoch": 0.05956320317670417,
+      "grad_norm": 0.1977909505367279,
+      "learning_rate": 0.00019954430639246735,
+      "loss": 1.0798,
+      "step": 810
+    },
+    {
+      "epoch": 0.05963673799544084,
+      "grad_norm": 0.21542273461818695,
+      "learning_rate": 0.00019954203251101933,
+      "loss": 1.0872,
+      "step": 811
+    },
+    {
+      "epoch": 0.05971027281417751,
+      "grad_norm": 0.21170756220817566,
+      "learning_rate": 0.0001995397529834422,
+      "loss": 1.0915,
+      "step": 812
+    },
+    {
+      "epoch": 0.05978380763291419,
+      "grad_norm": 0.30508801341056824,
+      "learning_rate": 0.0001995374678098652,
+      "loss": 1.1091,
+      "step": 813
+    },
+    {
+      "epoch": 0.059857342451650856,
+      "grad_norm": 0.3488905727863312,
+      "learning_rate": 0.000199535176990418,
+      "loss": 1.0951,
+      "step": 814
+    },
+    {
+      "epoch": 0.059930877270387525,
+      "grad_norm": 0.4338045120239258,
+      "learning_rate": 0.00019953288052523051,
+      "loss": 1.0612,
+      "step": 815
+    },
+    {
+      "epoch": 0.0600044120891242,
+      "grad_norm": 0.39033493399620056,
+      "learning_rate": 0.00019953057841443298,
+      "loss": 1.1085,
+      "step": 816
+    },
+    {
+      "epoch": 0.06007794690786087,
+      "grad_norm": 0.20895475149154663,
+      "learning_rate": 0.00019952827065815605,
+      "loss": 1.0775,
+      "step": 817
+    },
+    {
+      "epoch": 0.060151481726597546,
+      "grad_norm": 0.32304757833480835,
+      "learning_rate": 0.00019952595725653054,
+      "loss": 1.0805,
+      "step": 818
+    },
+    {
+      "epoch": 0.060225016545334215,
+      "grad_norm": 0.4716585576534271,
+      "learning_rate": 0.0001995236382096877,
+      "loss": 1.1157,
+      "step": 819
+    },
+    {
+      "epoch": 0.06029855136407089,
+      "grad_norm": 0.3339405059814453,
+      "learning_rate": 0.00019952131351775907,
+      "loss": 1.1247,
+      "step": 820
+    },
+    {
+      "epoch": 0.06037208618280756,
+      "grad_norm": 0.2096053510904312,
+      "learning_rate": 0.00019951898318087648,
+      "loss": 1.0855,
+      "step": 821
+    },
+    {
+      "epoch": 0.06044562100154423,
+      "grad_norm": 0.2807616889476776,
+      "learning_rate": 0.0001995166471991722,
+      "loss": 1.1212,
+      "step": 822
+    },
+    {
+      "epoch": 0.060519155820280904,
+      "grad_norm": 0.2965157926082611,
+      "learning_rate": 0.00019951430557277862,
+      "loss": 1.1111,
+      "step": 823
+    },
+    {
+      "epoch": 0.06059269063901757,
+      "grad_norm": 0.19426609575748444,
+      "learning_rate": 0.00019951195830182863,
+      "loss": 1.0863,
+      "step": 824
+    },
+    {
+      "epoch": 0.06066622545775425,
+      "grad_norm": 0.32013893127441406,
+      "learning_rate": 0.00019950960538645532,
+      "loss": 1.1091,
+      "step": 825
+    },
+    {
+      "epoch": 0.06073976027649092,
+      "grad_norm": 0.2480231672525406,
+      "learning_rate": 0.0001995072468267922,
+      "loss": 1.0951,
+      "step": 826
+    },
+    {
+      "epoch": 0.06081329509522759,
+      "grad_norm": 0.21880125999450684,
+      "learning_rate": 0.00019950488262297303,
+      "loss": 1.0832,
+      "step": 827
+    },
+    {
+      "epoch": 0.06088682991396426,
+      "grad_norm": 0.2135334610939026,
+      "learning_rate": 0.0001995025127751319,
+      "loss": 1.0856,
+      "step": 828
+    },
+    {
+      "epoch": 0.06096036473270093,
+      "grad_norm": 0.3305091857910156,
+      "learning_rate": 0.00019950013728340323,
+      "loss": 1.0285,
+      "step": 829
+    },
+    {
+      "epoch": 0.061033899551437606,
+      "grad_norm": 0.49997103214263916,
+      "learning_rate": 0.00019949775614792176,
+      "loss": 1.1035,
+      "step": 830
+    },
+    {
+      "epoch": 0.061107434370174275,
+      "grad_norm": 0.512563943862915,
+      "learning_rate": 0.00019949536936882258,
+      "loss": 1.0929,
+      "step": 831
+    },
+    {
+      "epoch": 0.06118096918891095,
+      "grad_norm": 0.28792399168014526,
+      "learning_rate": 0.00019949297694624103,
+      "loss": 1.0967,
+      "step": 832
+    },
+    {
+      "epoch": 0.06125450400764762,
+      "grad_norm": 0.2645791471004486,
+      "learning_rate": 0.00019949057888031283,
+      "loss": 1.0939,
+      "step": 833
+    },
+    {
+      "epoch": 0.061328038826384296,
+      "grad_norm": 0.3675289452075958,
+      "learning_rate": 0.000199488175171174,
+      "loss": 1.0736,
+      "step": 834
+    },
+    {
+      "epoch": 0.061401573645120965,
+      "grad_norm": 0.3173559904098511,
+      "learning_rate": 0.00019948576581896088,
+      "loss": 1.0454,
+      "step": 835
+    },
+    {
+      "epoch": 0.06147510846385763,
+      "grad_norm": 0.27554789185523987,
+      "learning_rate": 0.00019948335082381013,
+      "loss": 1.095,
+      "step": 836
+    },
+    {
+      "epoch": 0.06154864328259431,
+      "grad_norm": 0.221135213971138,
+      "learning_rate": 0.0001994809301858587,
+      "loss": 1.1299,
+      "step": 837
+    },
+    {
+      "epoch": 0.06162217810133098,
+      "grad_norm": 0.2252882421016693,
+      "learning_rate": 0.00019947850390524398,
+      "loss": 1.0843,
+      "step": 838
+    },
+    {
+      "epoch": 0.061695712920067654,
+      "grad_norm": 0.2281426191329956,
+      "learning_rate": 0.00019947607198210352,
+      "loss": 1.0864,
+      "step": 839
+    },
+    {
+      "epoch": 0.06176924773880432,
+      "grad_norm": 0.4098438024520874,
+      "learning_rate": 0.00019947363441657527,
+      "loss": 1.1066,
+      "step": 840
+    },
+    {
+      "epoch": 0.061842782557541,
+      "grad_norm": 0.6180359721183777,
+      "learning_rate": 0.0001994711912087975,
+      "loss": 1.0953,
+      "step": 841
+    },
+    {
+      "epoch": 0.06191631737627767,
+      "grad_norm": 0.7323246002197266,
+      "learning_rate": 0.00019946874235890878,
+      "loss": 1.074,
+      "step": 842
+    },
+    {
+      "epoch": 0.061989852195014336,
+      "grad_norm": 0.6787753701210022,
+      "learning_rate": 0.00019946628786704802,
+      "loss": 1.058,
+      "step": 843
+    },
+    {
+      "epoch": 0.06206338701375101,
+      "grad_norm": 0.4077757000923157,
+      "learning_rate": 0.00019946382773335444,
+      "loss": 1.0541,
+      "step": 844
+    },
+    {
+      "epoch": 0.06213692183248768,
+      "grad_norm": 0.23864954710006714,
+      "learning_rate": 0.0001994613619579676,
+      "loss": 1.0446,
+      "step": 845
+    },
+    {
+      "epoch": 0.062210456651224356,
+      "grad_norm": 0.3627992570400238,
+      "learning_rate": 0.00019945889054102733,
+      "loss": 1.0746,
+      "step": 846
+    },
+    {
+      "epoch": 0.062283991469961025,
+      "grad_norm": 0.5194274187088013,
+      "learning_rate": 0.00019945641348267384,
+      "loss": 1.0876,
+      "step": 847
+    },
+    {
+      "epoch": 0.0623575262886977,
+      "grad_norm": 0.48555782437324524,
+      "learning_rate": 0.0001994539307830476,
+      "loss": 1.1137,
+      "step": 848
+    },
+    {
+      "epoch": 0.06243106110743437,
+      "grad_norm": 0.22151942551136017,
+      "learning_rate": 0.00019945144244228944,
+      "loss": 1.0418,
+      "step": 849
+    },
+    {
+      "epoch": 0.06250459592617104,
+      "grad_norm": 0.6997348666191101,
+      "learning_rate": 0.0001994489484605405,
+      "loss": 1.036,
+      "step": 850
+    },
+    {
+      "epoch": 0.06257813074490771,
+      "grad_norm": 0.8677905797958374,
+      "learning_rate": 0.0001994464488379423,
+      "loss": 1.0929,
+      "step": 851
+    },
+    {
+      "epoch": 0.06265166556364439,
+      "grad_norm": 0.6879971623420715,
+      "learning_rate": 0.00019944394357463655,
+      "loss": 1.0616,
+      "step": 852
+    },
+    {
+      "epoch": 0.06272520038238105,
+      "grad_norm": 0.3544526696205139,
+      "learning_rate": 0.00019944143267076534,
+      "loss": 1.0719,
+      "step": 853
+    },
+    {
+      "epoch": 0.06279873520111773,
+      "grad_norm": 0.37275660037994385,
+      "learning_rate": 0.00019943891612647118,
+      "loss": 1.0396,
+      "step": 854
+    },
+    {
+      "epoch": 0.0628722700198544,
+      "grad_norm": 0.6234080195426941,
+      "learning_rate": 0.0001994363939418967,
+      "loss": 1.0589,
+      "step": 855
+    },
+    {
+      "epoch": 0.06294580483859108,
+      "grad_norm": 0.4823978543281555,
+      "learning_rate": 0.00019943386611718505,
+      "loss": 1.0655,
+      "step": 856
+    },
+    {
+      "epoch": 0.06301933965732774,
+      "grad_norm": 0.32655069231987,
+      "learning_rate": 0.00019943133265247959,
+      "loss": 1.0613,
+      "step": 857
+    },
+    {
+      "epoch": 0.06309287447606442,
+      "grad_norm": 0.6619205474853516,
+      "learning_rate": 0.000199428793547924,
+      "loss": 1.0708,
+      "step": 858
+    },
+    {
+      "epoch": 0.06316640929480109,
+      "grad_norm": 0.761160135269165,
+      "learning_rate": 0.00019942624880366227,
+      "loss": 1.0765,
+      "step": 859
+    },
+    {
+      "epoch": 0.06323994411353775,
+      "grad_norm": 0.670936644077301,
+      "learning_rate": 0.0001994236984198388,
+      "loss": 1.0906,
+      "step": 860
+    },
+    {
+      "epoch": 0.06331347893227443,
+      "grad_norm": 0.3210938274860382,
+      "learning_rate": 0.00019942114239659823,
+      "loss": 1.0958,
+      "step": 861
+    },
+    {
+      "epoch": 0.0633870137510111,
+      "grad_norm": 0.4674816131591797,
+      "learning_rate": 0.00019941858073408556,
+      "loss": 1.0434,
+      "step": 862
+    },
+    {
+      "epoch": 0.06346054856974778,
+      "grad_norm": 0.5286207795143127,
+      "learning_rate": 0.00019941601343244605,
+      "loss": 1.0298,
+      "step": 863
+    },
+    {
+      "epoch": 0.06353408338848444,
+      "grad_norm": 0.39457839727401733,
+      "learning_rate": 0.00019941344049182534,
+      "loss": 1.0867,
+      "step": 864
+    },
+    {
+      "epoch": 0.06360761820722112,
+      "grad_norm": 0.29489806294441223,
+      "learning_rate": 0.0001994108619123694,
+      "loss": 1.0498,
+      "step": 865
+    },
+    {
+      "epoch": 0.0636811530259578,
+      "grad_norm": 0.31229379773139954,
+      "learning_rate": 0.0001994082776942244,
+      "loss": 1.0516,
+      "step": 866
+    },
+    {
+      "epoch": 0.06375468784469446,
+      "grad_norm": 0.3862858712673187,
+      "learning_rate": 0.00019940568783753703,
+      "loss": 1.0458,
+      "step": 867
+    },
+    {
+      "epoch": 0.06382822266343113,
+      "grad_norm": 0.4977075159549713,
+      "learning_rate": 0.00019940309234245415,
+      "loss": 1.0359,
+      "step": 868
+    },
+    {
+      "epoch": 0.06390175748216781,
+      "grad_norm": 0.7233527302742004,
+      "learning_rate": 0.00019940049120912293,
+      "loss": 1.0583,
+      "step": 869
+    },
+    {
+      "epoch": 0.06397529230090448,
+      "grad_norm": 0.7782720923423767,
+      "learning_rate": 0.00019939788443769097,
+      "loss": 1.0497,
+      "step": 870
+    },
+    {
+      "epoch": 0.06404882711964115,
+      "grad_norm": 0.6333433389663696,
+      "learning_rate": 0.00019939527202830605,
+      "loss": 1.0278,
+      "step": 871
+    },
+    {
+      "epoch": 0.06412236193837782,
+      "grad_norm": 0.3762557804584503,
+      "learning_rate": 0.00019939265398111647,
+      "loss": 1.0271,
+      "step": 872
+    },
+    {
+      "epoch": 0.0641958967571145,
+      "grad_norm": 0.2479538917541504,
+      "learning_rate": 0.00019939003029627065,
+      "loss": 1.0385,
+      "step": 873
+    },
+    {
+      "epoch": 0.06426943157585116,
+      "grad_norm": 0.517724335193634,
+      "learning_rate": 0.0001993874009739174,
+      "loss": 1.0338,
+      "step": 874
+    },
+    {
+      "epoch": 0.06434296639458784,
+      "grad_norm": 0.8512571454048157,
+      "learning_rate": 0.0001993847660142059,
+      "loss": 1.0812,
+      "step": 875
+    },
+    {
+      "epoch": 0.06441650121332451,
+      "grad_norm": 1.1282378435134888,
+      "learning_rate": 0.0001993821254172856,
+      "loss": 1.0298,
+      "step": 876
+    },
+    {
+      "epoch": 0.06449003603206119,
+      "grad_norm": 0.8672948479652405,
+      "learning_rate": 0.00019937947918330622,
+      "loss": 1.0305,
+      "step": 877
+    },
+    {
+      "epoch": 0.06456357085079785,
+      "grad_norm": 0.42015644907951355,
+      "learning_rate": 0.00019937682731241792,
+      "loss": 0.9994,
+      "step": 878
+    },
+    {
+      "epoch": 0.06463710566953453,
+      "grad_norm": 0.4227016866207123,
+      "learning_rate": 0.00019937416980477112,
+      "loss": 1.0086,
+      "step": 879
+    },
+    {
+      "epoch": 0.0647106404882712,
+      "grad_norm": 0.9540536403656006,
+      "learning_rate": 0.0001993715066605165,
+      "loss": 1.0546,
+      "step": 880
+    },
+    {
+      "epoch": 0.06478417530700786,
+      "grad_norm": 1.2427648305892944,
+      "learning_rate": 0.00019936883787980518,
+      "loss": 1.0361,
+      "step": 881
+    },
+    {
+      "epoch": 0.06485771012574454,
+      "grad_norm": 0.3062010109424591,
+      "learning_rate": 0.0001993661634627885,
+      "loss": 1.0077,
+      "step": 882
+    },
+    {
+      "epoch": 0.06493124494448121,
+      "grad_norm": 0.8886560201644897,
+      "learning_rate": 0.00019936348340961817,
+      "loss": 1.0373,
+      "step": 883
+    },
+    {
+      "epoch": 0.06500477976321789,
+      "grad_norm": 1.1224278211593628,
+      "learning_rate": 0.0001993607977204462,
+      "loss": 1.0744,
+      "step": 884
+    },
+    {
+      "epoch": 0.06507831458195455,
+      "grad_norm": 0.4002835154533386,
+      "learning_rate": 0.0001993581063954249,
+      "loss": 1.0236,
+      "step": 885
+    },
+    {
+      "epoch": 0.06515184940069123,
+      "grad_norm": 0.7772980332374573,
+      "learning_rate": 0.00019935540943470694,
+      "loss": 1.0557,
+      "step": 886
+    },
+    {
+      "epoch": 0.0652253842194279,
+      "grad_norm": 0.7422829866409302,
+      "learning_rate": 0.00019935270683844533,
+      "loss": 1.0006,
+      "step": 887
+    },
+    {
+      "epoch": 0.06529891903816457,
+      "grad_norm": 0.3872947692871094,
+      "learning_rate": 0.00019934999860679334,
+      "loss": 1.0181,
+      "step": 888
+    },
+    {
+      "epoch": 0.06537245385690124,
+      "grad_norm": 0.924945592880249,
+      "learning_rate": 0.00019934728473990456,
+      "loss": 1.0153,
+      "step": 889
+    },
+    {
+      "epoch": 0.06544598867563792,
+      "grad_norm": 0.7350071668624878,
+      "learning_rate": 0.00019934456523793296,
+      "loss": 1.0069,
+      "step": 890
+    },
+    {
+      "epoch": 0.06551952349437459,
+      "grad_norm": 0.33767738938331604,
+      "learning_rate": 0.00019934184010103278,
+      "loss": 1.0058,
+      "step": 891
+    },
+    {
+      "epoch": 0.06559305831311125,
+      "grad_norm": 0.7754018902778625,
+      "learning_rate": 0.00019933910932935856,
+      "loss": 1.0085,
+      "step": 892
+    },
+    {
+      "epoch": 0.06566659313184793,
+      "grad_norm": 0.8883382678031921,
+      "learning_rate": 0.00019933637292306528,
+      "loss": 0.9888,
+      "step": 893
+    },
+    {
+      "epoch": 0.0657401279505846,
+      "grad_norm": 0.5617631077766418,
+      "learning_rate": 0.00019933363088230803,
+      "loss": 0.9777,
+      "step": 894
+    },
+    {
+      "epoch": 0.06581366276932127,
+      "grad_norm": 0.3998987376689911,
+      "learning_rate": 0.00019933088320724245,
+      "loss": 1.0055,
+      "step": 895
+    },
+    {
+      "epoch": 0.06588719758805794,
+      "grad_norm": 0.7149295806884766,
+      "learning_rate": 0.00019932812989802433,
+      "loss": 0.9925,
+      "step": 896
+    },
+    {
+      "epoch": 0.06596073240679462,
+      "grad_norm": 0.9402654767036438,
+      "learning_rate": 0.00019932537095480985,
+      "loss": 1.0002,
+      "step": 897
+    },
+    {
+      "epoch": 0.0660342672255313,
+      "grad_norm": 0.8194774985313416,
+      "learning_rate": 0.00019932260637775553,
+      "loss": 0.9815,
+      "step": 898
+    },
+    {
+      "epoch": 0.06610780204426796,
+      "grad_norm": 0.31211337447166443,
+      "learning_rate": 0.00019931983616701813,
+      "loss": 0.9866,
+      "step": 899
+    },
+    {
+      "epoch": 0.06618133686300463,
+      "grad_norm": 0.5808043479919434,
+      "learning_rate": 0.0001993170603227548,
+      "loss": 0.9739,
+      "step": 900
+    },
+    {
+      "epoch": 0.06625487168174131,
+      "grad_norm": 0.8424644470214844,
+      "learning_rate": 0.00019931427884512304,
+      "loss": 0.9908,
+      "step": 901
+    },
+    {
+      "epoch": 0.06632840650047797,
+      "grad_norm": 0.8208760023117065,
+      "learning_rate": 0.00019931149173428052,
+      "loss": 0.9953,
+      "step": 902
+    },
+    {
+      "epoch": 0.06640194131921465,
+      "grad_norm": 0.569511890411377,
+      "learning_rate": 0.0001993086989903854,
+      "loss": 0.9844,
+      "step": 903
+    },
+    {
+      "epoch": 0.06647547613795132,
+      "grad_norm": 0.41976502537727356,
+      "learning_rate": 0.0001993059006135961,
+      "loss": 1.0077,
+      "step": 904
+    },
+    {
+      "epoch": 0.066549010956688,
+      "grad_norm": 0.5832225680351257,
+      "learning_rate": 0.0001993030966040713,
+      "loss": 0.9824,
+      "step": 905
+    },
+    {
+      "epoch": 0.06662254577542466,
+      "grad_norm": 0.8282630443572998,
+      "learning_rate": 0.00019930028696197005,
+      "loss": 0.9555,
+      "step": 906
+    },
+    {
+      "epoch": 0.06669608059416134,
+      "grad_norm": 0.9183786511421204,
+      "learning_rate": 0.00019929747168745174,
+      "loss": 0.9736,
+      "step": 907
+    },
+    {
+      "epoch": 0.06676961541289801,
+      "grad_norm": 1.124592900276184,
+      "learning_rate": 0.00019929465078067602,
+      "loss": 0.9764,
+      "step": 908
+    },
+    {
+      "epoch": 0.06684315023163467,
+      "grad_norm": 0.9347695708274841,
+      "learning_rate": 0.00019929182424180292,
+      "loss": 0.9741,
+      "step": 909
+    },
+    {
+      "epoch": 0.06691668505037135,
+      "grad_norm": 0.6277939081192017,
+      "learning_rate": 0.00019928899207099279,
+      "loss": 0.962,
+      "step": 910
+    },
+    {
+      "epoch": 0.06699021986910803,
+      "grad_norm": 0.4454808831214905,
+      "learning_rate": 0.00019928615426840623,
+      "loss": 0.9794,
+      "step": 911
+    },
+    {
+      "epoch": 0.0670637546878447,
+      "grad_norm": 0.4241766631603241,
+      "learning_rate": 0.0001992833108342042,
+      "loss": 0.9676,
+      "step": 912
+    },
+    {
+      "epoch": 0.06713728950658136,
+      "grad_norm": 0.38915762305259705,
+      "learning_rate": 0.00019928046176854804,
+      "loss": 0.9839,
+      "step": 913
+    },
+    {
+      "epoch": 0.06721082432531804,
+      "grad_norm": 0.3738163411617279,
+      "learning_rate": 0.00019927760707159927,
+      "loss": 0.9736,
+      "step": 914
+    },
+    {
+      "epoch": 0.06728435914405471,
+      "grad_norm": 0.4163355231285095,
+      "learning_rate": 0.0001992747467435199,
+      "loss": 0.9438,
+      "step": 915
+    },
+    {
+      "epoch": 0.06735789396279138,
+      "grad_norm": 0.5021100640296936,
+      "learning_rate": 0.0001992718807844721,
+      "loss": 0.9439,
+      "step": 916
+    },
+    {
+      "epoch": 0.06743142878152805,
+      "grad_norm": 0.9198506474494934,
+      "learning_rate": 0.00019926900919461845,
+      "loss": 1.0119,
+      "step": 917
+    },
+    {
+      "epoch": 0.06750496360026473,
+      "grad_norm": 1.8818423748016357,
+      "learning_rate": 0.00019926613197412187,
+      "loss": 0.9911,
+      "step": 918
+    },
+    {
+      "epoch": 0.0675784984190014,
+      "grad_norm": 0.36781951785087585,
+      "learning_rate": 0.0001992632491231455,
+      "loss": 0.9626,
+      "step": 919
+    },
+    {
+      "epoch": 0.06765203323773807,
+      "grad_norm": 1.5763615369796753,
+      "learning_rate": 0.0001992603606418529,
+      "loss": 0.9598,
+      "step": 920
+    },
+    {
+      "epoch": 0.06772556805647474,
+      "grad_norm": 1.079985499382019,
+      "learning_rate": 0.00019925746653040786,
+      "loss": 0.9754,
+      "step": 921
+    },
+    {
+      "epoch": 0.06779910287521142,
+      "grad_norm": 0.4426579177379608,
+      "learning_rate": 0.00019925456678897461,
+      "loss": 0.9592,
+      "step": 922
+    },
+    {
+      "epoch": 0.06787263769394808,
+      "grad_norm": 0.4871717393398285,
+      "learning_rate": 0.00019925166141771753,
+      "loss": 0.9733,
+      "step": 923
+    },
+    {
+      "epoch": 0.06794617251268475,
+      "grad_norm": 0.9069750905036926,
+      "learning_rate": 0.00019924875041680152,
+      "loss": 0.9223,
+      "step": 924
+    },
+    {
+      "epoch": 0.06801970733142143,
+      "grad_norm": 1.5363870859146118,
+      "learning_rate": 0.00019924583378639162,
+      "loss": 0.9731,
+      "step": 925
+    },
+    {
+      "epoch": 0.0680932421501581,
+      "grad_norm": 0.5177852511405945,
+      "learning_rate": 0.00019924291152665329,
+      "loss": 0.9577,
+      "step": 926
+    },
+    {
+      "epoch": 0.06816677696889477,
+      "grad_norm": 0.6165874004364014,
+      "learning_rate": 0.00019923998363775226,
+      "loss": 0.9482,
+      "step": 927
+    },
+    {
+      "epoch": 0.06824031178763144,
+      "grad_norm": 1.4370137453079224,
+      "learning_rate": 0.00019923705011985467,
+      "loss": 0.9511,
+      "step": 928
+    },
+    {
+      "epoch": 0.06831384660636812,
+      "grad_norm": 0.8391589522361755,
+      "learning_rate": 0.00019923411097312686,
+      "loss": 0.9442,
+      "step": 929
+    },
+    {
+      "epoch": 0.06838738142510478,
+      "grad_norm": 0.7481157779693604,
+      "learning_rate": 0.00019923116619773552,
+      "loss": 0.9526,
+      "step": 930
+    },
+    {
+      "epoch": 0.06846091624384146,
+      "grad_norm": 0.9829418659210205,
+      "learning_rate": 0.00019922821579384774,
+      "loss": 0.9509,
+      "step": 931
+    },
+    {
+      "epoch": 0.06853445106257813,
+      "grad_norm": 1.4009909629821777,
+      "learning_rate": 0.00019922525976163083,
+      "loss": 0.9282,
+      "step": 932
+    },
+    {
+      "epoch": 0.06860798588131481,
+      "grad_norm": 0.6831022500991821,
+      "learning_rate": 0.00019922229810125246,
+      "loss": 0.9257,
+      "step": 933
+    },
+    {
+      "epoch": 0.06868152070005147,
+      "grad_norm": 0.49578967690467834,
+      "learning_rate": 0.00019921933081288063,
+      "loss": 0.9227,
+      "step": 934
+    },
+    {
+      "epoch": 0.06875505551878815,
+      "grad_norm": 0.7816241383552551,
+      "learning_rate": 0.00019921635789668367,
+      "loss": 0.9306,
+      "step": 935
+    },
+    {
+      "epoch": 0.06882859033752482,
+      "grad_norm": 1.2638639211654663,
+      "learning_rate": 0.0001992133793528302,
+      "loss": 0.9302,
+      "step": 936
+    },
+    {
+      "epoch": 0.06890212515626148,
+      "grad_norm": 0.938090443611145,
+      "learning_rate": 0.00019921039518148908,
+      "loss": 0.9161,
+      "step": 937
+    },
+    {
+      "epoch": 0.06897565997499816,
+      "grad_norm": 0.7142768502235413,
+      "learning_rate": 0.00019920740538282967,
+      "loss": 0.939,
+      "step": 938
+    },
+    {
+      "epoch": 0.06904919479373484,
+      "grad_norm": 0.9011443257331848,
+      "learning_rate": 0.00019920440995702156,
+      "loss": 0.9436,
+      "step": 939
+    },
+    {
+      "epoch": 0.06912272961247151,
+      "grad_norm": 1.6430141925811768,
+      "learning_rate": 0.0001992014089042346,
+      "loss": 0.9421,
+      "step": 940
+    },
+    {
+      "epoch": 0.06919626443120817,
+      "grad_norm": 0.7322592735290527,
+      "learning_rate": 0.00019919840222463903,
+      "loss": 0.9357,
+      "step": 941
+    },
+    {
+      "epoch": 0.06926979924994485,
+      "grad_norm": 0.5705530047416687,
+      "learning_rate": 0.00019919538991840542,
+      "loss": 0.9461,
+      "step": 942
+    },
+    {
+      "epoch": 0.06934333406868153,
+      "grad_norm": 1.3790953159332275,
+      "learning_rate": 0.0001991923719857046,
+      "loss": 0.9177,
+      "step": 943
+    },
+    {
+      "epoch": 0.06941686888741819,
+      "grad_norm": 1.1959846019744873,
+      "learning_rate": 0.00019918934842670774,
+      "loss": 0.9317,
+      "step": 944
+    },
+    {
+      "epoch": 0.06949040370615486,
+      "grad_norm": 0.8052881956100464,
+      "learning_rate": 0.00019918631924158636,
+      "loss": 0.9323,
+      "step": 945
+    },
+    {
+      "epoch": 0.06956393852489154,
+      "grad_norm": 0.6870926022529602,
+      "learning_rate": 0.00019918328443051229,
+      "loss": 0.9807,
+      "step": 946
+    },
+    {
+      "epoch": 0.06963747334362821,
+      "grad_norm": 0.693600058555603,
+      "learning_rate": 0.00019918024399365766,
+      "loss": 0.9324,
+      "step": 947
+    },
+    {
+      "epoch": 0.06971100816236488,
+      "grad_norm": 0.8876098990440369,
+      "learning_rate": 0.00019917719793119493,
+      "loss": 0.9491,
+      "step": 948
+    },
+    {
+      "epoch": 0.06978454298110155,
+      "grad_norm": 1.3374357223510742,
+      "learning_rate": 0.00019917414624329684,
+      "loss": 0.9053,
+      "step": 949
+    },
+    {
+      "epoch": 0.06985807779983823,
+      "grad_norm": 1.4268823862075806,
+      "learning_rate": 0.00019917108893013655,
+      "loss": 0.934,
+      "step": 950
+    },
+    {
+      "epoch": 0.06993161261857489,
+      "grad_norm": 0.6539322733879089,
+      "learning_rate": 0.00019916802599188738,
+      "loss": 0.9085,
+      "step": 951
+    },
+    {
+      "epoch": 0.07000514743731157,
+      "grad_norm": 0.5493115782737732,
+      "learning_rate": 0.00019916495742872316,
+      "loss": 0.9076,
+      "step": 952
+    },
+    {
+      "epoch": 0.07007868225604824,
+      "grad_norm": 0.46403685212135315,
+      "learning_rate": 0.00019916188324081789,
+      "loss": 0.9381,
+      "step": 953
+    },
+    {
+      "epoch": 0.07015221707478492,
+      "grad_norm": 0.5364179015159607,
+      "learning_rate": 0.00019915880342834597,
+      "loss": 0.9377,
+      "step": 954
+    },
+    {
+      "epoch": 0.07022575189352158,
+      "grad_norm": 0.5531928539276123,
+      "learning_rate": 0.00019915571799148204,
+      "loss": 0.9418,
+      "step": 955
+    },
+    {
+      "epoch": 0.07029928671225825,
+      "grad_norm": 0.5207899808883667,
+      "learning_rate": 0.00019915262693040117,
+      "loss": 0.9357,
+      "step": 956
+    },
+    {
+      "epoch": 0.07037282153099493,
+      "grad_norm": 0.5510200262069702,
+      "learning_rate": 0.00019914953024527865,
+      "loss": 0.9004,
+      "step": 957
+    },
+    {
+      "epoch": 0.07044635634973159,
+      "grad_norm": 1.1437119245529175,
+      "learning_rate": 0.00019914642793629014,
+      "loss": 0.899,
+      "step": 958
+    },
+    {
+      "epoch": 0.07051989116846827,
+      "grad_norm": 2.751384735107422,
+      "learning_rate": 0.0001991433200036116,
+      "loss": 0.9319,
+      "step": 959
+    },
+    {
+      "epoch": 0.07059342598720494,
+      "grad_norm": 0.895267128944397,
+      "learning_rate": 0.00019914020644741932,
+      "loss": 0.9104,
+      "step": 960
+    },
+    {
+      "epoch": 0.07066696080594162,
+      "grad_norm": 3.4362592697143555,
+      "learning_rate": 0.00019913708726788994,
+      "loss": 0.9534,
+      "step": 961
+    },
+    {
+      "epoch": 0.07074049562467828,
+      "grad_norm": 2.9869163036346436,
+      "learning_rate": 0.0001991339624652003,
+      "loss": 0.9059,
+      "step": 962
+    },
+    {
+      "epoch": 0.07081403044341496,
+      "grad_norm": 1.0330302715301514,
+      "learning_rate": 0.0001991308320395277,
+      "loss": 0.9479,
+      "step": 963
+    },
+    {
+      "epoch": 0.07088756526215163,
+      "grad_norm": 1.1759204864501953,
+      "learning_rate": 0.0001991276959910497,
+      "loss": 0.9136,
+      "step": 964
+    },
+    {
+      "epoch": 0.0709611000808883,
+      "grad_norm": 1.5788475275039673,
+      "learning_rate": 0.00019912455431994415,
+      "loss": 0.9554,
+      "step": 965
+    },
+    {
+      "epoch": 0.07103463489962497,
+      "grad_norm": 1.1545523405075073,
+      "learning_rate": 0.00019912140702638928,
+      "loss": 0.9285,
+      "step": 966
+    },
+    {
+      "epoch": 0.07110816971836165,
+      "grad_norm": 1.9769291877746582,
+      "learning_rate": 0.0001991182541105636,
+      "loss": 0.9098,
+      "step": 967
+    },
+    {
+      "epoch": 0.07118170453709832,
+      "grad_norm": 1.5429331064224243,
+      "learning_rate": 0.00019911509557264594,
+      "loss": 0.9078,
+      "step": 968
+    },
+    {
+      "epoch": 0.07125523935583498,
+      "grad_norm": 1.426108956336975,
+      "learning_rate": 0.00019911193141281544,
+      "loss": 0.9268,
+      "step": 969
+    },
+    {
+      "epoch": 0.07132877417457166,
+      "grad_norm": 1.3791168928146362,
+      "learning_rate": 0.00019910876163125162,
+      "loss": 0.9213,
+      "step": 970
+    },
+    {
+      "epoch": 0.07140230899330834,
+      "grad_norm": 1.3759832382202148,
+      "learning_rate": 0.00019910558622813424,
+      "loss": 0.8794,
+      "step": 971
+    },
+    {
+      "epoch": 0.071475843812045,
+      "grad_norm": 1.0466017723083496,
+      "learning_rate": 0.00019910240520364342,
+      "loss": 0.9513,
+      "step": 972
+    },
+    {
+      "epoch": 0.07154937863078167,
+      "grad_norm": 1.4120687246322632,
+      "learning_rate": 0.0001990992185579596,
+      "loss": 0.9346,
+      "step": 973
+    },
+    {
+      "epoch": 0.07162291344951835,
+      "grad_norm": 1.195578932762146,
+      "learning_rate": 0.0001990960262912635,
+      "loss": 0.9195,
+      "step": 974
+    },
+    {
+      "epoch": 0.07169644826825503,
+      "grad_norm": 1.2638912200927734,
+      "learning_rate": 0.0001990928284037362,
+      "loss": 0.9193,
+      "step": 975
+    },
+    {
+      "epoch": 0.07176998308699169,
+      "grad_norm": 1.2135369777679443,
+      "learning_rate": 0.0001990896248955591,
+      "loss": 0.9135,
+      "step": 976
+    },
+    {
+      "epoch": 0.07184351790572836,
+      "grad_norm": 1.2014213800430298,
+      "learning_rate": 0.00019908641576691394,
+      "loss": 0.9008,
+      "step": 977
+    },
+    {
+      "epoch": 0.07191705272446504,
+      "grad_norm": 0.9698330163955688,
+      "learning_rate": 0.0001990832010179827,
+      "loss": 0.8835,
+      "step": 978
+    },
+    {
+      "epoch": 0.0719905875432017,
+      "grad_norm": 1.0424320697784424,
+      "learning_rate": 0.00019907998064894772,
+      "loss": 0.9616,
+      "step": 979
+    },
+    {
+      "epoch": 0.07206412236193838,
+      "grad_norm": 0.6950603127479553,
+      "learning_rate": 0.00019907675465999166,
+      "loss": 0.9308,
+      "step": 980
+    },
+    {
+      "epoch": 0.07213765718067505,
+      "grad_norm": 1.0451945066452026,
+      "learning_rate": 0.0001990735230512975,
+      "loss": 0.9109,
+      "step": 981
+    },
+    {
+      "epoch": 0.07221119199941173,
+      "grad_norm": 0.7574619650840759,
+      "learning_rate": 0.00019907028582304858,
+      "loss": 0.9292,
+      "step": 982
+    },
+    {
+      "epoch": 0.07228472681814839,
+      "grad_norm": 0.747891902923584,
+      "learning_rate": 0.00019906704297542848,
+      "loss": 0.8655,
+      "step": 983
+    },
+    {
+      "epoch": 0.07235826163688507,
+      "grad_norm": 0.9905170798301697,
+      "learning_rate": 0.0001990637945086212,
+      "loss": 0.8715,
+      "step": 984
+    },
+    {
+      "epoch": 0.07243179645562174,
+      "grad_norm": 0.6438395977020264,
+      "learning_rate": 0.0001990605404228109,
+      "loss": 0.8932,
+      "step": 985
+    },
+    {
+      "epoch": 0.0725053312743584,
+      "grad_norm": 0.6783116459846497,
+      "learning_rate": 0.00019905728071818222,
+      "loss": 0.895,
+      "step": 986
+    },
+    {
+      "epoch": 0.07257886609309508,
+      "grad_norm": 0.7403161525726318,
+      "learning_rate": 0.00019905401539492003,
+      "loss": 0.9025,
+      "step": 987
+    },
+    {
+      "epoch": 0.07265240091183175,
+      "grad_norm": 0.8453176617622375,
+      "learning_rate": 0.00019905074445320957,
+      "loss": 0.913,
+      "step": 988
+    },
+    {
+      "epoch": 0.07272593573056843,
+      "grad_norm": 0.42744725942611694,
+      "learning_rate": 0.00019904746789323636,
+      "loss": 0.8906,
+      "step": 989
+    },
+    {
+      "epoch": 0.07279947054930509,
+      "grad_norm": 0.5625191330909729,
+      "learning_rate": 0.0001990441857151862,
+      "loss": 0.8853,
+      "step": 990
+    },
+    {
+      "epoch": 0.07287300536804177,
+      "grad_norm": 0.5175917148590088,
+      "learning_rate": 0.00019904089791924535,
+      "loss": 0.8733,
+      "step": 991
+    },
+    {
+      "epoch": 0.07294654018677844,
+      "grad_norm": 0.5560296773910522,
+      "learning_rate": 0.0001990376045056002,
+      "loss": 0.8642,
+      "step": 992
+    },
+    {
+      "epoch": 0.0730200750055151,
+      "grad_norm": 0.5291785597801208,
+      "learning_rate": 0.00019903430547443763,
+      "loss": 0.8944,
+      "step": 993
+    },
+    {
+      "epoch": 0.07309360982425178,
+      "grad_norm": 0.5972850918769836,
+      "learning_rate": 0.00019903100082594474,
+      "loss": 0.8772,
+      "step": 994
+    },
+    {
+      "epoch": 0.07316714464298846,
+      "grad_norm": 0.6258870363235474,
+      "learning_rate": 0.00019902769056030897,
+      "loss": 0.8909,
+      "step": 995
+    },
+    {
+      "epoch": 0.07324067946172513,
+      "grad_norm": 0.5990886688232422,
+      "learning_rate": 0.00019902437467771812,
+      "loss": 0.8981,
+      "step": 996
+    },
+    {
+      "epoch": 0.0733142142804618,
+      "grad_norm": 0.36634373664855957,
+      "learning_rate": 0.00019902105317836017,
+      "loss": 0.8732,
+      "step": 997
+    },
+    {
+      "epoch": 0.07338774909919847,
+      "grad_norm": 0.5527446269989014,
+      "learning_rate": 0.00019901772606242363,
+      "loss": 0.8666,
+      "step": 998
+    },
+    {
+      "epoch": 0.07346128391793515,
+      "grad_norm": 0.9354996681213379,
+      "learning_rate": 0.00019901439333009713,
+      "loss": 0.8965,
+      "step": 999
+    },
+    {
+      "epoch": 0.07353481873667181,
+      "grad_norm": 1.2674956321716309,
+      "learning_rate": 0.00019901105498156975,
+      "loss": 0.8872,
+      "step": 1000
+    },
+    {
+      "epoch": 0.07360835355540848,
+      "grad_norm": 0.7568721771240234,
+      "learning_rate": 0.00019900771101703086,
+      "loss": 0.8755,
+      "step": 1001
+    },
+    {
+      "epoch": 0.07368188837414516,
+      "grad_norm": 0.6215139627456665,
+      "learning_rate": 0.0001990043614366701,
+      "loss": 0.8863,
+      "step": 1002
+    },
+    {
+      "epoch": 0.07375542319288184,
+      "grad_norm": 0.9236053228378296,
+      "learning_rate": 0.0001990010062406775,
+      "loss": 0.8804,
+      "step": 1003
+    },
+    {
+      "epoch": 0.0738289580116185,
+      "grad_norm": 1.260148525238037,
+      "learning_rate": 0.00019899764542924333,
+      "loss": 0.8481,
+      "step": 1004
+    },
+    {
+      "epoch": 0.07390249283035517,
+      "grad_norm": 0.9443658590316772,
+      "learning_rate": 0.00019899427900255825,
+      "loss": 0.8661,
+      "step": 1005
+    },
+    {
+      "epoch": 0.07397602764909185,
+      "grad_norm": 1.1358203887939453,
+      "learning_rate": 0.00019899090696081316,
+      "loss": 0.8952,
+      "step": 1006
+    },
+    {
+      "epoch": 0.07404956246782851,
+      "grad_norm": 1.3354209661483765,
+      "learning_rate": 0.0001989875293041994,
+      "loss": 0.9036,
+      "step": 1007
+    },
+    {
+      "epoch": 0.07412309728656519,
+      "grad_norm": 0.7364773154258728,
+      "learning_rate": 0.00019898414603290848,
+      "loss": 0.874,
+      "step": 1008
+    },
+    {
+      "epoch": 0.07419663210530186,
+      "grad_norm": 0.43966877460479736,
+      "learning_rate": 0.00019898075714713234,
+      "loss": 0.8966,
+      "step": 1009
+    },
+    {
+      "epoch": 0.07427016692403854,
+      "grad_norm": 0.8051905035972595,
+      "learning_rate": 0.00019897736264706318,
+      "loss": 0.8568,
+      "step": 1010
+    },
+    {
+      "epoch": 0.0743437017427752,
+      "grad_norm": 1.77899968624115,
+      "learning_rate": 0.0001989739625328936,
+      "loss": 0.9052,
+      "step": 1011
+    },
+    {
+      "epoch": 0.07441723656151188,
+      "grad_norm": 0.4734535813331604,
+      "learning_rate": 0.00019897055680481637,
+      "loss": 0.9167,
+      "step": 1012
+    },
+    {
+      "epoch": 0.07449077138024855,
+      "grad_norm": 1.5958480834960938,
+      "learning_rate": 0.00019896714546302472,
+      "loss": 0.9232,
+      "step": 1013
+    },
+    {
+      "epoch": 0.07456430619898521,
+      "grad_norm": 1.2346314191818237,
+      "learning_rate": 0.00019896372850771216,
+      "loss": 0.8713,
+      "step": 1014
+    },
+    {
+      "epoch": 0.07463784101772189,
+      "grad_norm": 0.7043796181678772,
+      "learning_rate": 0.00019896030593907248,
+      "loss": 0.8842,
+      "step": 1015
+    },
+    {
+      "epoch": 0.07471137583645857,
+      "grad_norm": 0.48079049587249756,
+      "learning_rate": 0.00019895687775729978,
+      "loss": 0.9016,
+      "step": 1016
+    },
+    {
+      "epoch": 0.07478491065519524,
+      "grad_norm": 0.424186646938324,
+      "learning_rate": 0.00019895344396258854,
+      "loss": 0.8617,
+      "step": 1017
+    },
+    {
+      "epoch": 0.0748584454739319,
+      "grad_norm": 0.5821173191070557,
+      "learning_rate": 0.00019895000455513354,
+      "loss": 0.8555,
+      "step": 1018
+    },
+    {
+      "epoch": 0.07493198029266858,
+      "grad_norm": 1.2434591054916382,
+      "learning_rate": 0.00019894655953512987,
+      "loss": 0.8907,
+      "step": 1019
+    },
+    {
+      "epoch": 0.07500551511140525,
+      "grad_norm": 1.5707062482833862,
+      "learning_rate": 0.0001989431089027729,
+      "loss": 0.8953,
+      "step": 1020
+    },
+    {
+      "epoch": 0.07507904993014192,
+      "grad_norm": 0.5631011724472046,
+      "learning_rate": 0.00019893965265825838,
+      "loss": 0.995,
+      "step": 1021
+    },
+    {
+      "epoch": 0.07515258474887859,
+      "grad_norm": 0.4945397675037384,
+      "learning_rate": 0.00019893619080178234,
+      "loss": 1.0223,
+      "step": 1022
+    },
+    {
+      "epoch": 0.07522611956761527,
+      "grad_norm": 0.8126094937324524,
+      "learning_rate": 0.00019893272333354117,
+      "loss": 1.0409,
+      "step": 1023
+    },
+    {
+      "epoch": 0.07529965438635194,
+      "grad_norm": 1.7984756231307983,
+      "learning_rate": 0.0001989292502537315,
+      "loss": 1.014,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0753731892050886,
+      "grad_norm": 0.5944324135780334,
+      "learning_rate": 0.00019892577156255037,
+      "loss": 1.0104,
+      "step": 1025
+    },
+    {
+      "epoch": 0.07544672402382528,
+      "grad_norm": 0.9801954627037048,
+      "learning_rate": 0.00019892228726019507,
+      "loss": 1.0217,
+      "step": 1026
+    },
+    {
+      "epoch": 0.07552025884256196,
+      "grad_norm": 2.2698278427124023,
+      "learning_rate": 0.00019891879734686323,
+      "loss": 1.0459,
+      "step": 1027
+    },
+    {
+      "epoch": 0.07559379366129862,
+      "grad_norm": 1.2343313694000244,
+      "learning_rate": 0.00019891530182275283,
+      "loss": 0.9646,
+      "step": 1028
+    },
+    {
+      "epoch": 0.0756673284800353,
+      "grad_norm": 3.1793839931488037,
+      "learning_rate": 0.0001989118006880621,
+      "loss": 1.0355,
+      "step": 1029
+    },
+    {
+      "epoch": 0.07574086329877197,
+      "grad_norm": 3.057964324951172,
+      "learning_rate": 0.00019890829394298963,
+      "loss": 1.0224,
+      "step": 1030
+    },
+    {
+      "epoch": 0.07581439811750865,
+      "grad_norm": 0.8960400223731995,
+      "learning_rate": 0.0001989047815877344,
+      "loss": 1.0261,
+      "step": 1031
+    },
+    {
+      "epoch": 0.07588793293624531,
+      "grad_norm": 1.9378658533096313,
+      "learning_rate": 0.00019890126362249558,
+      "loss": 1.0349,
+      "step": 1032
+    },
+    {
+      "epoch": 0.07596146775498198,
+      "grad_norm": 1.266960620880127,
+      "learning_rate": 0.00019889774004747265,
+      "loss": 0.9963,
+      "step": 1033
+    },
+    {
+      "epoch": 0.07603500257371866,
+      "grad_norm": 1.3049352169036865,
+      "learning_rate": 0.00019889421086286558,
+      "loss": 1.0322,
+      "step": 1034
+    },
+    {
+      "epoch": 0.07610853739245532,
+      "grad_norm": 0.9681654572486877,
+      "learning_rate": 0.0001988906760688745,
+      "loss": 1.0113,
+      "step": 1035
+    },
+    {
+      "epoch": 0.076182072211192,
+      "grad_norm": 1.0815213918685913,
+      "learning_rate": 0.0001988871356656999,
+      "loss": 0.9982,
+      "step": 1036
+    },
+    {
+      "epoch": 0.07625560702992867,
+      "grad_norm": 0.8536231517791748,
+      "learning_rate": 0.00019888358965354259,
+      "loss": 0.9593,
+      "step": 1037
+    },
+    {
+      "epoch": 0.07632914184866535,
+      "grad_norm": 0.9906190633773804,
+      "learning_rate": 0.00019888003803260375,
+      "loss": 0.9714,
+      "step": 1038
+    },
+    {
+      "epoch": 0.07640267666740201,
+      "grad_norm": 0.7149401903152466,
+      "learning_rate": 0.00019887648080308478,
+      "loss": 0.9734,
+      "step": 1039
+    },
+    {
+      "epoch": 0.07647621148613869,
+      "grad_norm": 0.8319122791290283,
+      "learning_rate": 0.0001988729179651875,
+      "loss": 1.0172,
+      "step": 1040
+    },
+    {
+      "epoch": 0.07654974630487536,
+      "grad_norm": 0.6023457050323486,
+      "learning_rate": 0.00019886934951911394,
+      "loss": 1.059,
+      "step": 1041
+    },
+    {
+      "epoch": 0.07662328112361202,
+      "grad_norm": 0.7812548279762268,
+      "learning_rate": 0.00019886577546506654,
+      "loss": 0.9793,
+      "step": 1042
+    },
+    {
+      "epoch": 0.0766968159423487,
+      "grad_norm": 0.87541663646698,
+      "learning_rate": 0.000198862195803248,
+      "loss": 0.9759,
+      "step": 1043
+    },
+    {
+      "epoch": 0.07677035076108538,
+      "grad_norm": 0.6080463528633118,
+      "learning_rate": 0.00019885861053386142,
+      "loss": 0.9891,
+      "step": 1044
+    },
+    {
+      "epoch": 0.07684388557982205,
+      "grad_norm": 0.5907179713249207,
+      "learning_rate": 0.0001988550196571101,
+      "loss": 0.9499,
+      "step": 1045
+    },
+    {
+      "epoch": 0.07691742039855871,
+      "grad_norm": 0.5525895357131958,
+      "learning_rate": 0.00019885142317319773,
+      "loss": 0.9739,
+      "step": 1046
+    },
+    {
+      "epoch": 0.07699095521729539,
+      "grad_norm": 0.5882356762886047,
+      "learning_rate": 0.0001988478210823283,
+      "loss": 0.9922,
+      "step": 1047
+    },
+    {
+      "epoch": 0.07706449003603207,
+      "grad_norm": 0.44530415534973145,
+      "learning_rate": 0.0001988442133847062,
+      "loss": 1.0132,
+      "step": 1048
+    },
+    {
+      "epoch": 0.07713802485476873,
+      "grad_norm": 0.5377620458602905,
+      "learning_rate": 0.00019884060008053595,
+      "loss": 0.9297,
+      "step": 1049
+    },
+    {
+      "epoch": 0.0772115596735054,
+      "grad_norm": 0.5358130931854248,
+      "learning_rate": 0.0001988369811700226,
+      "loss": 1.0008,
+      "step": 1050
+    },
+    {
+      "epoch": 0.07728509449224208,
+      "grad_norm": 0.56438809633255,
+      "learning_rate": 0.00019883335665337132,
+      "loss": 1.0029,
+      "step": 1051
+    },
+    {
+      "epoch": 0.07735862931097875,
+      "grad_norm": 0.7399015426635742,
+      "learning_rate": 0.00019882972653078776,
+      "loss": 0.9544,
+      "step": 1052
+    },
+    {
+      "epoch": 0.07743216412971542,
+      "grad_norm": 1.6130253076553345,
+      "learning_rate": 0.00019882609080247782,
+      "loss": 0.9863,
+      "step": 1053
+    },
+    {
+      "epoch": 0.07750569894845209,
+      "grad_norm": 0.8740609884262085,
+      "learning_rate": 0.00019882244946864774,
+      "loss": 0.9709,
+      "step": 1054
+    },
+    {
+      "epoch": 0.07757923376718877,
+      "grad_norm": 0.5546737909317017,
+      "learning_rate": 0.000198818802529504,
+      "loss": 0.9639,
+      "step": 1055
+    },
+    {
+      "epoch": 0.07765276858592543,
+      "grad_norm": 0.5021244287490845,
+      "learning_rate": 0.0001988151499852535,
+      "loss": 0.9767,
+      "step": 1056
+    },
+    {
+      "epoch": 0.0777263034046621,
+      "grad_norm": 0.6235771179199219,
+      "learning_rate": 0.0001988114918361034,
+      "loss": 0.955,
+      "step": 1057
+    },
+    {
+      "epoch": 0.07779983822339878,
+      "grad_norm": 1.243605375289917,
+      "learning_rate": 0.00019880782808226122,
+      "loss": 1.0084,
+      "step": 1058
+    },
+    {
+      "epoch": 0.07787337304213546,
+      "grad_norm": 1.423438310623169,
+      "learning_rate": 0.00019880415872393473,
+      "loss": 0.9749,
+      "step": 1059
+    },
+    {
+      "epoch": 0.07794690786087212,
+      "grad_norm": 0.41725313663482666,
+      "learning_rate": 0.00019880048376133213,
+      "loss": 0.9476,
+      "step": 1060
+    },
+    {
+      "epoch": 0.0780204426796088,
+      "grad_norm": 0.8969585299491882,
+      "learning_rate": 0.0001987968031946618,
+      "loss": 0.9466,
+      "step": 1061
+    },
+    {
+      "epoch": 0.07809397749834547,
+      "grad_norm": 1.841230869293213,
+      "learning_rate": 0.00019879311702413252,
+      "loss": 1.0198,
+      "step": 1062
+    },
+    {
+      "epoch": 0.07816751231708213,
+      "grad_norm": 0.502857506275177,
+      "learning_rate": 0.0001987894252499534,
+      "loss": 0.952,
+      "step": 1063
+    },
+    {
+      "epoch": 0.07824104713581881,
+      "grad_norm": 1.5628248453140259,
+      "learning_rate": 0.0001987857278723338,
+      "loss": 0.938,
+      "step": 1064
+    },
+    {
+      "epoch": 0.07831458195455548,
+      "grad_norm": 1.0621662139892578,
+      "learning_rate": 0.00019878202489148348,
+      "loss": 0.9868,
+      "step": 1065
+    },
+    {
+      "epoch": 0.07838811677329216,
+      "grad_norm": 0.6004918217658997,
+      "learning_rate": 0.00019877831630761247,
+      "loss": 0.9811,
+      "step": 1066
+    },
+    {
+      "epoch": 0.07846165159202882,
+      "grad_norm": 0.8624292612075806,
+      "learning_rate": 0.00019877460212093108,
+      "loss": 0.9953,
+      "step": 1067
+    },
+    {
+      "epoch": 0.0785351864107655,
+      "grad_norm": 1.4515718221664429,
+      "learning_rate": 0.00019877088233165004,
+      "loss": 0.9622,
+      "step": 1068
+    },
+    {
+      "epoch": 0.07860872122950217,
+      "grad_norm": 0.8499951362609863,
+      "learning_rate": 0.0001987671569399803,
+      "loss": 0.9189,
+      "step": 1069
+    },
+    {
+      "epoch": 0.07868225604823884,
+      "grad_norm": 0.5747846364974976,
+      "learning_rate": 0.0001987634259461332,
+      "loss": 0.9591,
+      "step": 1070
+    },
+    {
+      "epoch": 0.07875579086697551,
+      "grad_norm": 0.5171135663986206,
+      "learning_rate": 0.00019875968935032036,
+      "loss": 0.9604,
+      "step": 1071
+    },
+    {
+      "epoch": 0.07882932568571219,
+      "grad_norm": 0.7115563154220581,
+      "learning_rate": 0.0001987559471527537,
+      "loss": 0.9128,
+      "step": 1072
+    },
+    {
+      "epoch": 0.07890286050444886,
+      "grad_norm": 0.8545862436294556,
+      "learning_rate": 0.0001987521993536455,
+      "loss": 0.9649,
+      "step": 1073
+    },
+    {
+      "epoch": 0.07897639532318552,
+      "grad_norm": 1.2763168811798096,
+      "learning_rate": 0.00019874844595320834,
+      "loss": 0.9329,
+      "step": 1074
+    },
+    {
+      "epoch": 0.0790499301419222,
+      "grad_norm": 0.9935707449913025,
+      "learning_rate": 0.00019874468695165509,
+      "loss": 0.9828,
+      "step": 1075
+    },
+    {
+      "epoch": 0.07912346496065888,
+      "grad_norm": 1.1774729490280151,
+      "learning_rate": 0.000198740922349199,
+      "loss": 0.9484,
+      "step": 1076
+    },
+    {
+      "epoch": 0.07919699977939554,
+      "grad_norm": 1.3150707483291626,
+      "learning_rate": 0.0001987371521460536,
+      "loss": 0.9498,
+      "step": 1077
+    },
+    {
+      "epoch": 0.07927053459813221,
+      "grad_norm": 0.813391387462616,
+      "learning_rate": 0.00019873337634243273,
+      "loss": 0.9288,
+      "step": 1078
+    },
+    {
+      "epoch": 0.07934406941686889,
+      "grad_norm": 0.5367438793182373,
+      "learning_rate": 0.00019872959493855055,
+      "loss": 0.9261,
+      "step": 1079
+    },
+    {
+      "epoch": 0.07941760423560557,
+      "grad_norm": 0.5286107659339905,
+      "learning_rate": 0.00019872580793462151,
+      "loss": 0.9364,
+      "step": 1080
+    },
+    {
+      "epoch": 0.07949113905434223,
+      "grad_norm": 0.6087942719459534,
+      "learning_rate": 0.00019872201533086052,
+      "loss": 0.936,
+      "step": 1081
+    },
+    {
+      "epoch": 0.0795646738730789,
+      "grad_norm": 1.0201727151870728,
+      "learning_rate": 0.00019871821712748262,
+      "loss": 0.9323,
+      "step": 1082
+    },
+    {
+      "epoch": 0.07963820869181558,
+      "grad_norm": 1.3712413311004639,
+      "learning_rate": 0.00019871441332470325,
+      "loss": 0.9354,
+      "step": 1083
+    },
+    {
+      "epoch": 0.07971174351055224,
+      "grad_norm": 0.6275050640106201,
+      "learning_rate": 0.00019871060392273818,
+      "loss": 0.9688,
+      "step": 1084
+    },
+    {
+      "epoch": 0.07978527832928892,
+      "grad_norm": 0.5905522704124451,
+      "learning_rate": 0.00019870678892180345,
+      "loss": 0.9347,
+      "step": 1085
+    },
+    {
+      "epoch": 0.07985881314802559,
+      "grad_norm": 0.5200135707855225,
+      "learning_rate": 0.00019870296832211553,
+      "loss": 0.9266,
+      "step": 1086
+    },
+    {
+      "epoch": 0.07993234796676227,
+      "grad_norm": 0.4588816463947296,
+      "learning_rate": 0.00019869914212389107,
+      "loss": 0.9419,
+      "step": 1087
+    },
+    {
+      "epoch": 0.08000588278549893,
+      "grad_norm": 0.5031083822250366,
+      "learning_rate": 0.0001986953103273471,
+      "loss": 0.9282,
+      "step": 1088
+    },
+    {
+      "epoch": 0.0800794176042356,
+      "grad_norm": 0.5862363576889038,
+      "learning_rate": 0.00019869147293270095,
+      "loss": 0.9336,
+      "step": 1089
+    },
+    {
+      "epoch": 0.08015295242297228,
+      "grad_norm": 1.0542545318603516,
+      "learning_rate": 0.00019868762994017036,
+      "loss": 0.9446,
+      "step": 1090
+    },
+    {
+      "epoch": 0.08022648724170894,
+      "grad_norm": 2.0128653049468994,
+      "learning_rate": 0.0001986837813499732,
+      "loss": 0.9364,
+      "step": 1091
+    },
+    {
+      "epoch": 0.08030002206044562,
+      "grad_norm": 0.7179802060127258,
+      "learning_rate": 0.00019867992716232784,
+      "loss": 0.9057,
+      "step": 1092
+    },
+    {
+      "epoch": 0.0803735568791823,
+      "grad_norm": 2.4752197265625,
+      "learning_rate": 0.00019867606737745285,
+      "loss": 0.9363,
+      "step": 1093
+    },
+    {
+      "epoch": 0.08044709169791897,
+      "grad_norm": 1.0299878120422363,
+      "learning_rate": 0.00019867220199556722,
+      "loss": 0.9185,
+      "step": 1094
+    },
+    {
+      "epoch": 0.08052062651665563,
+      "grad_norm": 2.829221487045288,
+      "learning_rate": 0.00019866833101689013,
+      "loss": 0.9388,
+      "step": 1095
+    },
+    {
+      "epoch": 0.08059416133539231,
+      "grad_norm": 1.9170441627502441,
+      "learning_rate": 0.00019866445444164117,
+      "loss": 0.9043,
+      "step": 1096
+    },
+    {
+      "epoch": 0.08066769615412898,
+      "grad_norm": 2.4133918285369873,
+      "learning_rate": 0.00019866057227004025,
+      "loss": 0.9403,
+      "step": 1097
+    },
+    {
+      "epoch": 0.08074123097286565,
+      "grad_norm": 1.765417456626892,
+      "learning_rate": 0.00019865668450230755,
+      "loss": 0.9405,
+      "step": 1098
+    },
+    {
+      "epoch": 0.08081476579160232,
+      "grad_norm": 2.411672353744507,
+      "learning_rate": 0.00019865279113866357,
+      "loss": 0.9722,
+      "step": 1099
+    },
+    {
+      "epoch": 0.080888300610339,
+      "grad_norm": 2.04662823677063,
+      "learning_rate": 0.00019864889217932917,
+      "loss": 0.9241,
+      "step": 1100
+    },
+    {
+      "epoch": 0.08096183542907567,
+      "grad_norm": 1.2599924802780151,
+      "learning_rate": 0.0001986449876245255,
+      "loss": 0.954,
+      "step": 1101
+    },
+    {
+      "epoch": 0.08103537024781234,
+      "grad_norm": 1.1086915731430054,
+      "learning_rate": 0.00019864107747447402,
+      "loss": 0.9435,
+      "step": 1102
+    },
+    {
+      "epoch": 0.08110890506654901,
+      "grad_norm": 1.7472800016403198,
+      "learning_rate": 0.00019863716172939655,
+      "loss": 0.9401,
+      "step": 1103
+    },
+    {
+      "epoch": 0.08118243988528569,
+      "grad_norm": 1.2390046119689941,
+      "learning_rate": 0.00019863324038951516,
+      "loss": 0.9333,
+      "step": 1104
+    },
+    {
+      "epoch": 0.08125597470402235,
+      "grad_norm": 2.219412326812744,
+      "learning_rate": 0.00019862931345505228,
+      "loss": 0.9892,
+      "step": 1105
+    },
+    {
+      "epoch": 0.08132950952275902,
+      "grad_norm": 1.7195812463760376,
+      "learning_rate": 0.00019862538092623065,
+      "loss": 0.9582,
+      "step": 1106
+    },
+    {
+      "epoch": 0.0814030443414957,
+      "grad_norm": 1.6628973484039307,
+      "learning_rate": 0.00019862144280327333,
+      "loss": 0.9593,
+      "step": 1107
+    },
+    {
+      "epoch": 0.08147657916023238,
+      "grad_norm": 1.594619870185852,
+      "learning_rate": 0.0001986174990864037,
+      "loss": 0.9168,
+      "step": 1108
+    },
+    {
+      "epoch": 0.08155011397896904,
+      "grad_norm": 1.6136348247528076,
+      "learning_rate": 0.00019861354977584542,
+      "loss": 0.9264,
+      "step": 1109
+    },
+    {
+      "epoch": 0.08162364879770571,
+      "grad_norm": 1.3548375368118286,
+      "learning_rate": 0.00019860959487182255,
+      "loss": 0.9258,
+      "step": 1110
+    },
+    {
+      "epoch": 0.08169718361644239,
+      "grad_norm": 1.4510983228683472,
+      "learning_rate": 0.0001986056343745594,
+      "loss": 0.935,
+      "step": 1111
+    },
+    {
+      "epoch": 0.08177071843517905,
+      "grad_norm": 1.1460691690444946,
+      "learning_rate": 0.0001986016682842806,
+      "loss": 0.8939,
+      "step": 1112
+    },
+    {
+      "epoch": 0.08184425325391573,
+      "grad_norm": 1.6069097518920898,
+      "learning_rate": 0.00019859769660121113,
+      "loss": 0.9461,
+      "step": 1113
+    },
+    {
+      "epoch": 0.0819177880726524,
+      "grad_norm": 1.496475338935852,
+      "learning_rate": 0.0001985937193255762,
+      "loss": 0.9517,
+      "step": 1114
+    },
+    {
+      "epoch": 0.08199132289138908,
+      "grad_norm": 1.3175703287124634,
+      "learning_rate": 0.00019858973645760152,
+      "loss": 0.9194,
+      "step": 1115
+    },
+    {
+      "epoch": 0.08206485771012574,
+      "grad_norm": 1.2758874893188477,
+      "learning_rate": 0.00019858574799751292,
+      "loss": 0.9259,
+      "step": 1116
+    },
+    {
+      "epoch": 0.08213839252886242,
+      "grad_norm": 1.117479920387268,
+      "learning_rate": 0.00019858175394553665,
+      "loss": 0.9571,
+      "step": 1117
+    },
+    {
+      "epoch": 0.08221192734759909,
+      "grad_norm": 0.9529320001602173,
+      "learning_rate": 0.00019857775430189925,
+      "loss": 0.9198,
+      "step": 1118
+    },
+    {
+      "epoch": 0.08228546216633575,
+      "grad_norm": 1.497534990310669,
+      "learning_rate": 0.00019857374906682762,
+      "loss": 0.9294,
+      "step": 1119
+    },
+    {
+      "epoch": 0.08235899698507243,
+      "grad_norm": 0.7723939418792725,
+      "learning_rate": 0.00019856973824054888,
+      "loss": 0.9553,
+      "step": 1120
+    },
+    {
+      "epoch": 0.0824325318038091,
+      "grad_norm": 1.529845118522644,
+      "learning_rate": 0.00019856572182329059,
+      "loss": 0.9169,
+      "step": 1121
+    },
+    {
+      "epoch": 0.08250606662254578,
+      "grad_norm": 1.0012229681015015,
+      "learning_rate": 0.00019856169981528053,
+      "loss": 0.9492,
+      "step": 1122
+    },
+    {
+      "epoch": 0.08257960144128244,
+      "grad_norm": 1.794523000717163,
+      "learning_rate": 0.00019855767221674682,
+      "loss": 0.9119,
+      "step": 1123
+    },
+    {
+      "epoch": 0.08265313626001912,
+      "grad_norm": 1.5841182470321655,
+      "learning_rate": 0.00019855363902791795,
+      "loss": 0.9367,
+      "step": 1124
+    },
+    {
+      "epoch": 0.0827266710787558,
+      "grad_norm": 1.2625213861465454,
+      "learning_rate": 0.00019854960024902267,
+      "loss": 0.9276,
+      "step": 1125
+    },
+    {
+      "epoch": 0.08280020589749246,
+      "grad_norm": 1.1957871913909912,
+      "learning_rate": 0.00019854555588029004,
+      "loss": 0.9145,
+      "step": 1126
+    },
+    {
+      "epoch": 0.08287374071622913,
+      "grad_norm": 1.1544164419174194,
+      "learning_rate": 0.00019854150592194948,
+      "loss": 0.9235,
+      "step": 1127
+    },
+    {
+      "epoch": 0.08294727553496581,
+      "grad_norm": 0.9589806795120239,
+      "learning_rate": 0.00019853745037423076,
+      "loss": 0.9261,
+      "step": 1128
+    },
+    {
+      "epoch": 0.08302081035370248,
+      "grad_norm": 1.130780577659607,
+      "learning_rate": 0.0001985333892373638,
+      "loss": 0.8982,
+      "step": 1129
+    },
+    {
+      "epoch": 0.08309434517243915,
+      "grad_norm": 0.9159256219863892,
+      "learning_rate": 0.00019852932251157904,
+      "loss": 0.8998,
+      "step": 1130
+    },
+    {
+      "epoch": 0.08316787999117582,
+      "grad_norm": 0.921309232711792,
+      "learning_rate": 0.00019852525019710716,
+      "loss": 0.8746,
+      "step": 1131
+    },
+    {
+      "epoch": 0.0832414148099125,
+      "grad_norm": 0.9993924498558044,
+      "learning_rate": 0.00019852117229417907,
+      "loss": 0.9024,
+      "step": 1132
+    },
+    {
+      "epoch": 0.08331494962864916,
+      "grad_norm": 0.7288271188735962,
+      "learning_rate": 0.0001985170888030261,
+      "loss": 0.9261,
+      "step": 1133
+    },
+    {
+      "epoch": 0.08338848444738584,
+      "grad_norm": 0.764900803565979,
+      "learning_rate": 0.00019851299972387993,
+      "loss": 0.8667,
+      "step": 1134
+    },
+    {
+      "epoch": 0.08346201926612251,
+      "grad_norm": 0.7006664872169495,
+      "learning_rate": 0.00019850890505697245,
+      "loss": 0.9168,
+      "step": 1135
+    },
+    {
+      "epoch": 0.08353555408485919,
+      "grad_norm": 0.6741101741790771,
+      "learning_rate": 0.00019850480480253588,
+      "loss": 0.8897,
+      "step": 1136
+    },
+    {
+      "epoch": 0.08360908890359585,
+      "grad_norm": 1.0485578775405884,
+      "learning_rate": 0.00019850069896080284,
+      "loss": 0.9326,
+      "step": 1137
+    },
+    {
+      "epoch": 0.08368262372233252,
+      "grad_norm": 0.5137832760810852,
+      "learning_rate": 0.0001984965875320062,
+      "loss": 0.9019,
+      "step": 1138
+    },
+    {
+      "epoch": 0.0837561585410692,
+      "grad_norm": 0.9950116872787476,
+      "learning_rate": 0.00019849247051637915,
+      "loss": 0.9233,
+      "step": 1139
+    },
+    {
+      "epoch": 0.08382969335980586,
+      "grad_norm": 1.138445258140564,
+      "learning_rate": 0.00019848834791415528,
+      "loss": 0.9504,
+      "step": 1140
+    },
+    {
+      "epoch": 0.08390322817854254,
+      "grad_norm": 0.5465790033340454,
+      "learning_rate": 0.00019848421972556833,
+      "loss": 0.9435,
+      "step": 1141
+    },
+    {
+      "epoch": 0.08397676299727921,
+      "grad_norm": 0.7563077211380005,
+      "learning_rate": 0.0001984800859508525,
+      "loss": 0.9138,
+      "step": 1142
+    },
+    {
+      "epoch": 0.08405029781601589,
+      "grad_norm": 0.7061545252799988,
+      "learning_rate": 0.0001984759465902423,
+      "loss": 0.9414,
+      "step": 1143
+    },
+    {
+      "epoch": 0.08412383263475255,
+      "grad_norm": 0.6039571762084961,
+      "learning_rate": 0.00019847180164397245,
+      "loss": 0.9219,
+      "step": 1144
+    },
+    {
+      "epoch": 0.08419736745348923,
+      "grad_norm": 0.49787408113479614,
+      "learning_rate": 0.00019846765111227813,
+      "loss": 0.8956,
+      "step": 1145
+    },
+    {
+      "epoch": 0.0842709022722259,
+      "grad_norm": 0.6304413676261902,
+      "learning_rate": 0.0001984634949953947,
+      "loss": 0.8993,
+      "step": 1146
+    },
+    {
+      "epoch": 0.08434443709096257,
+      "grad_norm": 0.6303095817565918,
+      "learning_rate": 0.0001984593332935579,
+      "loss": 0.9329,
+      "step": 1147
+    },
+    {
+      "epoch": 0.08441797190969924,
+      "grad_norm": 0.6377007961273193,
+      "learning_rate": 0.00019845516600700382,
+      "loss": 0.9049,
+      "step": 1148
+    },
+    {
+      "epoch": 0.08449150672843592,
+      "grad_norm": 0.7678728699684143,
+      "learning_rate": 0.00019845099313596883,
+      "loss": 0.8886,
+      "step": 1149
+    },
+    {
+      "epoch": 0.08456504154717259,
+      "grad_norm": 0.8491067290306091,
+      "learning_rate": 0.00019844681468068963,
+      "loss": 0.9191,
+      "step": 1150
+    },
+    {
+      "epoch": 0.08463857636590925,
+      "grad_norm": 1.1347951889038086,
+      "learning_rate": 0.00019844263064140315,
+      "loss": 0.8791,
+      "step": 1151
+    },
+    {
+      "epoch": 0.08471211118464593,
+      "grad_norm": 0.6839718818664551,
+      "learning_rate": 0.00019843844101834681,
+      "loss": 0.9142,
+      "step": 1152
+    },
+    {
+      "epoch": 0.0847856460033826,
+      "grad_norm": 0.5558001399040222,
+      "learning_rate": 0.0001984342458117582,
+      "loss": 0.8996,
+      "step": 1153
+    },
+    {
+      "epoch": 0.08485918082211927,
+      "grad_norm": 0.6527745723724365,
+      "learning_rate": 0.0001984300450218753,
+      "loss": 0.9251,
+      "step": 1154
+    },
+    {
+      "epoch": 0.08493271564085594,
+      "grad_norm": 0.8552815318107605,
+      "learning_rate": 0.00019842583864893633,
+      "loss": 0.91,
+      "step": 1155
+    },
+    {
+      "epoch": 0.08500625045959262,
+      "grad_norm": 0.8301436305046082,
+      "learning_rate": 0.00019842162669317997,
+      "loss": 0.8993,
+      "step": 1156
+    },
+    {
+      "epoch": 0.0850797852783293,
+      "grad_norm": 0.7786171436309814,
+      "learning_rate": 0.00019841740915484506,
+      "loss": 0.9475,
+      "step": 1157
+    },
+    {
+      "epoch": 0.08515332009706596,
+      "grad_norm": 0.6035689115524292,
+      "learning_rate": 0.00019841318603417083,
+      "loss": 0.9408,
+      "step": 1158
+    },
+    {
+      "epoch": 0.08522685491580263,
+      "grad_norm": 0.6750732064247131,
+      "learning_rate": 0.0001984089573313968,
+      "loss": 0.9016,
+      "step": 1159
+    },
+    {
+      "epoch": 0.08530038973453931,
+      "grad_norm": 1.5600361824035645,
+      "learning_rate": 0.0001984047230467629,
+      "loss": 0.8841,
+      "step": 1160
+    },
+    {
+      "epoch": 0.08537392455327597,
+      "grad_norm": 0.5783163905143738,
+      "learning_rate": 0.00019840048318050925,
+      "loss": 0.9276,
+      "step": 1161
+    },
+    {
+      "epoch": 0.08544745937201265,
+      "grad_norm": 0.7939640283584595,
+      "learning_rate": 0.00019839623773287632,
+      "loss": 0.9269,
+      "step": 1162
+    },
+    {
+      "epoch": 0.08552099419074932,
+      "grad_norm": 1.7710728645324707,
+      "learning_rate": 0.00019839198670410497,
+      "loss": 0.92,
+      "step": 1163
+    },
+    {
+      "epoch": 0.085594529009486,
+      "grad_norm": 0.5471442341804504,
+      "learning_rate": 0.0001983877300944363,
+      "loss": 0.8977,
+      "step": 1164
+    },
+    {
+      "epoch": 0.08566806382822266,
+      "grad_norm": 1.8655799627304077,
+      "learning_rate": 0.0001983834679041117,
+      "loss": 0.8643,
+      "step": 1165
+    },
+    {
+      "epoch": 0.08574159864695934,
+      "grad_norm": 0.6184312105178833,
+      "learning_rate": 0.00019837920013337303,
+      "loss": 0.9398,
+      "step": 1166
+    },
+    {
+      "epoch": 0.08581513346569601,
+      "grad_norm": 1.1598464250564575,
+      "learning_rate": 0.00019837492678246227,
+      "loss": 0.8759,
+      "step": 1167
+    },
+    {
+      "epoch": 0.08588866828443267,
+      "grad_norm": 1.5156643390655518,
+      "learning_rate": 0.00019837064785162186,
+      "loss": 0.895,
+      "step": 1168
+    },
+    {
+      "epoch": 0.08596220310316935,
+      "grad_norm": 0.6237832307815552,
+      "learning_rate": 0.00019836636334109452,
+      "loss": 0.8962,
+      "step": 1169
+    },
+    {
+      "epoch": 0.08603573792190602,
+      "grad_norm": 0.7645578980445862,
+      "learning_rate": 0.00019836207325112318,
+      "loss": 0.9132,
+      "step": 1170
+    },
+    {
+      "epoch": 0.0861092727406427,
+      "grad_norm": 1.5563639402389526,
+      "learning_rate": 0.00019835777758195127,
+      "loss": 0.9173,
+      "step": 1171
+    },
+    {
+      "epoch": 0.08618280755937936,
+      "grad_norm": 0.7761930227279663,
+      "learning_rate": 0.0001983534763338224,
+      "loss": 0.8848,
+      "step": 1172
+    },
+    {
+      "epoch": 0.08625634237811604,
+      "grad_norm": 1.3360595703125,
+      "learning_rate": 0.00019834916950698056,
+      "loss": 0.9297,
+      "step": 1173
+    },
+    {
+      "epoch": 0.08632987719685271,
+      "grad_norm": 1.5478949546813965,
+      "learning_rate": 0.00019834485710167006,
+      "loss": 0.9078,
+      "step": 1174
+    },
+    {
+      "epoch": 0.08640341201558938,
+      "grad_norm": 0.6345556974411011,
+      "learning_rate": 0.00019834053911813547,
+      "loss": 0.9042,
+      "step": 1175
+    },
+    {
+      "epoch": 0.08647694683432605,
+      "grad_norm": 2.225137233734131,
+      "learning_rate": 0.0001983362155566217,
+      "loss": 0.879,
+      "step": 1176
+    },
+    {
+      "epoch": 0.08655048165306273,
+      "grad_norm": 0.710138201713562,
+      "learning_rate": 0.000198331886417374,
+      "loss": 0.8737,
+      "step": 1177
+    },
+    {
+      "epoch": 0.0866240164717994,
+      "grad_norm": 2.1969826221466064,
+      "learning_rate": 0.00019832755170063796,
+      "loss": 0.9661,
+      "step": 1178
+    },
+    {
+      "epoch": 0.08669755129053607,
+      "grad_norm": 1.2045308351516724,
+      "learning_rate": 0.00019832321140665938,
+      "loss": 0.8646,
+      "step": 1179
+    },
+    {
+      "epoch": 0.08677108610927274,
+      "grad_norm": 2.6736440658569336,
+      "learning_rate": 0.0001983188655356845,
+      "loss": 0.9163,
+      "step": 1180
+    },
+    {
+      "epoch": 0.08684462092800942,
+      "grad_norm": 2.4958488941192627,
+      "learning_rate": 0.00019831451408795984,
+      "loss": 0.9147,
+      "step": 1181
+    },
+    {
+      "epoch": 0.08691815574674608,
+      "grad_norm": 1.2818974256515503,
+      "learning_rate": 0.00019831015706373215,
+      "loss": 0.8882,
+      "step": 1182
+    },
+    {
+      "epoch": 0.08699169056548275,
+      "grad_norm": 1.359810709953308,
+      "learning_rate": 0.00019830579446324857,
+      "loss": 0.9108,
+      "step": 1183
+    },
+    {
+      "epoch": 0.08706522538421943,
+      "grad_norm": 1.38339102268219,
+      "learning_rate": 0.00019830142628675664,
+      "loss": 0.8846,
+      "step": 1184
+    },
+    {
+      "epoch": 0.0871387602029561,
+      "grad_norm": 1.0123505592346191,
+      "learning_rate": 0.00019829705253450404,
+      "loss": 0.8973,
+      "step": 1185
+    },
+    {
+      "epoch": 0.08721229502169277,
+      "grad_norm": 1.9262945652008057,
+      "learning_rate": 0.0001982926732067389,
+      "loss": 0.8956,
+      "step": 1186
+    },
+    {
+      "epoch": 0.08728582984042944,
+      "grad_norm": 1.6566966772079468,
+      "learning_rate": 0.0001982882883037096,
+      "loss": 0.881,
+      "step": 1187
+    },
+    {
+      "epoch": 0.08735936465916612,
+      "grad_norm": 1.717227578163147,
+      "learning_rate": 0.00019828389782566483,
+      "loss": 0.8969,
+      "step": 1188
+    },
+    {
+      "epoch": 0.08743289947790278,
+      "grad_norm": 1.4622933864593506,
+      "learning_rate": 0.00019827950177285367,
+      "loss": 0.9457,
+      "step": 1189
+    },
+    {
+      "epoch": 0.08750643429663946,
+      "grad_norm": 1.2970147132873535,
+      "learning_rate": 0.00019827510014552545,
+      "loss": 0.902,
+      "step": 1190
+    },
+    {
+      "epoch": 0.08757996911537613,
+      "grad_norm": 1.2494728565216064,
+      "learning_rate": 0.00019827069294392987,
+      "loss": 0.9145,
+      "step": 1191
+    },
+    {
+      "epoch": 0.08765350393411281,
+      "grad_norm": 1.4582123756408691,
+      "learning_rate": 0.00019826628016831683,
+      "loss": 0.9011,
+      "step": 1192
+    },
+    {
+      "epoch": 0.08772703875284947,
+      "grad_norm": 1.071927785873413,
+      "learning_rate": 0.00019826186181893672,
+      "loss": 0.9186,
+      "step": 1193
+    },
+    {
+      "epoch": 0.08780057357158615,
+      "grad_norm": 1.6156177520751953,
+      "learning_rate": 0.0001982574378960401,
+      "loss": 0.9269,
+      "step": 1194
+    },
+    {
+      "epoch": 0.08787410839032282,
+      "grad_norm": 1.3256275653839111,
+      "learning_rate": 0.00019825300839987787,
+      "loss": 0.9045,
+      "step": 1195
+    },
+    {
+      "epoch": 0.08794764320905948,
+      "grad_norm": 1.4961538314819336,
+      "learning_rate": 0.00019824857333070132,
+      "loss": 0.9231,
+      "step": 1196
+    },
+    {
+      "epoch": 0.08802117802779616,
+      "grad_norm": 1.3811380863189697,
+      "learning_rate": 0.00019824413268876203,
+      "loss": 0.8907,
+      "step": 1197
+    },
+    {
+      "epoch": 0.08809471284653284,
+      "grad_norm": 1.392310380935669,
+      "learning_rate": 0.0001982396864743118,
+      "loss": 0.8855,
+      "step": 1198
+    },
+    {
+      "epoch": 0.08816824766526951,
+      "grad_norm": 1.1857701539993286,
+      "learning_rate": 0.00019823523468760293,
+      "loss": 0.8998,
+      "step": 1199
+    },
+    {
+      "epoch": 0.08824178248400617,
+      "grad_norm": 1.4353760480880737,
+      "learning_rate": 0.0001982307773288879,
+      "loss": 0.9017,
+      "step": 1200
+    },
+    {
+      "epoch": 0.08831531730274285,
+      "grad_norm": 1.1064549684524536,
+      "learning_rate": 0.00019822631439841946,
+      "loss": 0.8945,
+      "step": 1201
+    },
+    {
+      "epoch": 0.08838885212147952,
+      "grad_norm": 1.5415904521942139,
+      "learning_rate": 0.00019822184589645084,
+      "loss": 0.8685,
+      "step": 1202
+    },
+    {
+      "epoch": 0.08846238694021619,
+      "grad_norm": 1.3537554740905762,
+      "learning_rate": 0.0001982173718232354,
+      "loss": 0.8814,
+      "step": 1203
+    },
+    {
+      "epoch": 0.08853592175895286,
+      "grad_norm": 1.5644854307174683,
+      "learning_rate": 0.00019821289217902702,
+      "loss": 0.9041,
+      "step": 1204
+    },
+    {
+      "epoch": 0.08860945657768954,
+      "grad_norm": 1.5359615087509155,
+      "learning_rate": 0.00019820840696407977,
+      "loss": 0.9055,
+      "step": 1205
+    },
+    {
+      "epoch": 0.08868299139642621,
+      "grad_norm": 1.241362452507019,
+      "learning_rate": 0.00019820391617864802,
+      "loss": 0.9248,
+      "step": 1206
+    },
+    {
+      "epoch": 0.08875652621516288,
+      "grad_norm": 0.965174674987793,
+      "learning_rate": 0.0001981994198229865,
+      "loss": 0.8596,
+      "step": 1207
+    },
+    {
+      "epoch": 0.08883006103389955,
+      "grad_norm": 1.4349634647369385,
+      "learning_rate": 0.00019819491789735026,
+      "loss": 0.8778,
+      "step": 1208
+    },
+    {
+      "epoch": 0.08890359585263623,
+      "grad_norm": 0.9509841203689575,
+      "learning_rate": 0.0001981904104019946,
+      "loss": 0.8557,
+      "step": 1209
+    },
+    {
+      "epoch": 0.08897713067137289,
+      "grad_norm": 1.5240931510925293,
+      "learning_rate": 0.0001981858973371753,
+      "loss": 0.8955,
+      "step": 1210
+    },
+    {
+      "epoch": 0.08905066549010957,
+      "grad_norm": 1.1514242887496948,
+      "learning_rate": 0.00019818137870314826,
+      "loss": 0.8697,
+      "step": 1211
+    },
+    {
+      "epoch": 0.08912420030884624,
+      "grad_norm": 1.5162341594696045,
+      "learning_rate": 0.0001981768545001698,
+      "loss": 0.8803,
+      "step": 1212
+    },
+    {
+      "epoch": 0.08919773512758292,
+      "grad_norm": 1.2575844526290894,
+      "learning_rate": 0.00019817232472849657,
+      "loss": 0.8612,
+      "step": 1213
+    },
+    {
+      "epoch": 0.08927126994631958,
+      "grad_norm": 1.43647038936615,
+      "learning_rate": 0.00019816778938838546,
+      "loss": 0.9028,
+      "step": 1214
+    },
+    {
+      "epoch": 0.08934480476505625,
+      "grad_norm": 1.3336704969406128,
+      "learning_rate": 0.0001981632484800937,
+      "loss": 0.8986,
+      "step": 1215
+    },
+    {
+      "epoch": 0.08941833958379293,
+      "grad_norm": 1.5550183057785034,
+      "learning_rate": 0.00019815870200387892,
+      "loss": 0.8573,
+      "step": 1216
+    },
+    {
+      "epoch": 0.08949187440252959,
+      "grad_norm": 1.2657097578048706,
+      "learning_rate": 0.00019815414995999898,
+      "loss": 0.8347,
+      "step": 1217
+    },
+    {
+      "epoch": 0.08956540922126627,
+      "grad_norm": 1.462835431098938,
+      "learning_rate": 0.00019814959234871205,
+      "loss": 0.8384,
+      "step": 1218
+    },
+    {
+      "epoch": 0.08963894404000294,
+      "grad_norm": 1.1220712661743164,
+      "learning_rate": 0.00019814502917027668,
+      "loss": 0.8736,
+      "step": 1219
+    },
+    {
+      "epoch": 0.08971247885873962,
+      "grad_norm": 1.3983973264694214,
+      "learning_rate": 0.00019814046042495166,
+      "loss": 0.8559,
+      "step": 1220
+    },
+    {
+      "epoch": 0.08978601367747628,
+      "grad_norm": 1.1078206300735474,
+      "learning_rate": 0.00019813588611299616,
+      "loss": 0.8336,
+      "step": 1221
+    },
+    {
+      "epoch": 0.08985954849621296,
+      "grad_norm": 1.3603934049606323,
+      "learning_rate": 0.00019813130623466964,
+      "loss": 0.8996,
+      "step": 1222
+    },
+    {
+      "epoch": 0.08993308331494963,
+      "grad_norm": 1.0905827283859253,
+      "learning_rate": 0.00019812672079023186,
+      "loss": 0.8899,
+      "step": 1223
+    },
+    {
+      "epoch": 0.0900066181336863,
+      "grad_norm": 1.3743280172348022,
+      "learning_rate": 0.00019812212977994293,
+      "loss": 0.8966,
+      "step": 1224
+    },
+    {
+      "epoch": 0.09008015295242297,
+      "grad_norm": 1.0087875127792358,
+      "learning_rate": 0.00019811753320406325,
+      "loss": 0.8591,
+      "step": 1225
+    },
+    {
+      "epoch": 0.09015368777115965,
+      "grad_norm": 1.4524458646774292,
+      "learning_rate": 0.00019811293106285351,
+      "loss": 0.8556,
+      "step": 1226
+    },
+    {
+      "epoch": 0.09022722258989632,
+      "grad_norm": 1.2741235494613647,
+      "learning_rate": 0.0001981083233565748,
+      "loss": 0.8649,
+      "step": 1227
+    },
+    {
+      "epoch": 0.09030075740863298,
+      "grad_norm": 1.6416369676589966,
+      "learning_rate": 0.0001981037100854884,
+      "loss": 0.8398,
+      "step": 1228
+    },
+    {
+      "epoch": 0.09037429222736966,
+      "grad_norm": 1.2055248022079468,
+      "learning_rate": 0.00019809909124985612,
+      "loss": 0.892,
+      "step": 1229
+    },
+    {
+      "epoch": 0.09044782704610634,
+      "grad_norm": 1.0601824522018433,
+      "learning_rate": 0.00019809446684993981,
+      "loss": 0.8747,
+      "step": 1230
+    },
+    {
+      "epoch": 0.090521361864843,
+      "grad_norm": 0.9545323848724365,
+      "learning_rate": 0.00019808983688600183,
+      "loss": 0.8739,
+      "step": 1231
+    },
+    {
+      "epoch": 0.09059489668357967,
+      "grad_norm": 0.6252270936965942,
+      "learning_rate": 0.00019808520135830478,
+      "loss": 0.8859,
+      "step": 1232
+    },
+    {
+      "epoch": 0.09066843150231635,
+      "grad_norm": 0.7750497460365295,
+      "learning_rate": 0.00019808056026711157,
+      "loss": 0.8879,
+      "step": 1233
+    },
+    {
+      "epoch": 0.09074196632105302,
+      "grad_norm": 0.7908056974411011,
+      "learning_rate": 0.00019807591361268548,
+      "loss": 0.855,
+      "step": 1234
+    },
+    {
+      "epoch": 0.09081550113978969,
+      "grad_norm": 0.8334416151046753,
+      "learning_rate": 0.00019807126139529012,
+      "loss": 0.8861,
+      "step": 1235
+    },
+    {
+      "epoch": 0.09088903595852636,
+      "grad_norm": 0.7295163869857788,
+      "learning_rate": 0.00019806660361518924,
+      "loss": 0.8762,
+      "step": 1236
+    },
+    {
+      "epoch": 0.09096257077726304,
+      "grad_norm": 0.804947018623352,
+      "learning_rate": 0.00019806194027264715,
+      "loss": 0.9153,
+      "step": 1237
+    },
+    {
+      "epoch": 0.0910361055959997,
+      "grad_norm": 0.7026077508926392,
+      "learning_rate": 0.0001980572713679283,
+      "loss": 0.9007,
+      "step": 1238
+    },
+    {
+      "epoch": 0.09110964041473638,
+      "grad_norm": 0.7753552198410034,
+      "learning_rate": 0.00019805259690129756,
+      "loss": 0.8932,
+      "step": 1239
+    },
+    {
+      "epoch": 0.09118317523347305,
+      "grad_norm": 0.8694870471954346,
+      "learning_rate": 0.00019804791687302005,
+      "loss": 0.9137,
+      "step": 1240
+    },
+    {
+      "epoch": 0.09125671005220973,
+      "grad_norm": 0.6515732407569885,
+      "learning_rate": 0.00019804323128336118,
+      "loss": 0.9239,
+      "step": 1241
+    },
+    {
+      "epoch": 0.09133024487094639,
+      "grad_norm": 1.0462077856063843,
+      "learning_rate": 0.00019803854013258679,
+      "loss": 0.8675,
+      "step": 1242
+    },
+    {
+      "epoch": 0.09140377968968307,
+      "grad_norm": 0.668405294418335,
+      "learning_rate": 0.00019803384342096295,
+      "loss": 0.8699,
+      "step": 1243
+    },
+    {
+      "epoch": 0.09147731450841974,
+      "grad_norm": 0.7597375512123108,
+      "learning_rate": 0.00019802914114875602,
+      "loss": 0.8817,
+      "step": 1244
+    },
+    {
+      "epoch": 0.0915508493271564,
+      "grad_norm": 1.0701277256011963,
+      "learning_rate": 0.00019802443331623276,
+      "loss": 0.8859,
+      "step": 1245
+    },
+    {
+      "epoch": 0.09162438414589308,
+      "grad_norm": 0.8760750889778137,
+      "learning_rate": 0.0001980197199236602,
+      "loss": 0.8868,
+      "step": 1246
+    },
+    {
+      "epoch": 0.09169791896462975,
+      "grad_norm": 1.2690455913543701,
+      "learning_rate": 0.00019801500097130567,
+      "loss": 0.9038,
+      "step": 1247
+    },
+    {
+      "epoch": 0.09177145378336643,
+      "grad_norm": 0.9959099292755127,
+      "learning_rate": 0.00019801027645943687,
+      "loss": 0.899,
+      "step": 1248
+    },
+    {
+      "epoch": 0.09184498860210309,
+      "grad_norm": 0.9615917801856995,
+      "learning_rate": 0.00019800554638832173,
+      "loss": 0.9045,
+      "step": 1249
+    },
+    {
+      "epoch": 0.09191852342083977,
+      "grad_norm": 0.815589964389801,
+      "learning_rate": 0.00019800081075822858,
+      "loss": 0.8821,
+      "step": 1250
+    },
+    {
+      "epoch": 0.09199205823957644,
+      "grad_norm": 1.6022166013717651,
+      "learning_rate": 0.00019799606956942602,
+      "loss": 0.9093,
+      "step": 1251
+    },
+    {
+      "epoch": 0.0920655930583131,
+      "grad_norm": 0.6787603497505188,
+      "learning_rate": 0.00019799132282218298,
+      "loss": 0.9074,
+      "step": 1252
+    },
+    {
+      "epoch": 0.09213912787704978,
+      "grad_norm": 1.8932757377624512,
+      "learning_rate": 0.00019798657051676869,
+      "loss": 0.889,
+      "step": 1253
+    },
+    {
+      "epoch": 0.09221266269578646,
+      "grad_norm": 1.1454704999923706,
+      "learning_rate": 0.00019798181265345274,
+      "loss": 0.8915,
+      "step": 1254
+    },
+    {
+      "epoch": 0.09228619751452313,
+      "grad_norm": 1.6034833192825317,
+      "learning_rate": 0.0001979770492325049,
+      "loss": 0.9376,
+      "step": 1255
+    },
+    {
+      "epoch": 0.0923597323332598,
+      "grad_norm": 1.4970453977584839,
+      "learning_rate": 0.0001979722802541955,
+      "loss": 0.91,
+      "step": 1256
+    },
+    {
+      "epoch": 0.09243326715199647,
+      "grad_norm": 0.9006907939910889,
+      "learning_rate": 0.00019796750571879492,
+      "loss": 0.9143,
+      "step": 1257
+    },
+    {
+      "epoch": 0.09250680197073315,
+      "grad_norm": 0.8779823184013367,
+      "learning_rate": 0.00019796272562657406,
+      "loss": 0.9376,
+      "step": 1258
+    },
+    {
+      "epoch": 0.09258033678946981,
+      "grad_norm": 0.7195339202880859,
+      "learning_rate": 0.000197957939977804,
+      "loss": 0.9346,
+      "step": 1259
+    },
+    {
+      "epoch": 0.09265387160820648,
+      "grad_norm": 1.1694622039794922,
+      "learning_rate": 0.00019795314877275622,
+      "loss": 0.9411,
+      "step": 1260
+    },
+    {
+      "epoch": 0.09272740642694316,
+      "grad_norm": 0.9884313941001892,
+      "learning_rate": 0.00019794835201170243,
+      "loss": 0.9456,
+      "step": 1261
+    },
+    {
+      "epoch": 0.09280094124567984,
+      "grad_norm": 0.9206125736236572,
+      "learning_rate": 0.0001979435496949148,
+      "loss": 0.9107,
+      "step": 1262
+    },
+    {
+      "epoch": 0.0928744760644165,
+      "grad_norm": 1.2713993787765503,
+      "learning_rate": 0.00019793874182266562,
+      "loss": 0.9018,
+      "step": 1263
+    },
+    {
+      "epoch": 0.09294801088315317,
+      "grad_norm": 1.1546505689620972,
+      "learning_rate": 0.00019793392839522766,
+      "loss": 0.9418,
+      "step": 1264
+    },
+    {
+      "epoch": 0.09302154570188985,
+      "grad_norm": 0.8445000052452087,
+      "learning_rate": 0.00019792910941287395,
+      "loss": 0.9176,
+      "step": 1265
+    },
+    {
+      "epoch": 0.09309508052062651,
+      "grad_norm": 0.9914439916610718,
+      "learning_rate": 0.00019792428487587778,
+      "loss": 0.9077,
+      "step": 1266
+    },
+    {
+      "epoch": 0.09316861533936319,
+      "grad_norm": 1.369713544845581,
+      "learning_rate": 0.00019791945478451282,
+      "loss": 0.8871,
+      "step": 1267
+    },
+    {
+      "epoch": 0.09324215015809986,
+      "grad_norm": 0.5963691473007202,
+      "learning_rate": 0.00019791461913905306,
+      "loss": 0.9085,
+      "step": 1268
+    },
+    {
+      "epoch": 0.09331568497683654,
+      "grad_norm": 1.0341975688934326,
+      "learning_rate": 0.00019790977793977279,
+      "loss": 0.913,
+      "step": 1269
+    },
+    {
+      "epoch": 0.0933892197955732,
+      "grad_norm": 1.283675193786621,
+      "learning_rate": 0.00019790493118694658,
+      "loss": 0.9116,
+      "step": 1270
+    },
+    {
+      "epoch": 0.09346275461430988,
+      "grad_norm": 0.776107668876648,
+      "learning_rate": 0.00019790007888084935,
+      "loss": 0.9035,
+      "step": 1271
+    },
+    {
+      "epoch": 0.09353628943304655,
+      "grad_norm": 0.5132461786270142,
+      "learning_rate": 0.0001978952210217563,
+      "loss": 0.8848,
+      "step": 1272
+    },
+    {
+      "epoch": 0.09360982425178321,
+      "grad_norm": 0.5506516098976135,
+      "learning_rate": 0.000197890357609943,
+      "loss": 0.8908,
+      "step": 1273
+    },
+    {
+      "epoch": 0.09368335907051989,
+      "grad_norm": 0.6367899775505066,
+      "learning_rate": 0.00019788548864568535,
+      "loss": 0.8987,
+      "step": 1274
+    },
+    {
+      "epoch": 0.09375689388925657,
+      "grad_norm": 0.8592755794525146,
+      "learning_rate": 0.00019788061412925943,
+      "loss": 0.8992,
+      "step": 1275
+    },
+    {
+      "epoch": 0.09383042870799324,
+      "grad_norm": 1.1213024854660034,
+      "learning_rate": 0.0001978757340609418,
+      "loss": 0.9102,
+      "step": 1276
+    },
+    {
+      "epoch": 0.0939039635267299,
+      "grad_norm": 0.5707221627235413,
+      "learning_rate": 0.00019787084844100926,
+      "loss": 0.9204,
+      "step": 1277
+    },
+    {
+      "epoch": 0.09397749834546658,
+      "grad_norm": 1.024032473564148,
+      "learning_rate": 0.0001978659572697389,
+      "loss": 0.9139,
+      "step": 1278
+    },
+    {
+      "epoch": 0.09405103316420325,
+      "grad_norm": 1.5735774040222168,
+      "learning_rate": 0.00019786106054740815,
+      "loss": 0.9172,
+      "step": 1279
+    },
+    {
+      "epoch": 0.09412456798293992,
+      "grad_norm": 0.7786800265312195,
+      "learning_rate": 0.0001978561582742948,
+      "loss": 0.933,
+      "step": 1280
+    },
+    {
+      "epoch": 0.09419810280167659,
+      "grad_norm": 0.8323783874511719,
+      "learning_rate": 0.00019785125045067683,
+      "loss": 0.8965,
+      "step": 1281
+    },
+    {
+      "epoch": 0.09427163762041327,
+      "grad_norm": 1.3607269525527954,
+      "learning_rate": 0.0001978463370768327,
+      "loss": 0.9015,
+      "step": 1282
+    },
+    {
+      "epoch": 0.09434517243914994,
+      "grad_norm": 1.3546491861343384,
+      "learning_rate": 0.00019784141815304106,
+      "loss": 0.8619,
+      "step": 1283
+    },
+    {
+      "epoch": 0.0944187072578866,
+      "grad_norm": 0.6951857805252075,
+      "learning_rate": 0.00019783649367958095,
+      "loss": 0.8803,
+      "step": 1284
+    },
+    {
+      "epoch": 0.09449224207662328,
+      "grad_norm": 1.8383643627166748,
+      "learning_rate": 0.00019783156365673165,
+      "loss": 0.9104,
+      "step": 1285
+    },
+    {
+      "epoch": 0.09456577689535996,
+      "grad_norm": 0.8697600364685059,
+      "learning_rate": 0.0001978266280847728,
+      "loss": 0.9237,
+      "step": 1286
+    },
+    {
+      "epoch": 0.09463931171409662,
+      "grad_norm": 2.2567882537841797,
+      "learning_rate": 0.0001978216869639844,
+      "loss": 0.9076,
+      "step": 1287
+    },
+    {
+      "epoch": 0.0947128465328333,
+      "grad_norm": 1.8043625354766846,
+      "learning_rate": 0.00019781674029464665,
+      "loss": 0.8932,
+      "step": 1288
+    },
+    {
+      "epoch": 0.09478638135156997,
+      "grad_norm": 1.504205584526062,
+      "learning_rate": 0.00019781178807704017,
+      "loss": 0.9108,
+      "step": 1289
+    },
+    {
+      "epoch": 0.09485991617030665,
+      "grad_norm": 1.1635217666625977,
+      "learning_rate": 0.00019780683031144584,
+      "loss": 0.8793,
+      "step": 1290
+    },
+    {
+      "epoch": 0.09493345098904331,
+      "grad_norm": 0.685609757900238,
+      "learning_rate": 0.00019780186699814488,
+      "loss": 0.9101,
+      "step": 1291
+    },
+    {
+      "epoch": 0.09500698580777998,
+      "grad_norm": 1.0636645555496216,
+      "learning_rate": 0.0001977968981374188,
+      "loss": 0.8772,
+      "step": 1292
+    },
+    {
+      "epoch": 0.09508052062651666,
+      "grad_norm": 0.8581905364990234,
+      "learning_rate": 0.00019779192372954945,
+      "loss": 0.8635,
+      "step": 1293
+    },
+    {
+      "epoch": 0.09515405544525332,
+      "grad_norm": 1.681488037109375,
+      "learning_rate": 0.000197786943774819,
+      "loss": 0.8813,
+      "step": 1294
+    },
+    {
+      "epoch": 0.09522759026399,
+      "grad_norm": 0.9313204288482666,
+      "learning_rate": 0.00019778195827350987,
+      "loss": 0.9034,
+      "step": 1295
+    },
+    {
+      "epoch": 0.09530112508272667,
+      "grad_norm": 1.9232078790664673,
+      "learning_rate": 0.0001977769672259049,
+      "loss": 0.9111,
+      "step": 1296
+    },
+    {
+      "epoch": 0.09537465990146335,
+      "grad_norm": 1.254762887954712,
+      "learning_rate": 0.00019777197063228715,
+      "loss": 0.9062,
+      "step": 1297
+    },
+    {
+      "epoch": 0.09544819472020001,
+      "grad_norm": 2.0289361476898193,
+      "learning_rate": 0.00019776696849294005,
+      "loss": 0.8706,
+      "step": 1298
+    },
+    {
+      "epoch": 0.09552172953893669,
+      "grad_norm": 0.992583155632019,
+      "learning_rate": 0.0001977619608081473,
+      "loss": 0.9399,
+      "step": 1299
+    },
+    {
+      "epoch": 0.09559526435767336,
+      "grad_norm": 1.2118895053863525,
+      "learning_rate": 0.000197756947578193,
+      "loss": 0.8904,
+      "step": 1300
+    },
+    {
+      "epoch": 0.09566879917641002,
+      "grad_norm": 0.8679304122924805,
+      "learning_rate": 0.00019775192880336145,
+      "loss": 0.8691,
+      "step": 1301
+    },
+    {
+      "epoch": 0.0957423339951467,
+      "grad_norm": 0.9420344829559326,
+      "learning_rate": 0.00019774690448393732,
+      "loss": 0.935,
+      "step": 1302
+    },
+    {
+      "epoch": 0.09581586881388338,
+      "grad_norm": 1.341454029083252,
+      "learning_rate": 0.0001977418746202056,
+      "loss": 0.8545,
+      "step": 1303
+    },
+    {
+      "epoch": 0.09588940363262005,
+      "grad_norm": 0.8569790720939636,
+      "learning_rate": 0.00019773683921245163,
+      "loss": 0.8935,
+      "step": 1304
+    },
+    {
+      "epoch": 0.09596293845135671,
+      "grad_norm": 1.6506984233856201,
+      "learning_rate": 0.000197731798260961,
+      "loss": 0.9455,
+      "step": 1305
+    },
+    {
+      "epoch": 0.09603647327009339,
+      "grad_norm": 1.226545810699463,
+      "learning_rate": 0.0001977267517660196,
+      "loss": 0.9237,
+      "step": 1306
+    },
+    {
+      "epoch": 0.09611000808883007,
+      "grad_norm": 0.9278528094291687,
+      "learning_rate": 0.00019772169972791375,
+      "loss": 0.8906,
+      "step": 1307
+    },
+    {
+      "epoch": 0.09618354290756673,
+      "grad_norm": 2.1230297088623047,
+      "learning_rate": 0.0001977166421469299,
+      "loss": 0.8677,
+      "step": 1308
+    },
+    {
+      "epoch": 0.0962570777263034,
+      "grad_norm": 1.8557162284851074,
+      "learning_rate": 0.00019771157902335502,
+      "loss": 0.9745,
+      "step": 1309
+    },
+    {
+      "epoch": 0.09633061254504008,
+      "grad_norm": 2.444681167602539,
+      "learning_rate": 0.00019770651035747627,
+      "loss": 0.9578,
+      "step": 1310
+    },
+    {
+      "epoch": 0.09640414736377675,
+      "grad_norm": 2.2395617961883545,
+      "learning_rate": 0.00019770143614958113,
+      "loss": 0.9289,
+      "step": 1311
+    },
+    {
+      "epoch": 0.09647768218251342,
+      "grad_norm": 1.029771327972412,
+      "learning_rate": 0.0001976963563999574,
+      "loss": 0.954,
+      "step": 1312
+    },
+    {
+      "epoch": 0.09655121700125009,
+      "grad_norm": 1.494878888130188,
+      "learning_rate": 0.00019769127110889326,
+      "loss": 0.9412,
+      "step": 1313
+    },
+    {
+      "epoch": 0.09662475181998677,
+      "grad_norm": 0.7268884181976318,
+      "learning_rate": 0.00019768618027667708,
+      "loss": 0.9432,
+      "step": 1314
+    },
+    {
+      "epoch": 0.09669828663872343,
+      "grad_norm": 1.0709764957427979,
+      "learning_rate": 0.00019768108390359773,
+      "loss": 0.9825,
+      "step": 1315
+    },
+    {
+      "epoch": 0.0967718214574601,
+      "grad_norm": 1.7663569450378418,
+      "learning_rate": 0.00019767598198994416,
+      "loss": 0.9493,
+      "step": 1316
+    },
+    {
+      "epoch": 0.09684535627619678,
+      "grad_norm": 1.2842752933502197,
+      "learning_rate": 0.00019767087453600586,
+      "loss": 0.9357,
+      "step": 1317
+    },
+    {
+      "epoch": 0.09691889109493346,
+      "grad_norm": 1.0778850317001343,
+      "learning_rate": 0.00019766576154207243,
+      "loss": 0.9291,
+      "step": 1318
+    },
+    {
+      "epoch": 0.09699242591367012,
+      "grad_norm": 2.250833749771118,
+      "learning_rate": 0.00019766064300843396,
+      "loss": 0.9157,
+      "step": 1319
+    },
+    {
+      "epoch": 0.0970659607324068,
+      "grad_norm": 1.5535889863967896,
+      "learning_rate": 0.00019765551893538075,
+      "loss": 0.9788,
+      "step": 1320
+    },
+    {
+      "epoch": 0.09713949555114347,
+      "grad_norm": 2.2681171894073486,
+      "learning_rate": 0.00019765038932320343,
+      "loss": 0.8741,
+      "step": 1321
+    },
+    {
+      "epoch": 0.09721303036988013,
+      "grad_norm": 2.309908390045166,
+      "learning_rate": 0.00019764525417219299,
+      "loss": 0.9498,
+      "step": 1322
+    },
+    {
+      "epoch": 0.09728656518861681,
+      "grad_norm": 1.2105283737182617,
+      "learning_rate": 0.0001976401134826407,
+      "loss": 0.9316,
+      "step": 1323
+    },
+    {
+      "epoch": 0.09736010000735348,
+      "grad_norm": 1.0542690753936768,
+      "learning_rate": 0.0001976349672548381,
+      "loss": 0.9011,
+      "step": 1324
+    },
+    {
+      "epoch": 0.09743363482609016,
+      "grad_norm": 1.3493800163269043,
+      "learning_rate": 0.00019762981548907712,
+      "loss": 0.9135,
+      "step": 1325
+    },
+    {
+      "epoch": 0.09750716964482682,
+      "grad_norm": 0.8972010016441345,
+      "learning_rate": 0.00019762465818565,
+      "loss": 0.9141,
+      "step": 1326
+    },
+    {
+      "epoch": 0.0975807044635635,
+      "grad_norm": 1.5897903442382812,
+      "learning_rate": 0.00019761949534484924,
+      "loss": 0.9258,
+      "step": 1327
+    },
+    {
+      "epoch": 0.09765423928230017,
+      "grad_norm": 0.7470709085464478,
+      "learning_rate": 0.00019761432696696764,
+      "loss": 0.9096,
+      "step": 1328
+    },
+    {
+      "epoch": 0.09772777410103684,
+      "grad_norm": 1.1395912170410156,
+      "learning_rate": 0.0001976091530522984,
+      "loss": 0.9185,
+      "step": 1329
+    },
+    {
+      "epoch": 0.09780130891977351,
+      "grad_norm": 1.160820722579956,
+      "learning_rate": 0.00019760397360113504,
+      "loss": 0.9274,
+      "step": 1330
+    },
+    {
+      "epoch": 0.09787484373851019,
+      "grad_norm": 0.8843382000923157,
+      "learning_rate": 0.00019759878861377126,
+      "loss": 0.9221,
+      "step": 1331
+    },
+    {
+      "epoch": 0.09794837855724686,
+      "grad_norm": 0.820218563079834,
+      "learning_rate": 0.00019759359809050117,
+      "loss": 0.9462,
+      "step": 1332
+    },
+    {
+      "epoch": 0.09802191337598352,
+      "grad_norm": 0.5992587208747864,
+      "learning_rate": 0.00019758840203161924,
+      "loss": 0.9028,
+      "step": 1333
+    },
+    {
+      "epoch": 0.0980954481947202,
+      "grad_norm": 1.0978137254714966,
+      "learning_rate": 0.00019758320043742014,
+      "loss": 0.8973,
+      "step": 1334
+    },
+    {
+      "epoch": 0.09816898301345688,
+      "grad_norm": 1.1506785154342651,
+      "learning_rate": 0.00019757799330819893,
+      "loss": 0.9114,
+      "step": 1335
+    },
+    {
+      "epoch": 0.09824251783219354,
+      "grad_norm": 0.9300787448883057,
+      "learning_rate": 0.00019757278064425098,
+      "loss": 0.9127,
+      "step": 1336
+    },
+    {
+      "epoch": 0.09831605265093021,
+      "grad_norm": 0.8314434289932251,
+      "learning_rate": 0.0001975675624458719,
+      "loss": 0.8936,
+      "step": 1337
+    },
+    {
+      "epoch": 0.09838958746966689,
+      "grad_norm": 1.0089665651321411,
+      "learning_rate": 0.00019756233871335774,
+      "loss": 0.9093,
+      "step": 1338
+    },
+    {
+      "epoch": 0.09846312228840357,
+      "grad_norm": 0.6183509230613708,
+      "learning_rate": 0.00019755710944700474,
+      "loss": 0.9235,
+      "step": 1339
+    },
+    {
+      "epoch": 0.09853665710714023,
+      "grad_norm": 1.2781784534454346,
+      "learning_rate": 0.00019755187464710956,
+      "loss": 0.9321,
+      "step": 1340
+    },
+    {
+      "epoch": 0.0986101919258769,
+      "grad_norm": 1.0182865858078003,
+      "learning_rate": 0.0001975466343139691,
+      "loss": 0.8882,
+      "step": 1341
+    },
+    {
+      "epoch": 0.09868372674461358,
+      "grad_norm": 0.959937334060669,
+      "learning_rate": 0.0001975413884478806,
+      "loss": 0.9103,
+      "step": 1342
+    },
+    {
+      "epoch": 0.09875726156335024,
+      "grad_norm": 1.4001015424728394,
+      "learning_rate": 0.00019753613704914157,
+      "loss": 0.9038,
+      "step": 1343
+    },
+    {
+      "epoch": 0.09883079638208692,
+      "grad_norm": 1.0472127199172974,
+      "learning_rate": 0.00019753088011804993,
+      "loss": 0.9345,
+      "step": 1344
+    },
+    {
+      "epoch": 0.09890433120082359,
+      "grad_norm": 0.7107047438621521,
+      "learning_rate": 0.00019752561765490388,
+      "loss": 0.8873,
+      "step": 1345
+    },
+    {
+      "epoch": 0.09897786601956027,
+      "grad_norm": 1.8709827661514282,
+      "learning_rate": 0.00019752034966000185,
+      "loss": 0.9249,
+      "step": 1346
+    },
+    {
+      "epoch": 0.09905140083829693,
+      "grad_norm": 0.7095510363578796,
+      "learning_rate": 0.00019751507613364265,
+      "loss": 0.918,
+      "step": 1347
+    },
+    {
+      "epoch": 0.0991249356570336,
+      "grad_norm": 2.17832088470459,
+      "learning_rate": 0.00019750979707612544,
+      "loss": 0.9642,
+      "step": 1348
+    },
+    {
+      "epoch": 0.09919847047577028,
+      "grad_norm": 1.058514952659607,
+      "learning_rate": 0.00019750451248774962,
+      "loss": 0.9253,
+      "step": 1349
+    },
+    {
+      "epoch": 0.09927200529450694,
+      "grad_norm": 2.1694207191467285,
+      "learning_rate": 0.00019749922236881497,
+      "loss": 0.9148,
+      "step": 1350
+    },
+    {
+      "epoch": 0.09934554011324362,
+      "grad_norm": 1.622204065322876,
+      "learning_rate": 0.0001974939267196215,
+      "loss": 0.9385,
+      "step": 1351
+    },
+    {
+      "epoch": 0.0994190749319803,
+      "grad_norm": 2.0359880924224854,
+      "learning_rate": 0.00019748862554046966,
+      "loss": 0.9198,
+      "step": 1352
+    },
+    {
+      "epoch": 0.09949260975071697,
+      "grad_norm": 1.5862441062927246,
+      "learning_rate": 0.00019748331883166005,
+      "loss": 0.9026,
+      "step": 1353
+    },
+    {
+      "epoch": 0.09956614456945363,
+      "grad_norm": 1.5793598890304565,
+      "learning_rate": 0.00019747800659349375,
+      "loss": 0.9194,
+      "step": 1354
+    },
+    {
+      "epoch": 0.09963967938819031,
+      "grad_norm": 1.1820104122161865,
+      "learning_rate": 0.00019747268882627203,
+      "loss": 0.9189,
+      "step": 1355
+    },
+    {
+      "epoch": 0.09971321420692698,
+      "grad_norm": 2.0449838638305664,
+      "learning_rate": 0.00019746736553029654,
+      "loss": 0.9019,
+      "step": 1356
+    },
+    {
+      "epoch": 0.09978674902566365,
+      "grad_norm": 1.775578260421753,
+      "learning_rate": 0.00019746203670586922,
+      "loss": 0.9426,
+      "step": 1357
+    },
+    {
+      "epoch": 0.09986028384440032,
+      "grad_norm": 1.7284780740737915,
+      "learning_rate": 0.00019745670235329232,
+      "loss": 0.8944,
+      "step": 1358
+    },
+    {
+      "epoch": 0.099933818663137,
+      "grad_norm": 1.5911685228347778,
+      "learning_rate": 0.0001974513624728684,
+      "loss": 0.9238,
+      "step": 1359
+    },
+    {
+      "epoch": 0.10000735348187367,
+      "grad_norm": 1.2739384174346924,
+      "learning_rate": 0.00019744601706490035,
+      "loss": 0.9238,
+      "step": 1360
+    },
+    {
+      "epoch": 0.10008088830061034,
+      "grad_norm": 1.3835335969924927,
+      "learning_rate": 0.0001974406661296914,
+      "loss": 0.9059,
+      "step": 1361
+    },
+    {
+      "epoch": 0.10015442311934701,
+      "grad_norm": 1.5276079177856445,
+      "learning_rate": 0.00019743530966754503,
+      "loss": 0.9026,
+      "step": 1362
+    },
+    {
+      "epoch": 0.10022795793808369,
+      "grad_norm": 0.920304000377655,
+      "learning_rate": 0.00019742994767876505,
+      "loss": 0.9036,
+      "step": 1363
+    },
+    {
+      "epoch": 0.10030149275682035,
+      "grad_norm": 0.8257358074188232,
+      "learning_rate": 0.00019742458016365564,
+      "loss": 0.9337,
+      "step": 1364
+    },
+    {
+      "epoch": 0.10037502757555702,
+      "grad_norm": 0.784819483757019,
+      "learning_rate": 0.0001974192071225212,
+      "loss": 0.8833,
+      "step": 1365
+    },
+    {
+      "epoch": 0.1004485623942937,
+      "grad_norm": 0.98670893907547,
+      "learning_rate": 0.00019741382855566656,
+      "loss": 0.9403,
+      "step": 1366
+    },
+    {
+      "epoch": 0.10052209721303038,
+      "grad_norm": 1.113434076309204,
+      "learning_rate": 0.00019740844446339672,
+      "loss": 0.9208,
+      "step": 1367
+    },
+    {
+      "epoch": 0.10059563203176704,
+      "grad_norm": 0.7199136018753052,
+      "learning_rate": 0.0001974030548460171,
+      "loss": 0.8872,
+      "step": 1368
+    },
+    {
+      "epoch": 0.10066916685050371,
+      "grad_norm": 1.163401484489441,
+      "learning_rate": 0.00019739765970383343,
+      "loss": 0.9455,
+      "step": 1369
+    },
+    {
+      "epoch": 0.10074270166924039,
+      "grad_norm": 0.8042177557945251,
+      "learning_rate": 0.00019739225903715174,
+      "loss": 0.898,
+      "step": 1370
+    },
+    {
+      "epoch": 0.10081623648797705,
+      "grad_norm": 0.7965889573097229,
+      "learning_rate": 0.00019738685284627832,
+      "loss": 0.9238,
+      "step": 1371
+    },
+    {
+      "epoch": 0.10088977130671373,
+      "grad_norm": 0.9893072247505188,
+      "learning_rate": 0.00019738144113151985,
+      "loss": 0.9304,
+      "step": 1372
+    },
+    {
+      "epoch": 0.1009633061254504,
+      "grad_norm": 1.085226058959961,
+      "learning_rate": 0.00019737602389318322,
+      "loss": 0.9507,
+      "step": 1373
+    },
+    {
+      "epoch": 0.10103684094418708,
+      "grad_norm": 1.2602131366729736,
+      "learning_rate": 0.00019737060113157577,
+      "loss": 0.9717,
+      "step": 1374
+    },
+    {
+      "epoch": 0.10111037576292374,
+      "grad_norm": 1.5903875827789307,
+      "learning_rate": 0.0001973651728470051,
+      "loss": 0.9304,
+      "step": 1375
+    },
+    {
+      "epoch": 0.10118391058166042,
+      "grad_norm": 0.6397833824157715,
+      "learning_rate": 0.00019735973903977903,
+      "loss": 0.933,
+      "step": 1376
+    },
+    {
+      "epoch": 0.10125744540039709,
+      "grad_norm": 1.6206276416778564,
+      "learning_rate": 0.00019735429971020583,
+      "loss": 0.9167,
+      "step": 1377
+    },
+    {
+      "epoch": 0.10133098021913375,
+      "grad_norm": 1.4400920867919922,
+      "learning_rate": 0.000197348854858594,
+      "loss": 0.9737,
+      "step": 1378
+    },
+    {
+      "epoch": 0.10140451503787043,
+      "grad_norm": 1.1403254270553589,
+      "learning_rate": 0.0001973434044852524,
+      "loss": 0.8984,
+      "step": 1379
+    },
+    {
+      "epoch": 0.1014780498566071,
+      "grad_norm": 2.3616204261779785,
+      "learning_rate": 0.00019733794859049016,
+      "loss": 0.9403,
+      "step": 1380
+    },
+    {
+      "epoch": 0.10155158467534378,
+      "grad_norm": 1.7736711502075195,
+      "learning_rate": 0.00019733248717461674,
+      "loss": 0.9125,
+      "step": 1381
+    },
+    {
+      "epoch": 0.10162511949408044,
+      "grad_norm": 1.5461795330047607,
+      "learning_rate": 0.00019732702023794193,
+      "loss": 0.9102,
+      "step": 1382
+    },
+    {
+      "epoch": 0.10169865431281712,
+      "grad_norm": 1.7316179275512695,
+      "learning_rate": 0.00019732154778077585,
+      "loss": 0.938,
+      "step": 1383
+    },
+    {
+      "epoch": 0.1017721891315538,
+      "grad_norm": 1.6892077922821045,
+      "learning_rate": 0.00019731606980342883,
+      "loss": 0.951,
+      "step": 1384
+    },
+    {
+      "epoch": 0.10184572395029046,
+      "grad_norm": 1.2948482036590576,
+      "learning_rate": 0.00019731058630621164,
+      "loss": 0.8892,
+      "step": 1385
+    },
+    {
+      "epoch": 0.10191925876902713,
+      "grad_norm": 1.855499029159546,
+      "learning_rate": 0.00019730509728943535,
+      "loss": 0.9078,
+      "step": 1386
+    },
+    {
+      "epoch": 0.10199279358776381,
+      "grad_norm": 1.2775121927261353,
+      "learning_rate": 0.00019729960275341118,
+      "loss": 0.8584,
+      "step": 1387
+    },
+    {
+      "epoch": 0.10206632840650048,
+      "grad_norm": 2.050262928009033,
+      "learning_rate": 0.0001972941026984509,
+      "loss": 0.8974,
+      "step": 1388
+    },
+    {
+      "epoch": 0.10213986322523715,
+      "grad_norm": 1.9408633708953857,
+      "learning_rate": 0.0001972885971248664,
+      "loss": 0.9332,
+      "step": 1389
+    },
+    {
+      "epoch": 0.10221339804397382,
+      "grad_norm": 1.4905468225479126,
+      "learning_rate": 0.00019728308603297003,
+      "loss": 0.937,
+      "step": 1390
+    },
+    {
+      "epoch": 0.1022869328627105,
+      "grad_norm": 1.3605905771255493,
+      "learning_rate": 0.00019727756942307435,
+      "loss": 0.9012,
+      "step": 1391
+    },
+    {
+      "epoch": 0.10236046768144716,
+      "grad_norm": 1.516457200050354,
+      "learning_rate": 0.00019727204729549227,
+      "loss": 0.9157,
+      "step": 1392
+    },
+    {
+      "epoch": 0.10243400250018384,
+      "grad_norm": 1.3758314847946167,
+      "learning_rate": 0.000197266519650537,
+      "loss": 0.9122,
+      "step": 1393
+    },
+    {
+      "epoch": 0.10250753731892051,
+      "grad_norm": 1.3712369203567505,
+      "learning_rate": 0.0001972609864885221,
+      "loss": 0.8844,
+      "step": 1394
+    },
+    {
+      "epoch": 0.10258107213765719,
+      "grad_norm": 1.3384265899658203,
+      "learning_rate": 0.0001972554478097614,
+      "loss": 0.9263,
+      "step": 1395
+    },
+    {
+      "epoch": 0.10265460695639385,
+      "grad_norm": 1.3354859352111816,
+      "learning_rate": 0.00019724990361456906,
+      "loss": 0.9218,
+      "step": 1396
+    },
+    {
+      "epoch": 0.10272814177513052,
+      "grad_norm": 1.346414566040039,
+      "learning_rate": 0.00019724435390325955,
+      "loss": 0.9084,
+      "step": 1397
+    },
+    {
+      "epoch": 0.1028016765938672,
+      "grad_norm": 1.3093671798706055,
+      "learning_rate": 0.00019723879867614765,
+      "loss": 0.8885,
+      "step": 1398
+    },
+    {
+      "epoch": 0.10287521141260386,
+      "grad_norm": 1.1431535482406616,
+      "learning_rate": 0.00019723323793354852,
+      "loss": 0.9037,
+      "step": 1399
+    },
+    {
+      "epoch": 0.10294874623134054,
+      "grad_norm": 1.1548740863800049,
+      "learning_rate": 0.00019722767167577746,
+      "loss": 0.9083,
+      "step": 1400
+    },
+    {
+      "epoch": 0.10302228105007721,
+      "grad_norm": 1.2153936624526978,
+      "learning_rate": 0.00019722209990315028,
+      "loss": 0.8966,
+      "step": 1401
+    },
+    {
+      "epoch": 0.10309581586881389,
+      "grad_norm": 0.8711536526679993,
+      "learning_rate": 0.000197216522615983,
+      "loss": 0.8947,
+      "step": 1402
+    },
+    {
+      "epoch": 0.10316935068755055,
+      "grad_norm": 1.2730368375778198,
+      "learning_rate": 0.00019721093981459194,
+      "loss": 0.9161,
+      "step": 1403
+    },
+    {
+      "epoch": 0.10324288550628723,
+      "grad_norm": 0.7239508032798767,
+      "learning_rate": 0.0001972053514992938,
+      "loss": 0.9162,
+      "step": 1404
+    },
+    {
+      "epoch": 0.1033164203250239,
+      "grad_norm": 1.3071931600570679,
+      "learning_rate": 0.00019719975767040554,
+      "loss": 0.8679,
+      "step": 1405
+    },
+    {
+      "epoch": 0.10338995514376056,
+      "grad_norm": 0.7745002508163452,
+      "learning_rate": 0.0001971941583282444,
+      "loss": 0.8895,
+      "step": 1406
+    },
+    {
+      "epoch": 0.10346348996249724,
+      "grad_norm": 1.6250338554382324,
+      "learning_rate": 0.00019718855347312808,
+      "loss": 0.8906,
+      "step": 1407
+    },
+    {
+      "epoch": 0.10353702478123392,
+      "grad_norm": 1.0553187131881714,
+      "learning_rate": 0.0001971829431053744,
+      "loss": 0.918,
+      "step": 1408
+    },
+    {
+      "epoch": 0.10361055959997059,
+      "grad_norm": 1.7217857837677002,
+      "learning_rate": 0.00019717732722530166,
+      "loss": 0.9183,
+      "step": 1409
+    },
+    {
+      "epoch": 0.10368409441870725,
+      "grad_norm": 1.7668135166168213,
+      "learning_rate": 0.00019717170583322834,
+      "loss": 0.9011,
+      "step": 1410
+    },
+    {
+      "epoch": 0.10375762923744393,
+      "grad_norm": 0.9553186297416687,
+      "learning_rate": 0.00019716607892947337,
+      "loss": 0.8437,
+      "step": 1411
+    },
+    {
+      "epoch": 0.1038311640561806,
+      "grad_norm": 0.9647707343101501,
+      "learning_rate": 0.0001971604465143558,
+      "loss": 0.9168,
+      "step": 1412
+    },
+    {
+      "epoch": 0.10390469887491727,
+      "grad_norm": 0.9313908815383911,
+      "learning_rate": 0.00019715480858819518,
+      "loss": 0.8806,
+      "step": 1413
+    },
+    {
+      "epoch": 0.10397823369365394,
+      "grad_norm": 2.29312801361084,
+      "learning_rate": 0.00019714916515131128,
+      "loss": 0.9158,
+      "step": 1414
+    },
+    {
+      "epoch": 0.10405176851239062,
+      "grad_norm": 0.8423621654510498,
+      "learning_rate": 0.00019714351620402423,
+      "loss": 0.8717,
+      "step": 1415
+    },
+    {
+      "epoch": 0.1041253033311273,
+      "grad_norm": 1.9242128133773804,
+      "learning_rate": 0.0001971378617466544,
+      "loss": 0.8918,
+      "step": 1416
+    },
+    {
+      "epoch": 0.10419883814986396,
+      "grad_norm": 0.9675571322441101,
+      "learning_rate": 0.00019713220177952257,
+      "loss": 0.8891,
+      "step": 1417
+    },
+    {
+      "epoch": 0.10427237296860063,
+      "grad_norm": 1.4491127729415894,
+      "learning_rate": 0.00019712653630294972,
+      "loss": 0.873,
+      "step": 1418
+    },
+    {
+      "epoch": 0.10434590778733731,
+      "grad_norm": 1.1429775953292847,
+      "learning_rate": 0.00019712086531725723,
+      "loss": 0.9457,
+      "step": 1419
+    },
+    {
+      "epoch": 0.10441944260607397,
+      "grad_norm": 1.693390965461731,
+      "learning_rate": 0.00019711518882276676,
+      "loss": 0.8665,
+      "step": 1420
+    },
+    {
+      "epoch": 0.10449297742481065,
+      "grad_norm": 1.0453791618347168,
+      "learning_rate": 0.0001971095068198003,
+      "loss": 0.87,
+      "step": 1421
+    },
+    {
+      "epoch": 0.10456651224354732,
+      "grad_norm": 1.204046607017517,
+      "learning_rate": 0.0001971038193086801,
+      "loss": 0.8792,
+      "step": 1422
+    },
+    {
+      "epoch": 0.104640047062284,
+      "grad_norm": 0.9971487522125244,
+      "learning_rate": 0.00019709812628972883,
+      "loss": 0.8394,
+      "step": 1423
+    },
+    {
+      "epoch": 0.10471358188102066,
+      "grad_norm": 0.9890929460525513,
+      "learning_rate": 0.00019709242776326936,
+      "loss": 0.9098,
+      "step": 1424
+    },
+    {
+      "epoch": 0.10478711669975734,
+      "grad_norm": 1.2027051448822021,
+      "learning_rate": 0.0001970867237296249,
+      "loss": 0.8724,
+      "step": 1425
+    },
+    {
+      "epoch": 0.10486065151849401,
+      "grad_norm": 1.1249390840530396,
+      "learning_rate": 0.00019708101418911903,
+      "loss": 0.8391,
+      "step": 1426
+    },
+    {
+      "epoch": 0.10493418633723067,
+      "grad_norm": 0.8577402234077454,
+      "learning_rate": 0.0001970752991420756,
+      "loss": 0.8798,
+      "step": 1427
+    },
+    {
+      "epoch": 0.10500772115596735,
+      "grad_norm": 0.8483748435974121,
+      "learning_rate": 0.00019706957858881872,
+      "loss": 0.8856,
+      "step": 1428
+    },
+    {
+      "epoch": 0.10508125597470402,
+      "grad_norm": 0.9420272707939148,
+      "learning_rate": 0.0001970638525296729,
+      "loss": 0.9023,
+      "step": 1429
+    },
+    {
+      "epoch": 0.1051547907934407,
+      "grad_norm": 1.0620026588439941,
+      "learning_rate": 0.00019705812096496294,
+      "loss": 0.8791,
+      "step": 1430
+    },
+    {
+      "epoch": 0.10522832561217736,
+      "grad_norm": 1.0137559175491333,
+      "learning_rate": 0.00019705238389501392,
+      "loss": 0.8765,
+      "step": 1431
+    },
+    {
+      "epoch": 0.10530186043091404,
+      "grad_norm": 0.7742050886154175,
+      "learning_rate": 0.00019704664132015126,
+      "loss": 0.8798,
+      "step": 1432
+    },
+    {
+      "epoch": 0.10537539524965071,
+      "grad_norm": 0.6888828873634338,
+      "learning_rate": 0.00019704089324070072,
+      "loss": 0.9084,
+      "step": 1433
+    },
+    {
+      "epoch": 0.10544893006838738,
+      "grad_norm": 0.7414498925209045,
+      "learning_rate": 0.00019703513965698828,
+      "loss": 0.8688,
+      "step": 1434
+    },
+    {
+      "epoch": 0.10552246488712405,
+      "grad_norm": 0.8080383539199829,
+      "learning_rate": 0.00019702938056934032,
+      "loss": 0.8705,
+      "step": 1435
+    },
+    {
+      "epoch": 0.10559599970586073,
+      "grad_norm": 1.3728694915771484,
+      "learning_rate": 0.0001970236159780835,
+      "loss": 0.9001,
+      "step": 1436
+    },
+    {
+      "epoch": 0.1056695345245974,
+      "grad_norm": 1.370190978050232,
+      "learning_rate": 0.00019701784588354478,
+      "loss": 0.9226,
+      "step": 1437
+    },
+    {
+      "epoch": 0.10574306934333406,
+      "grad_norm": 0.8220975399017334,
+      "learning_rate": 0.00019701207028605144,
+      "loss": 0.892,
+      "step": 1438
+    },
+    {
+      "epoch": 0.10581660416207074,
+      "grad_norm": 0.5943233370780945,
+      "learning_rate": 0.00019700628918593112,
+      "loss": 0.8822,
+      "step": 1439
+    },
+    {
+      "epoch": 0.10589013898080742,
+      "grad_norm": 0.7888864874839783,
+      "learning_rate": 0.0001970005025835117,
+      "loss": 0.9109,
+      "step": 1440
+    },
+    {
+      "epoch": 0.10596367379954408,
+      "grad_norm": 0.9658022522926331,
+      "learning_rate": 0.0001969947104791214,
+      "loss": 0.906,
+      "step": 1441
+    },
+    {
+      "epoch": 0.10603720861828075,
+      "grad_norm": 2.0499372482299805,
+      "learning_rate": 0.00019698891287308872,
+      "loss": 0.8802,
+      "step": 1442
+    },
+    {
+      "epoch": 0.10611074343701743,
+      "grad_norm": 0.7951045632362366,
+      "learning_rate": 0.00019698310976574258,
+      "loss": 0.877,
+      "step": 1443
+    },
+    {
+      "epoch": 0.1061842782557541,
+      "grad_norm": 2.9741456508636475,
+      "learning_rate": 0.0001969773011574121,
+      "loss": 0.9043,
+      "step": 1444
+    },
+    {
+      "epoch": 0.10625781307449077,
+      "grad_norm": 1.1555216312408447,
+      "learning_rate": 0.00019697148704842676,
+      "loss": 0.8703,
+      "step": 1445
+    },
+    {
+      "epoch": 0.10633134789322744,
+      "grad_norm": 3.377225160598755,
+      "learning_rate": 0.00019696566743911635,
+      "loss": 0.9096,
+      "step": 1446
+    },
+    {
+      "epoch": 0.10640488271196412,
+      "grad_norm": 2.921715497970581,
+      "learning_rate": 0.00019695984232981093,
+      "loss": 0.9328,
+      "step": 1447
+    },
+    {
+      "epoch": 0.10647841753070078,
+      "grad_norm": 1.6621732711791992,
+      "learning_rate": 0.0001969540117208409,
+      "loss": 0.904,
+      "step": 1448
+    },
+    {
+      "epoch": 0.10655195234943746,
+      "grad_norm": 2.5339465141296387,
+      "learning_rate": 0.00019694817561253706,
+      "loss": 0.9378,
+      "step": 1449
+    },
+    {
+      "epoch": 0.10662548716817413,
+      "grad_norm": 2.1396477222442627,
+      "learning_rate": 0.00019694233400523037,
+      "loss": 0.9463,
+      "step": 1450
+    },
+    {
+      "epoch": 0.10669902198691081,
+      "grad_norm": 1.9694663286209106,
+      "learning_rate": 0.00019693648689925216,
+      "loss": 0.8984,
+      "step": 1451
+    },
+    {
+      "epoch": 0.10677255680564747,
+      "grad_norm": 2.012608528137207,
+      "learning_rate": 0.00019693063429493415,
+      "loss": 0.9127,
+      "step": 1452
+    },
+    {
+      "epoch": 0.10684609162438415,
+      "grad_norm": 1.4470584392547607,
+      "learning_rate": 0.00019692477619260825,
+      "loss": 0.8715,
+      "step": 1453
+    },
+    {
+      "epoch": 0.10691962644312082,
+      "grad_norm": 1.6820907592773438,
+      "learning_rate": 0.00019691891259260676,
+      "loss": 0.8817,
+      "step": 1454
+    },
+    {
+      "epoch": 0.10699316126185748,
+      "grad_norm": 1.3522907495498657,
+      "learning_rate": 0.00019691304349526225,
+      "loss": 0.9168,
+      "step": 1455
+    },
+    {
+      "epoch": 0.10706669608059416,
+      "grad_norm": 1.3091238737106323,
+      "learning_rate": 0.00019690716890090765,
+      "loss": 0.8896,
+      "step": 1456
+    },
+    {
+      "epoch": 0.10714023089933083,
+      "grad_norm": 1.399861454963684,
+      "learning_rate": 0.00019690128880987614,
+      "loss": 0.9126,
+      "step": 1457
+    },
+    {
+      "epoch": 0.10721376571806751,
+      "grad_norm": 1.0162453651428223,
+      "learning_rate": 0.0001968954032225013,
+      "loss": 0.9128,
+      "step": 1458
+    },
+    {
+      "epoch": 0.10728730053680417,
+      "grad_norm": 1.7608469724655151,
+      "learning_rate": 0.0001968895121391169,
+      "loss": 0.9003,
+      "step": 1459
+    },
+    {
+      "epoch": 0.10736083535554085,
+      "grad_norm": 1.2385790348052979,
+      "learning_rate": 0.00019688361556005716,
+      "loss": 0.8711,
+      "step": 1460
+    },
+    {
+      "epoch": 0.10743437017427752,
+      "grad_norm": 1.1409460306167603,
+      "learning_rate": 0.00019687771348565644,
+      "loss": 0.8997,
+      "step": 1461
+    },
+    {
+      "epoch": 0.10750790499301419,
+      "grad_norm": 1.028370976448059,
+      "learning_rate": 0.00019687180591624964,
+      "loss": 0.9253,
+      "step": 1462
+    },
+    {
+      "epoch": 0.10758143981175086,
+      "grad_norm": 1.0634444952011108,
+      "learning_rate": 0.00019686589285217173,
+      "loss": 0.9038,
+      "step": 1463
+    },
+    {
+      "epoch": 0.10765497463048754,
+      "grad_norm": 1.95552396774292,
+      "learning_rate": 0.00019685997429375816,
+      "loss": 0.9527,
+      "step": 1464
+    },
+    {
+      "epoch": 0.10772850944922421,
+      "grad_norm": 1.2820218801498413,
+      "learning_rate": 0.00019685405024134464,
+      "loss": 0.9184,
+      "step": 1465
+    },
+    {
+      "epoch": 0.10780204426796088,
+      "grad_norm": 2.303712844848633,
+      "learning_rate": 0.00019684812069526716,
+      "loss": 0.9058,
+      "step": 1466
+    },
+    {
+      "epoch": 0.10787557908669755,
+      "grad_norm": 1.5525286197662354,
+      "learning_rate": 0.00019684218565586207,
+      "loss": 0.8697,
+      "step": 1467
+    },
+    {
+      "epoch": 0.10794911390543423,
+      "grad_norm": 1.6222412586212158,
+      "learning_rate": 0.000196836245123466,
+      "loss": 0.8897,
+      "step": 1468
+    },
+    {
+      "epoch": 0.10802264872417089,
+      "grad_norm": 1.1414673328399658,
+      "learning_rate": 0.00019683029909841593,
+      "loss": 0.9361,
+      "step": 1469
+    },
+    {
+      "epoch": 0.10809618354290756,
+      "grad_norm": 1.8357977867126465,
+      "learning_rate": 0.00019682434758104908,
+      "loss": 0.9248,
+      "step": 1470
+    },
+    {
+      "epoch": 0.10816971836164424,
+      "grad_norm": 0.9569761753082275,
+      "learning_rate": 0.00019681839057170307,
+      "loss": 0.8759,
+      "step": 1471
+    },
+    {
+      "epoch": 0.10824325318038092,
+      "grad_norm": 1.6061382293701172,
+      "learning_rate": 0.0001968124280707158,
+      "loss": 0.9292,
+      "step": 1472
+    },
+    {
+      "epoch": 0.10831678799911758,
+      "grad_norm": 1.252079963684082,
+      "learning_rate": 0.00019680646007842542,
+      "loss": 0.9052,
+      "step": 1473
+    },
+    {
+      "epoch": 0.10839032281785425,
+      "grad_norm": 1.258362889289856,
+      "learning_rate": 0.00019680048659517042,
+      "loss": 0.9269,
+      "step": 1474
+    },
+    {
+      "epoch": 0.10846385763659093,
+      "grad_norm": 1.595488429069519,
+      "learning_rate": 0.00019679450762128967,
+      "loss": 0.9062,
+      "step": 1475
+    },
+    {
+      "epoch": 0.10853739245532759,
+      "grad_norm": 1.5255414247512817,
+      "learning_rate": 0.00019678852315712235,
+      "loss": 0.9143,
+      "step": 1476
+    },
+    {
+      "epoch": 0.10861092727406427,
+      "grad_norm": 1.8744109869003296,
+      "learning_rate": 0.00019678253320300782,
+      "loss": 0.9059,
+      "step": 1477
+    },
+    {
+      "epoch": 0.10868446209280094,
+      "grad_norm": 1.8928077220916748,
+      "learning_rate": 0.00019677653775928587,
+      "loss": 0.8759,
+      "step": 1478
+    },
+    {
+      "epoch": 0.10875799691153762,
+      "grad_norm": 1.3073502779006958,
+      "learning_rate": 0.00019677053682629655,
+      "loss": 0.9109,
+      "step": 1479
+    },
+    {
+      "epoch": 0.10883153173027428,
+      "grad_norm": 1.5995789766311646,
+      "learning_rate": 0.00019676453040438026,
+      "loss": 0.9038,
+      "step": 1480
+    },
+    {
+      "epoch": 0.10890506654901096,
+      "grad_norm": 1.202123761177063,
+      "learning_rate": 0.0001967585184938777,
+      "loss": 0.9408,
+      "step": 1481
+    },
+    {
+      "epoch": 0.10897860136774763,
+      "grad_norm": 1.2111676931381226,
+      "learning_rate": 0.00019675250109512987,
+      "loss": 0.9133,
+      "step": 1482
+    },
+    {
+      "epoch": 0.1090521361864843,
+      "grad_norm": 1.3049373626708984,
+      "learning_rate": 0.00019674647820847805,
+      "loss": 0.9255,
+      "step": 1483
+    },
+    {
+      "epoch": 0.10912567100522097,
+      "grad_norm": 1.2724905014038086,
+      "learning_rate": 0.0001967404498342639,
+      "loss": 0.8926,
+      "step": 1484
+    },
+    {
+      "epoch": 0.10919920582395765,
+      "grad_norm": 1.8511401414871216,
+      "learning_rate": 0.00019673441597282932,
+      "loss": 0.8914,
+      "step": 1485
+    },
+    {
+      "epoch": 0.10927274064269432,
+      "grad_norm": 1.6952948570251465,
+      "learning_rate": 0.00019672837662451659,
+      "loss": 0.9257,
+      "step": 1486
+    },
+    {
+      "epoch": 0.10934627546143098,
+      "grad_norm": 1.093451976776123,
+      "learning_rate": 0.00019672233178966825,
+      "loss": 0.8947,
+      "step": 1487
+    },
+    {
+      "epoch": 0.10941981028016766,
+      "grad_norm": 1.059073805809021,
+      "learning_rate": 0.0001967162814686272,
+      "loss": 0.885,
+      "step": 1488
+    },
+    {
+      "epoch": 0.10949334509890433,
+      "grad_norm": 1.3910630941390991,
+      "learning_rate": 0.00019671022566173656,
+      "loss": 0.8935,
+      "step": 1489
+    },
+    {
+      "epoch": 0.109566879917641,
+      "grad_norm": 0.8187393546104431,
+      "learning_rate": 0.00019670416436933988,
+      "loss": 0.9424,
+      "step": 1490
+    },
+    {
+      "epoch": 0.10964041473637767,
+      "grad_norm": 0.7321296334266663,
+      "learning_rate": 0.00019669809759178094,
+      "loss": 0.8439,
+      "step": 1491
+    },
+    {
+      "epoch": 0.10971394955511435,
+      "grad_norm": 0.8418588042259216,
+      "learning_rate": 0.00019669202532940382,
+      "loss": 0.8782,
+      "step": 1492
+    },
+    {
+      "epoch": 0.10978748437385102,
+      "grad_norm": 0.6813521981239319,
+      "learning_rate": 0.00019668594758255302,
+      "loss": 0.8912,
+      "step": 1493
+    },
+    {
+      "epoch": 0.10986101919258769,
+      "grad_norm": 0.8613452315330505,
+      "learning_rate": 0.00019667986435157323,
+      "loss": 0.9393,
+      "step": 1494
+    },
+    {
+      "epoch": 0.10993455401132436,
+      "grad_norm": 0.8516898155212402,
+      "learning_rate": 0.00019667377563680952,
+      "loss": 0.8853,
+      "step": 1495
+    },
+    {
+      "epoch": 0.11000808883006104,
+      "grad_norm": 0.6053500175476074,
+      "learning_rate": 0.0001966676814386072,
+      "loss": 0.8948,
+      "step": 1496
+    },
+    {
+      "epoch": 0.1100816236487977,
+      "grad_norm": 0.7933889627456665,
+      "learning_rate": 0.000196661581757312,
+      "loss": 0.8866,
+      "step": 1497
+    },
+    {
+      "epoch": 0.11015515846753438,
+      "grad_norm": 0.5835139155387878,
+      "learning_rate": 0.00019665547659326985,
+      "loss": 0.8932,
+      "step": 1498
+    },
+    {
+      "epoch": 0.11022869328627105,
+      "grad_norm": 0.6790846586227417,
+      "learning_rate": 0.00019664936594682706,
+      "loss": 0.8884,
+      "step": 1499
+    },
+    {
+      "epoch": 0.11030222810500773,
+      "grad_norm": 1.2909609079360962,
+      "learning_rate": 0.00019664324981833026,
+      "loss": 0.8752,
+      "step": 1500
+    },
+    {
+      "epoch": 0.11037576292374439,
+      "grad_norm": 1.4551030397415161,
+      "learning_rate": 0.00019663712820812632,
+      "loss": 0.8864,
+      "step": 1501
+    },
+    {
+      "epoch": 0.11044929774248106,
+      "grad_norm": 0.6914474368095398,
+      "learning_rate": 0.0001966310011165625,
+      "loss": 0.9032,
+      "step": 1502
+    },
+    {
+      "epoch": 0.11052283256121774,
+      "grad_norm": 1.1030083894729614,
+      "learning_rate": 0.0001966248685439863,
+      "loss": 0.8983,
+      "step": 1503
+    },
+    {
+      "epoch": 0.1105963673799544,
+      "grad_norm": 1.0917330980300903,
+      "learning_rate": 0.00019661873049074558,
+      "loss": 0.9132,
+      "step": 1504
+    },
+    {
+      "epoch": 0.11066990219869108,
+      "grad_norm": 1.1616206169128418,
+      "learning_rate": 0.0001966125869571885,
+      "loss": 0.8984,
+      "step": 1505
+    },
+    {
+      "epoch": 0.11074343701742775,
+      "grad_norm": 0.9483665823936462,
+      "learning_rate": 0.00019660643794366353,
+      "loss": 0.8749,
+      "step": 1506
+    },
+    {
+      "epoch": 0.11081697183616443,
+      "grad_norm": 1.1939057111740112,
+      "learning_rate": 0.00019660028345051947,
+      "loss": 0.8742,
+      "step": 1507
+    },
+    {
+      "epoch": 0.11089050665490109,
+      "grad_norm": 0.765306830406189,
+      "learning_rate": 0.00019659412347810537,
+      "loss": 0.8748,
+      "step": 1508
+    },
+    {
+      "epoch": 0.11096404147363777,
+      "grad_norm": 1.448372721672058,
+      "learning_rate": 0.00019658795802677065,
+      "loss": 0.8651,
+      "step": 1509
+    },
+    {
+      "epoch": 0.11103757629237444,
+      "grad_norm": 1.3845897912979126,
+      "learning_rate": 0.00019658178709686501,
+      "loss": 0.9047,
+      "step": 1510
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 0.989128589630127,
+      "learning_rate": 0.00019657561068873848,
+      "loss": 0.9156,
+      "step": 1511
+    },
+    {
+      "epoch": 0.11118464592984778,
+      "grad_norm": 2.0974552631378174,
+      "learning_rate": 0.0001965694288027414,
+      "loss": 0.891,
+      "step": 1512
+    },
+    {
+      "epoch": 0.11125818074858446,
+      "grad_norm": 1.4651072025299072,
+      "learning_rate": 0.0001965632414392244,
+      "loss": 0.9462,
+      "step": 1513
+    },
+    {
+      "epoch": 0.11133171556732113,
+      "grad_norm": 1.8048090934753418,
+      "learning_rate": 0.00019655704859853845,
+      "loss": 0.9296,
+      "step": 1514
+    },
+    {
+      "epoch": 0.1114052503860578,
+      "grad_norm": 1.4310939311981201,
+      "learning_rate": 0.00019655085028103477,
+      "loss": 0.8883,
+      "step": 1515
+    },
+    {
+      "epoch": 0.11147878520479447,
+      "grad_norm": 1.8090015649795532,
+      "learning_rate": 0.00019654464648706502,
+      "loss": 0.8898,
+      "step": 1516
+    },
+    {
+      "epoch": 0.11155232002353115,
+      "grad_norm": 1.246786117553711,
+      "learning_rate": 0.00019653843721698097,
+      "loss": 0.8802,
+      "step": 1517
+    },
+    {
+      "epoch": 0.11162585484226781,
+      "grad_norm": 1.3009222745895386,
+      "learning_rate": 0.0001965322224711349,
+      "loss": 0.9335,
+      "step": 1518
+    },
+    {
+      "epoch": 0.11169938966100448,
+      "grad_norm": 1.3562233448028564,
+      "learning_rate": 0.00019652600224987933,
+      "loss": 0.8592,
+      "step": 1519
+    },
+    {
+      "epoch": 0.11177292447974116,
+      "grad_norm": 0.7404664158821106,
+      "learning_rate": 0.00019651977655356704,
+      "loss": 0.8676,
+      "step": 1520
+    },
+    {
+      "epoch": 0.11184645929847783,
+      "grad_norm": 0.8865517377853394,
+      "learning_rate": 0.0001965135453825511,
+      "loss": 0.9154,
+      "step": 1521
+    },
+    {
+      "epoch": 0.1119199941172145,
+      "grad_norm": 1.5888679027557373,
+      "learning_rate": 0.00019650730873718508,
+      "loss": 0.8894,
+      "step": 1522
+    },
+    {
+      "epoch": 0.11199352893595117,
+      "grad_norm": 0.9649756550788879,
+      "learning_rate": 0.00019650106661782264,
+      "loss": 0.9054,
+      "step": 1523
+    },
+    {
+      "epoch": 0.11206706375468785,
+      "grad_norm": 1.5268908739089966,
+      "learning_rate": 0.00019649481902481785,
+      "loss": 0.8751,
+      "step": 1524
+    },
+    {
+      "epoch": 0.11214059857342451,
+      "grad_norm": 1.2616695165634155,
+      "learning_rate": 0.00019648856595852511,
+      "loss": 0.8677,
+      "step": 1525
+    },
+    {
+      "epoch": 0.11221413339216119,
+      "grad_norm": 1.6797529458999634,
+      "learning_rate": 0.00019648230741929905,
+      "loss": 0.867,
+      "step": 1526
+    },
+    {
+      "epoch": 0.11228766821089786,
+      "grad_norm": 1.1374331712722778,
+      "learning_rate": 0.00019647604340749473,
+      "loss": 0.8787,
+      "step": 1527
+    },
+    {
+      "epoch": 0.11236120302963454,
+      "grad_norm": 1.4214528799057007,
+      "learning_rate": 0.00019646977392346736,
+      "loss": 0.8769,
+      "step": 1528
+    },
+    {
+      "epoch": 0.1124347378483712,
+      "grad_norm": 1.104591727256775,
+      "learning_rate": 0.00019646349896757266,
+      "loss": 0.8735,
+      "step": 1529
+    },
+    {
+      "epoch": 0.11250827266710788,
+      "grad_norm": 0.7824739217758179,
+      "learning_rate": 0.00019645721854016647,
+      "loss": 0.8383,
+      "step": 1530
+    },
+    {
+      "epoch": 0.11258180748584455,
+      "grad_norm": 1.491888403892517,
+      "learning_rate": 0.00019645093264160506,
+      "loss": 0.8788,
+      "step": 1531
+    },
+    {
+      "epoch": 0.11265534230458121,
+      "grad_norm": 0.7540958523750305,
+      "learning_rate": 0.00019644464127224496,
+      "loss": 0.8698,
+      "step": 1532
+    },
+    {
+      "epoch": 0.11272887712331789,
+      "grad_norm": 0.9714244604110718,
+      "learning_rate": 0.000196438344432443,
+      "loss": 0.8618,
+      "step": 1533
+    },
+    {
+      "epoch": 0.11280241194205456,
+      "grad_norm": 1.9782365560531616,
+      "learning_rate": 0.00019643204212255638,
+      "loss": 0.8816,
+      "step": 1534
+    },
+    {
+      "epoch": 0.11287594676079124,
+      "grad_norm": 0.969748854637146,
+      "learning_rate": 0.0001964257343429426,
+      "loss": 0.8969,
+      "step": 1535
+    },
+    {
+      "epoch": 0.1129494815795279,
+      "grad_norm": 2.685028076171875,
+      "learning_rate": 0.00019641942109395938,
+      "loss": 0.9354,
+      "step": 1536
+    },
+    {
+      "epoch": 0.11302301639826458,
+      "grad_norm": 1.7907739877700806,
+      "learning_rate": 0.00019641310237596482,
+      "loss": 0.9036,
+      "step": 1537
+    },
+    {
+      "epoch": 0.11309655121700125,
+      "grad_norm": 1.9487887620925903,
+      "learning_rate": 0.00019640677818931734,
+      "loss": 0.8963,
+      "step": 1538
+    },
+    {
+      "epoch": 0.11317008603573792,
+      "grad_norm": 1.5944583415985107,
+      "learning_rate": 0.0001964004485343757,
+      "loss": 0.8983,
+      "step": 1539
+    },
+    {
+      "epoch": 0.11324362085447459,
+      "grad_norm": 2.2190825939178467,
+      "learning_rate": 0.00019639411341149888,
+      "loss": 0.8927,
+      "step": 1540
+    },
+    {
+      "epoch": 0.11331715567321127,
+      "grad_norm": 1.519472599029541,
+      "learning_rate": 0.0001963877728210462,
+      "loss": 0.874,
+      "step": 1541
+    },
+    {
+      "epoch": 0.11339069049194794,
+      "grad_norm": 1.7229207754135132,
+      "learning_rate": 0.0001963814267633773,
+      "loss": 0.9169,
+      "step": 1542
+    },
+    {
+      "epoch": 0.1134642253106846,
+      "grad_norm": 1.249028205871582,
+      "learning_rate": 0.0001963750752388522,
+      "loss": 0.866,
+      "step": 1543
+    },
+    {
+      "epoch": 0.11353776012942128,
+      "grad_norm": 1.8265897035598755,
+      "learning_rate": 0.0001963687182478311,
+      "loss": 0.8855,
+      "step": 1544
+    },
+    {
+      "epoch": 0.11361129494815796,
+      "grad_norm": 1.1729581356048584,
+      "learning_rate": 0.0001963623557906746,
+      "loss": 0.8302,
+      "step": 1545
+    },
+    {
+      "epoch": 0.11368482976689462,
+      "grad_norm": 2.019524335861206,
+      "learning_rate": 0.00019635598786774358,
+      "loss": 0.8689,
+      "step": 1546
+    },
+    {
+      "epoch": 0.1137583645856313,
+      "grad_norm": 1.2850593328475952,
+      "learning_rate": 0.00019634961447939926,
+      "loss": 0.8524,
+      "step": 1547
+    },
+    {
+      "epoch": 0.11383189940436797,
+      "grad_norm": 2.2539286613464355,
+      "learning_rate": 0.00019634323562600308,
+      "loss": 0.8723,
+      "step": 1548
+    },
+    {
+      "epoch": 0.11390543422310465,
+      "grad_norm": 1.8272838592529297,
+      "learning_rate": 0.00019633685130791698,
+      "loss": 0.8789,
+      "step": 1549
+    },
+    {
+      "epoch": 0.11397896904184131,
+      "grad_norm": 1.567038655281067,
+      "learning_rate": 0.00019633046152550294,
+      "loss": 0.8429,
+      "step": 1550
+    },
+    {
+      "epoch": 0.11405250386057798,
+      "grad_norm": 1.5974955558776855,
+      "learning_rate": 0.00019632406627912346,
+      "loss": 0.8813,
+      "step": 1551
+    },
+    {
+      "epoch": 0.11412603867931466,
+      "grad_norm": 1.0610891580581665,
+      "learning_rate": 0.00019631766556914133,
+      "loss": 0.8455,
+      "step": 1552
+    },
+    {
+      "epoch": 0.11419957349805132,
+      "grad_norm": 1.2768560647964478,
+      "learning_rate": 0.00019631125939591955,
+      "loss": 0.8491,
+      "step": 1553
+    },
+    {
+      "epoch": 0.114273108316788,
+      "grad_norm": 1.1397154331207275,
+      "learning_rate": 0.00019630484775982147,
+      "loss": 0.8782,
+      "step": 1554
+    },
+    {
+      "epoch": 0.11434664313552467,
+      "grad_norm": 1.0940841436386108,
+      "learning_rate": 0.00019629843066121082,
+      "loss": 0.8411,
+      "step": 1555
+    },
+    {
+      "epoch": 0.11442017795426135,
+      "grad_norm": 1.1627635955810547,
+      "learning_rate": 0.00019629200810045153,
+      "loss": 0.8663,
+      "step": 1556
+    },
+    {
+      "epoch": 0.11449371277299801,
+      "grad_norm": 0.9046681523323059,
+      "learning_rate": 0.00019628558007790794,
+      "loss": 0.844,
+      "step": 1557
+    },
+    {
+      "epoch": 0.11456724759173469,
+      "grad_norm": 1.3891067504882812,
+      "learning_rate": 0.0001962791465939446,
+      "loss": 0.8667,
+      "step": 1558
+    },
+    {
+      "epoch": 0.11464078241047136,
+      "grad_norm": 1.036232590675354,
+      "learning_rate": 0.00019627270764892648,
+      "loss": 0.8357,
+      "step": 1559
+    },
+    {
+      "epoch": 0.11471431722920802,
+      "grad_norm": 1.9901095628738403,
+      "learning_rate": 0.0001962662632432188,
+      "loss": 0.8899,
+      "step": 1560
+    },
+    {
+      "epoch": 0.1147878520479447,
+      "grad_norm": 1.7478631734848022,
+      "learning_rate": 0.00019625981337718703,
+      "loss": 0.8477,
+      "step": 1561
+    },
+    {
+      "epoch": 0.11486138686668138,
+      "grad_norm": 1.3817931413650513,
+      "learning_rate": 0.0001962533580511971,
+      "loss": 0.8601,
+      "step": 1562
+    },
+    {
+      "epoch": 0.11493492168541805,
+      "grad_norm": 1.2834928035736084,
+      "learning_rate": 0.00019624689726561512,
+      "loss": 0.8676,
+      "step": 1563
+    },
+    {
+      "epoch": 0.11500845650415471,
+      "grad_norm": 1.6230931282043457,
+      "learning_rate": 0.0001962404310208075,
+      "loss": 0.8608,
+      "step": 1564
+    },
+    {
+      "epoch": 0.11508199132289139,
+      "grad_norm": 1.0849311351776123,
+      "learning_rate": 0.00019623395931714114,
+      "loss": 0.8356,
+      "step": 1565
+    },
+    {
+      "epoch": 0.11515552614162806,
+      "grad_norm": 1.648608684539795,
+      "learning_rate": 0.000196227482154983,
+      "loss": 0.8768,
+      "step": 1566
+    },
+    {
+      "epoch": 0.11522906096036473,
+      "grad_norm": 1.2513844966888428,
+      "learning_rate": 0.00019622099953470055,
+      "loss": 0.8311,
+      "step": 1567
+    },
+    {
+      "epoch": 0.1153025957791014,
+      "grad_norm": 1.7076654434204102,
+      "learning_rate": 0.00019621451145666142,
+      "loss": 0.8541,
+      "step": 1568
+    },
+    {
+      "epoch": 0.11537613059783808,
+      "grad_norm": 1.4567228555679321,
+      "learning_rate": 0.00019620801792123369,
+      "loss": 0.8383,
+      "step": 1569
+    },
+    {
+      "epoch": 0.11544966541657475,
+      "grad_norm": 1.1875765323638916,
+      "learning_rate": 0.00019620151892878567,
+      "loss": 0.836,
+      "step": 1570
+    },
+    {
+      "epoch": 0.11552320023531142,
+      "grad_norm": 0.9912257194519043,
+      "learning_rate": 0.00019619501447968592,
+      "loss": 0.8647,
+      "step": 1571
+    },
+    {
+      "epoch": 0.11559673505404809,
+      "grad_norm": 1.6818130016326904,
+      "learning_rate": 0.00019618850457430345,
+      "loss": 0.8314,
+      "step": 1572
+    },
+    {
+      "epoch": 0.11567026987278477,
+      "grad_norm": 1.0181254148483276,
+      "learning_rate": 0.0001961819892130075,
+      "loss": 0.8317,
+      "step": 1573
+    },
+    {
+      "epoch": 0.11574380469152143,
+      "grad_norm": 1.4218688011169434,
+      "learning_rate": 0.00019617546839616762,
+      "loss": 0.8781,
+      "step": 1574
+    },
+    {
+      "epoch": 0.1158173395102581,
+      "grad_norm": 1.1426608562469482,
+      "learning_rate": 0.00019616894212415366,
+      "loss": 0.8392,
+      "step": 1575
+    },
+    {
+      "epoch": 0.11589087432899478,
+      "grad_norm": 1.4509615898132324,
+      "learning_rate": 0.00019616241039733582,
+      "loss": 0.8814,
+      "step": 1576
+    },
+    {
+      "epoch": 0.11596440914773146,
+      "grad_norm": 1.2703207731246948,
+      "learning_rate": 0.00019615587321608459,
+      "loss": 0.8744,
+      "step": 1577
+    },
+    {
+      "epoch": 0.11603794396646812,
+      "grad_norm": 1.4068892002105713,
+      "learning_rate": 0.00019614933058077072,
+      "loss": 0.872,
+      "step": 1578
+    },
+    {
+      "epoch": 0.1161114787852048,
+      "grad_norm": 1.0297458171844482,
+      "learning_rate": 0.00019614278249176537,
+      "loss": 0.8636,
+      "step": 1579
+    },
+    {
+      "epoch": 0.11618501360394147,
+      "grad_norm": 1.5851505994796753,
+      "learning_rate": 0.00019613622894943994,
+      "loss": 0.8456,
+      "step": 1580
+    },
+    {
+      "epoch": 0.11625854842267813,
+      "grad_norm": 1.2108354568481445,
+      "learning_rate": 0.00019612966995416614,
+      "loss": 0.853,
+      "step": 1581
+    },
+    {
+      "epoch": 0.11633208324141481,
+      "grad_norm": 1.1483020782470703,
+      "learning_rate": 0.000196123105506316,
+      "loss": 0.8657,
+      "step": 1582
+    },
+    {
+      "epoch": 0.11640561806015148,
+      "grad_norm": 1.215212345123291,
+      "learning_rate": 0.00019611653560626187,
+      "loss": 0.8557,
+      "step": 1583
+    },
+    {
+      "epoch": 0.11647915287888816,
+      "grad_norm": 1.1399013996124268,
+      "learning_rate": 0.00019610996025437645,
+      "loss": 0.8476,
+      "step": 1584
+    },
+    {
+      "epoch": 0.11655268769762482,
+      "grad_norm": 1.4342464208602905,
+      "learning_rate": 0.00019610337945103262,
+      "loss": 0.8265,
+      "step": 1585
+    },
+    {
+      "epoch": 0.1166262225163615,
+      "grad_norm": 0.8745381236076355,
+      "learning_rate": 0.0001960967931966037,
+      "loss": 0.8379,
+      "step": 1586
+    },
+    {
+      "epoch": 0.11669975733509817,
+      "grad_norm": 1.7802222967147827,
+      "learning_rate": 0.00019609020149146322,
+      "loss": 0.8745,
+      "step": 1587
+    },
+    {
+      "epoch": 0.11677329215383483,
+      "grad_norm": 1.088546633720398,
+      "learning_rate": 0.00019608360433598514,
+      "loss": 0.8592,
+      "step": 1588
+    },
+    {
+      "epoch": 0.11684682697257151,
+      "grad_norm": 1.8353729248046875,
+      "learning_rate": 0.0001960770017305436,
+      "loss": 0.8513,
+      "step": 1589
+    },
+    {
+      "epoch": 0.11692036179130819,
+      "grad_norm": 1.5316413640975952,
+      "learning_rate": 0.00019607039367551315,
+      "loss": 0.8625,
+      "step": 1590
+    },
+    {
+      "epoch": 0.11699389661004486,
+      "grad_norm": 1.3382794857025146,
+      "learning_rate": 0.00019606378017126855,
+      "loss": 0.8446,
+      "step": 1591
+    },
+    {
+      "epoch": 0.11706743142878152,
+      "grad_norm": 1.0647205114364624,
+      "learning_rate": 0.000196057161218185,
+      "loss": 0.817,
+      "step": 1592
+    },
+    {
+      "epoch": 0.1171409662475182,
+      "grad_norm": 1.3138344287872314,
+      "learning_rate": 0.00019605053681663788,
+      "loss": 0.8361,
+      "step": 1593
+    },
+    {
+      "epoch": 0.11721450106625488,
+      "grad_norm": 0.979292094707489,
+      "learning_rate": 0.00019604390696700293,
+      "loss": 0.8173,
+      "step": 1594
+    },
+    {
+      "epoch": 0.11728803588499154,
+      "grad_norm": 1.6679821014404297,
+      "learning_rate": 0.0001960372716696562,
+      "loss": 0.8408,
+      "step": 1595
+    },
+    {
+      "epoch": 0.11736157070372821,
+      "grad_norm": 1.209737777709961,
+      "learning_rate": 0.0001960306309249741,
+      "loss": 0.8655,
+      "step": 1596
+    },
+    {
+      "epoch": 0.11743510552246489,
+      "grad_norm": 1.5650124549865723,
+      "learning_rate": 0.00019602398473333326,
+      "loss": 0.8241,
+      "step": 1597
+    },
+    {
+      "epoch": 0.11750864034120156,
+      "grad_norm": 1.5448863506317139,
+      "learning_rate": 0.00019601733309511067,
+      "loss": 0.8434,
+      "step": 1598
+    },
+    {
+      "epoch": 0.11758217515993823,
+      "grad_norm": 1.4208863973617554,
+      "learning_rate": 0.00019601067601068364,
+      "loss": 0.8891,
+      "step": 1599
+    },
+    {
+      "epoch": 0.1176557099786749,
+      "grad_norm": 1.3386183977127075,
+      "learning_rate": 0.0001960040134804297,
+      "loss": 0.8647,
+      "step": 1600
+    },
+    {
+      "epoch": 0.11772924479741158,
+      "grad_norm": 1.5525360107421875,
+      "learning_rate": 0.00019599734550472683,
+      "loss": 0.871,
+      "step": 1601
+    },
+    {
+      "epoch": 0.11780277961614824,
+      "grad_norm": 1.0561774969100952,
+      "learning_rate": 0.0001959906720839532,
+      "loss": 0.8221,
+      "step": 1602
+    },
+    {
+      "epoch": 0.11787631443488492,
+      "grad_norm": 1.6779098510742188,
+      "learning_rate": 0.00019598399321848732,
+      "loss": 0.8302,
+      "step": 1603
+    },
+    {
+      "epoch": 0.11794984925362159,
+      "grad_norm": 1.2319687604904175,
+      "learning_rate": 0.00019597730890870806,
+      "loss": 0.8425,
+      "step": 1604
+    },
+    {
+      "epoch": 0.11802338407235827,
+      "grad_norm": 1.669732928276062,
+      "learning_rate": 0.0001959706191549946,
+      "loss": 0.8573,
+      "step": 1605
+    },
+    {
+      "epoch": 0.11809691889109493,
+      "grad_norm": 1.2987724542617798,
+      "learning_rate": 0.00019596392395772628,
+      "loss": 0.8096,
+      "step": 1606
+    },
+    {
+      "epoch": 0.1181704537098316,
+      "grad_norm": 1.5951318740844727,
+      "learning_rate": 0.00019595722331728295,
+      "loss": 0.8336,
+      "step": 1607
+    },
+    {
+      "epoch": 0.11824398852856828,
+      "grad_norm": 1.214902400970459,
+      "learning_rate": 0.00019595051723404465,
+      "loss": 0.822,
+      "step": 1608
+    },
+    {
+      "epoch": 0.11831752334730494,
+      "grad_norm": 1.4320347309112549,
+      "learning_rate": 0.00019594380570839173,
+      "loss": 0.8404,
+      "step": 1609
+    },
+    {
+      "epoch": 0.11839105816604162,
+      "grad_norm": 1.0845303535461426,
+      "learning_rate": 0.00019593708874070492,
+      "loss": 0.8421,
+      "step": 1610
+    },
+    {
+      "epoch": 0.1184645929847783,
+      "grad_norm": 1.4716999530792236,
+      "learning_rate": 0.00019593036633136517,
+      "loss": 0.8571,
+      "step": 1611
+    },
+    {
+      "epoch": 0.11853812780351497,
+      "grad_norm": 0.9756578207015991,
+      "learning_rate": 0.0001959236384807538,
+      "loss": 0.8382,
+      "step": 1612
+    },
+    {
+      "epoch": 0.11861166262225163,
+      "grad_norm": 1.568200945854187,
+      "learning_rate": 0.00019591690518925244,
+      "loss": 0.8891,
+      "step": 1613
+    },
+    {
+      "epoch": 0.11868519744098831,
+      "grad_norm": 0.9394587278366089,
+      "learning_rate": 0.000195910166457243,
+      "loss": 0.81,
+      "step": 1614
+    },
+    {
+      "epoch": 0.11875873225972498,
+      "grad_norm": 1.3995333909988403,
+      "learning_rate": 0.00019590342228510766,
+      "loss": 0.8248,
+      "step": 1615
+    },
+    {
+      "epoch": 0.11883226707846165,
+      "grad_norm": 0.8745999336242676,
+      "learning_rate": 0.00019589667267322903,
+      "loss": 0.7954,
+      "step": 1616
+    },
+    {
+      "epoch": 0.11890580189719832,
+      "grad_norm": 1.382752776145935,
+      "learning_rate": 0.0001958899176219899,
+      "loss": 0.8404,
+      "step": 1617
+    },
+    {
+      "epoch": 0.118979336715935,
+      "grad_norm": 0.711279571056366,
+      "learning_rate": 0.0001958831571317735,
+      "loss": 0.808,
+      "step": 1618
+    },
+    {
+      "epoch": 0.11905287153467167,
+      "grad_norm": 1.498473048210144,
+      "learning_rate": 0.00019587639120296316,
+      "loss": 0.8328,
+      "step": 1619
+    },
+    {
+      "epoch": 0.11912640635340833,
+      "grad_norm": 0.8667972087860107,
+      "learning_rate": 0.0001958696198359428,
+      "loss": 0.8186,
+      "step": 1620
+    },
+    {
+      "epoch": 0.11919994117214501,
+      "grad_norm": 1.0260272026062012,
+      "learning_rate": 0.00019586284303109636,
+      "loss": 0.8325,
+      "step": 1621
+    },
+    {
+      "epoch": 0.11927347599088169,
+      "grad_norm": 1.1999397277832031,
+      "learning_rate": 0.00019585606078880834,
+      "loss": 0.8295,
+      "step": 1622
+    },
+    {
+      "epoch": 0.11934701080961835,
+      "grad_norm": 0.929821789264679,
+      "learning_rate": 0.0001958492731094634,
+      "loss": 0.8132,
+      "step": 1623
+    },
+    {
+      "epoch": 0.11942054562835502,
+      "grad_norm": 1.665909767150879,
+      "learning_rate": 0.0001958424799934465,
+      "loss": 0.8357,
+      "step": 1624
+    },
+    {
+      "epoch": 0.1194940804470917,
+      "grad_norm": 1.0747026205062866,
+      "learning_rate": 0.000195835681441143,
+      "loss": 0.8106,
+      "step": 1625
+    },
+    {
+      "epoch": 0.11956761526582838,
+      "grad_norm": 1.7649494409561157,
+      "learning_rate": 0.0001958288774529385,
+      "loss": 0.8282,
+      "step": 1626
+    },
+    {
+      "epoch": 0.11964115008456504,
+      "grad_norm": 1.8486988544464111,
+      "learning_rate": 0.00019582206802921896,
+      "loss": 0.8274,
+      "step": 1627
+    },
+    {
+      "epoch": 0.11971468490330171,
+      "grad_norm": 0.9450544714927673,
+      "learning_rate": 0.00019581525317037058,
+      "loss": 0.824,
+      "step": 1628
+    },
+    {
+      "epoch": 0.11978821972203839,
+      "grad_norm": 1.1530989408493042,
+      "learning_rate": 0.00019580843287677995,
+      "loss": 0.8052,
+      "step": 1629
+    },
+    {
+      "epoch": 0.11986175454077505,
+      "grad_norm": 0.9755038022994995,
+      "learning_rate": 0.00019580160714883387,
+      "loss": 0.8206,
+      "step": 1630
+    },
+    {
+      "epoch": 0.11993528935951173,
+      "grad_norm": 1.0235357284545898,
+      "learning_rate": 0.00019579477598691955,
+      "loss": 0.8002,
+      "step": 1631
+    },
+    {
+      "epoch": 0.1200088241782484,
+      "grad_norm": 1.4192819595336914,
+      "learning_rate": 0.0001957879393914244,
+      "loss": 0.8082,
+      "step": 1632
+    },
+    {
+      "epoch": 0.12008235899698508,
+      "grad_norm": 1.148432970046997,
+      "learning_rate": 0.00019578109736273627,
+      "loss": 0.8003,
+      "step": 1633
+    },
+    {
+      "epoch": 0.12015589381572174,
+      "grad_norm": 1.6818832159042358,
+      "learning_rate": 0.0001957742499012432,
+      "loss": 0.8179,
+      "step": 1634
+    },
+    {
+      "epoch": 0.12022942863445842,
+      "grad_norm": 1.609099268913269,
+      "learning_rate": 0.00019576739700733362,
+      "loss": 0.7847,
+      "step": 1635
+    },
+    {
+      "epoch": 0.12030296345319509,
+      "grad_norm": 1.0396372079849243,
+      "learning_rate": 0.00019576053868139624,
+      "loss": 0.8128,
+      "step": 1636
+    },
+    {
+      "epoch": 0.12037649827193175,
+      "grad_norm": 0.9049035310745239,
+      "learning_rate": 0.00019575367492382,
+      "loss": 0.8005,
+      "step": 1637
+    },
+    {
+      "epoch": 0.12045003309066843,
+      "grad_norm": 1.2967690229415894,
+      "learning_rate": 0.00019574680573499427,
+      "loss": 0.79,
+      "step": 1638
+    },
+    {
+      "epoch": 0.1205235679094051,
+      "grad_norm": 0.6737204194068909,
+      "learning_rate": 0.0001957399311153087,
+      "loss": 0.8224,
+      "step": 1639
+    },
+    {
+      "epoch": 0.12059710272814178,
+      "grad_norm": 0.9417632818222046,
+      "learning_rate": 0.00019573305106515315,
+      "loss": 0.814,
+      "step": 1640
+    },
+    {
+      "epoch": 0.12067063754687844,
+      "grad_norm": 0.7143616676330566,
+      "learning_rate": 0.00019572616558491795,
+      "loss": 0.8273,
+      "step": 1641
+    },
+    {
+      "epoch": 0.12074417236561512,
+      "grad_norm": 0.6060929298400879,
+      "learning_rate": 0.0001957192746749936,
+      "loss": 0.8007,
+      "step": 1642
+    },
+    {
+      "epoch": 0.1208177071843518,
+      "grad_norm": 0.5946378707885742,
+      "learning_rate": 0.00019571237833577098,
+      "loss": 0.8274,
+      "step": 1643
+    },
+    {
+      "epoch": 0.12089124200308846,
+      "grad_norm": 0.7428919076919556,
+      "learning_rate": 0.00019570547656764124,
+      "loss": 0.8372,
+      "step": 1644
+    },
+    {
+      "epoch": 0.12096477682182513,
+      "grad_norm": 0.9110476970672607,
+      "learning_rate": 0.0001956985693709959,
+      "loss": 0.8084,
+      "step": 1645
+    },
+    {
+      "epoch": 0.12103831164056181,
+      "grad_norm": 0.9914560914039612,
+      "learning_rate": 0.00019569165674622666,
+      "loss": 0.8069,
+      "step": 1646
+    },
+    {
+      "epoch": 0.12111184645929848,
+      "grad_norm": 0.6678136587142944,
+      "learning_rate": 0.00019568473869372571,
+      "loss": 0.7969,
+      "step": 1647
+    },
+    {
+      "epoch": 0.12118538127803515,
+      "grad_norm": 1.334033489227295,
+      "learning_rate": 0.0001956778152138854,
+      "loss": 0.8113,
+      "step": 1648
+    },
+    {
+      "epoch": 0.12125891609677182,
+      "grad_norm": 0.7146121859550476,
+      "learning_rate": 0.0001956708863070984,
+      "loss": 0.8221,
+      "step": 1649
+    },
+    {
+      "epoch": 0.1213324509155085,
+      "grad_norm": 1.6840062141418457,
+      "learning_rate": 0.0001956639519737578,
+      "loss": 0.8518,
+      "step": 1650
+    },
+    {
+      "epoch": 0.12140598573424516,
+      "grad_norm": 1.0042973756790161,
+      "learning_rate": 0.00019565701221425688,
+      "loss": 0.8443,
+      "step": 1651
+    },
+    {
+      "epoch": 0.12147952055298183,
+      "grad_norm": 1.8196910619735718,
+      "learning_rate": 0.00019565006702898925,
+      "loss": 0.833,
+      "step": 1652
+    },
+    {
+      "epoch": 0.12155305537171851,
+      "grad_norm": 1.756319522857666,
+      "learning_rate": 0.0001956431164183489,
+      "loss": 0.8238,
+      "step": 1653
+    },
+    {
+      "epoch": 0.12162659019045519,
+      "grad_norm": 1.058904767036438,
+      "learning_rate": 0.00019563616038273003,
+      "loss": 0.8271,
+      "step": 1654
+    },
+    {
+      "epoch": 0.12170012500919185,
+      "grad_norm": 0.9911401271820068,
+      "learning_rate": 0.00019562919892252722,
+      "loss": 0.8199,
+      "step": 1655
+    },
+    {
+      "epoch": 0.12177365982792852,
+      "grad_norm": 0.9724599719047546,
+      "learning_rate": 0.00019562223203813535,
+      "loss": 0.838,
+      "step": 1656
+    },
+    {
+      "epoch": 0.1218471946466652,
+      "grad_norm": 1.0892648696899414,
+      "learning_rate": 0.00019561525972994952,
+      "loss": 0.8138,
+      "step": 1657
+    },
+    {
+      "epoch": 0.12192072946540186,
+      "grad_norm": 1.034786343574524,
+      "learning_rate": 0.0001956082819983653,
+      "loss": 0.8136,
+      "step": 1658
+    },
+    {
+      "epoch": 0.12199426428413854,
+      "grad_norm": 1.3634626865386963,
+      "learning_rate": 0.0001956012988437784,
+      "loss": 0.7944,
+      "step": 1659
+    },
+    {
+      "epoch": 0.12206779910287521,
+      "grad_norm": 1.0958759784698486,
+      "learning_rate": 0.0001955943102665849,
+      "loss": 0.7783,
+      "step": 1660
+    },
+    {
+      "epoch": 0.12214133392161189,
+      "grad_norm": 1.2753406763076782,
+      "learning_rate": 0.0001955873162671813,
+      "loss": 0.8353,
+      "step": 1661
+    },
+    {
+      "epoch": 0.12221486874034855,
+      "grad_norm": 1.0364642143249512,
+      "learning_rate": 0.00019558031684596423,
+      "loss": 0.8081,
+      "step": 1662
+    },
+    {
+      "epoch": 0.12228840355908523,
+      "grad_norm": 0.993880569934845,
+      "learning_rate": 0.00019557331200333072,
+      "loss": 0.7896,
+      "step": 1663
+    },
+    {
+      "epoch": 0.1223619383778219,
+      "grad_norm": 0.9834621548652649,
+      "learning_rate": 0.00019556630173967807,
+      "loss": 0.7953,
+      "step": 1664
+    },
+    {
+      "epoch": 0.12243547319655856,
+      "grad_norm": 1.1589241027832031,
+      "learning_rate": 0.00019555928605540394,
+      "loss": 0.7916,
+      "step": 1665
+    },
+    {
+      "epoch": 0.12250900801529524,
+      "grad_norm": 0.7051790356636047,
+      "learning_rate": 0.00019555226495090625,
+      "loss": 0.8094,
+      "step": 1666
+    },
+    {
+      "epoch": 0.12258254283403192,
+      "grad_norm": 0.7266932725906372,
+      "learning_rate": 0.00019554523842658329,
+      "loss": 0.8214,
+      "step": 1667
+    },
+    {
+      "epoch": 0.12265607765276859,
+      "grad_norm": 0.6855170726776123,
+      "learning_rate": 0.00019553820648283353,
+      "loss": 0.7993,
+      "step": 1668
+    },
+    {
+      "epoch": 0.12272961247150525,
+      "grad_norm": 0.6495578289031982,
+      "learning_rate": 0.00019553116912005586,
+      "loss": 0.7829,
+      "step": 1669
+    },
+    {
+      "epoch": 0.12280314729024193,
+      "grad_norm": 0.686926543712616,
+      "learning_rate": 0.00019552412633864948,
+      "loss": 0.7635,
+      "step": 1670
+    },
+    {
+      "epoch": 0.1228766821089786,
+      "grad_norm": 0.8450779914855957,
+      "learning_rate": 0.00019551707813901387,
+      "loss": 0.8128,
+      "step": 1671
+    },
+    {
+      "epoch": 0.12295021692771527,
+      "grad_norm": 1.197353482246399,
+      "learning_rate": 0.00019551002452154875,
+      "loss": 0.8125,
+      "step": 1672
+    },
+    {
+      "epoch": 0.12302375174645194,
+      "grad_norm": 0.72205650806427,
+      "learning_rate": 0.00019550296548665426,
+      "loss": 0.83,
+      "step": 1673
+    },
+    {
+      "epoch": 0.12309728656518862,
+      "grad_norm": 1.2823433876037598,
+      "learning_rate": 0.00019549590103473078,
+      "loss": 0.8142,
+      "step": 1674
+    },
+    {
+      "epoch": 0.1231708213839253,
+      "grad_norm": 0.658149242401123,
+      "learning_rate": 0.00019548883116617902,
+      "loss": 0.7814,
+      "step": 1675
+    },
+    {
+      "epoch": 0.12324435620266196,
+      "grad_norm": 1.1555169820785522,
+      "learning_rate": 0.00019548175588139996,
+      "loss": 0.7956,
+      "step": 1676
+    },
+    {
+      "epoch": 0.12331789102139863,
+      "grad_norm": 0.7551974654197693,
+      "learning_rate": 0.00019547467518079493,
+      "loss": 0.8097,
+      "step": 1677
+    },
+    {
+      "epoch": 0.12339142584013531,
+      "grad_norm": 0.6732266545295715,
+      "learning_rate": 0.0001954675890647656,
+      "loss": 0.7809,
+      "step": 1678
+    },
+    {
+      "epoch": 0.12346496065887197,
+      "grad_norm": 0.7144074440002441,
+      "learning_rate": 0.00019546049753371388,
+      "loss": 0.7786,
+      "step": 1679
+    },
+    {
+      "epoch": 0.12353849547760865,
+      "grad_norm": 1.0866152048110962,
+      "learning_rate": 0.00019545340058804196,
+      "loss": 0.7922,
+      "step": 1680
+    },
+    {
+      "epoch": 0.12361203029634532,
+      "grad_norm": 1.1235846281051636,
+      "learning_rate": 0.0001954462982281524,
+      "loss": 0.7631,
+      "step": 1681
+    },
+    {
+      "epoch": 0.123685565115082,
+      "grad_norm": 0.6182257533073425,
+      "learning_rate": 0.00019543919045444813,
+      "loss": 0.7967,
+      "step": 1682
+    },
+    {
+      "epoch": 0.12375909993381866,
+      "grad_norm": 1.4904069900512695,
+      "learning_rate": 0.00019543207726733223,
+      "loss": 0.7913,
+      "step": 1683
+    },
+    {
+      "epoch": 0.12383263475255533,
+      "grad_norm": 0.5907593369483948,
+      "learning_rate": 0.0001954249586672082,
+      "loss": 0.7721,
+      "step": 1684
+    },
+    {
+      "epoch": 0.12390616957129201,
+      "grad_norm": 1.3172903060913086,
+      "learning_rate": 0.00019541783465447984,
+      "loss": 0.812,
+      "step": 1685
+    },
+    {
+      "epoch": 0.12397970439002867,
+      "grad_norm": 0.745670735836029,
+      "learning_rate": 0.00019541070522955116,
+      "loss": 0.8118,
+      "step": 1686
+    },
+    {
+      "epoch": 0.12405323920876535,
+      "grad_norm": 0.6827477216720581,
+      "learning_rate": 0.00019540357039282658,
+      "loss": 0.7825,
+      "step": 1687
+    },
+    {
+      "epoch": 0.12412677402750202,
+      "grad_norm": 0.9526305794715881,
+      "learning_rate": 0.00019539643014471084,
+      "loss": 0.7685,
+      "step": 1688
+    },
+    {
+      "epoch": 0.1242003088462387,
+      "grad_norm": 1.1133743524551392,
+      "learning_rate": 0.0001953892844856089,
+      "loss": 0.7855,
+      "step": 1689
+    },
+    {
+      "epoch": 0.12427384366497536,
+      "grad_norm": 0.6895951628684998,
+      "learning_rate": 0.00019538213341592603,
+      "loss": 0.7759,
+      "step": 1690
+    },
+    {
+      "epoch": 0.12434737848371204,
+      "grad_norm": 0.893835723400116,
+      "learning_rate": 0.00019537497693606793,
+      "loss": 0.7766,
+      "step": 1691
+    },
+    {
+      "epoch": 0.12442091330244871,
+      "grad_norm": 1.5720313787460327,
+      "learning_rate": 0.0001953678150464405,
+      "loss": 0.8135,
+      "step": 1692
+    },
+    {
+      "epoch": 0.12449444812118537,
+      "grad_norm": 0.8618747591972351,
+      "learning_rate": 0.00019536064774744994,
+      "loss": 0.7924,
+      "step": 1693
+    },
+    {
+      "epoch": 0.12456798293992205,
+      "grad_norm": 1.5932917594909668,
+      "learning_rate": 0.0001953534750395028,
+      "loss": 0.8047,
+      "step": 1694
+    },
+    {
+      "epoch": 0.12464151775865873,
+      "grad_norm": 1.0366971492767334,
+      "learning_rate": 0.0001953462969230059,
+      "loss": 0.7592,
+      "step": 1695
+    },
+    {
+      "epoch": 0.1247150525773954,
+      "grad_norm": 1.867653250694275,
+      "learning_rate": 0.00019533911339836643,
+      "loss": 0.7956,
+      "step": 1696
+    },
+    {
+      "epoch": 0.12478858739613206,
+      "grad_norm": 1.3899480104446411,
+      "learning_rate": 0.00019533192446599185,
+      "loss": 0.8062,
+      "step": 1697
+    },
+    {
+      "epoch": 0.12486212221486874,
+      "grad_norm": 1.9249718189239502,
+      "learning_rate": 0.0001953247301262899,
+      "loss": 0.7605,
+      "step": 1698
+    },
+    {
+      "epoch": 0.12493565703360542,
+      "grad_norm": 1.408210277557373,
+      "learning_rate": 0.00019531753037966865,
+      "loss": 0.7766,
+      "step": 1699
+    },
+    {
+      "epoch": 0.12500919185234208,
+      "grad_norm": 1.4330681562423706,
+      "learning_rate": 0.00019531032522653648,
+      "loss": 0.764,
+      "step": 1700
+    },
+    {
+      "epoch": 0.12508272667107875,
+      "grad_norm": 1.3606981039047241,
+      "learning_rate": 0.0001953031146673021,
+      "loss": 0.8107,
+      "step": 1701
+    },
+    {
+      "epoch": 0.12515626148981543,
+      "grad_norm": 1.6727102994918823,
+      "learning_rate": 0.00019529589870237443,
+      "loss": 0.7671,
+      "step": 1702
+    },
+    {
+      "epoch": 0.1252297963085521,
+      "grad_norm": 1.3221168518066406,
+      "learning_rate": 0.00019528867733216288,
+      "loss": 0.7608,
+      "step": 1703
+    },
+    {
+      "epoch": 0.12530333112728878,
+      "grad_norm": 1.6415146589279175,
+      "learning_rate": 0.00019528145055707695,
+      "loss": 0.7789,
+      "step": 1704
+    },
+    {
+      "epoch": 0.12537686594602546,
+      "grad_norm": 1.2395414113998413,
+      "learning_rate": 0.0001952742183775266,
+      "loss": 0.7856,
+      "step": 1705
+    },
+    {
+      "epoch": 0.1254504007647621,
+      "grad_norm": 1.4304600954055786,
+      "learning_rate": 0.00019526698079392203,
+      "loss": 0.8079,
+      "step": 1706
+    },
+    {
+      "epoch": 0.12552393558349878,
+      "grad_norm": 1.0059683322906494,
+      "learning_rate": 0.00019525973780667378,
+      "loss": 0.7807,
+      "step": 1707
+    },
+    {
+      "epoch": 0.12559747040223546,
+      "grad_norm": 1.3990800380706787,
+      "learning_rate": 0.00019525248941619267,
+      "loss": 0.7468,
+      "step": 1708
+    },
+    {
+      "epoch": 0.12567100522097213,
+      "grad_norm": 0.9708283543586731,
+      "learning_rate": 0.00019524523562288983,
+      "loss": 0.7653,
+      "step": 1709
+    },
+    {
+      "epoch": 0.1257445400397088,
+      "grad_norm": 1.8022875785827637,
+      "learning_rate": 0.0001952379764271767,
+      "loss": 0.7666,
+      "step": 1710
+    },
+    {
+      "epoch": 0.12581807485844548,
+      "grad_norm": 1.058467149734497,
+      "learning_rate": 0.00019523071182946507,
+      "loss": 0.7783,
+      "step": 1711
+    },
+    {
+      "epoch": 0.12589160967718216,
+      "grad_norm": 1.405788540840149,
+      "learning_rate": 0.00019522344183016695,
+      "loss": 0.7868,
+      "step": 1712
+    },
+    {
+      "epoch": 0.1259651444959188,
+      "grad_norm": 0.9785199165344238,
+      "learning_rate": 0.00019521616642969471,
+      "loss": 0.7571,
+      "step": 1713
+    },
+    {
+      "epoch": 0.12603867931465548,
+      "grad_norm": 1.6926510334014893,
+      "learning_rate": 0.00019520888562846102,
+      "loss": 0.8221,
+      "step": 1714
+    },
+    {
+      "epoch": 0.12611221413339216,
+      "grad_norm": 1.0594542026519775,
+      "learning_rate": 0.00019520159942687886,
+      "loss": 0.7944,
+      "step": 1715
+    },
+    {
+      "epoch": 0.12618574895212883,
+      "grad_norm": 1.2328039407730103,
+      "learning_rate": 0.00019519430782536152,
+      "loss": 0.7503,
+      "step": 1716
+    },
+    {
+      "epoch": 0.1262592837708655,
+      "grad_norm": 0.918054461479187,
+      "learning_rate": 0.00019518701082432257,
+      "loss": 0.7943,
+      "step": 1717
+    },
+    {
+      "epoch": 0.12633281858960219,
+      "grad_norm": 0.7576400637626648,
+      "learning_rate": 0.0001951797084241759,
+      "loss": 0.7532,
+      "step": 1718
+    },
+    {
+      "epoch": 0.12640635340833886,
+      "grad_norm": 0.8530510663986206,
+      "learning_rate": 0.00019517240062533574,
+      "loss": 0.7764,
+      "step": 1719
+    },
+    {
+      "epoch": 0.1264798882270755,
+      "grad_norm": 0.6186944246292114,
+      "learning_rate": 0.00019516508742821657,
+      "loss": 0.7948,
+      "step": 1720
+    },
+    {
+      "epoch": 0.12655342304581219,
+      "grad_norm": 0.6279460787773132,
+      "learning_rate": 0.0001951577688332332,
+      "loss": 0.7218,
+      "step": 1721
+    },
+    {
+      "epoch": 0.12662695786454886,
+      "grad_norm": 0.8177329301834106,
+      "learning_rate": 0.00019515044484080078,
+      "loss": 0.7608,
+      "step": 1722
+    },
+    {
+      "epoch": 0.12670049268328554,
+      "grad_norm": 0.8306900262832642,
+      "learning_rate": 0.0001951431154513347,
+      "loss": 0.7713,
+      "step": 1723
+    },
+    {
+      "epoch": 0.1267740275020222,
+      "grad_norm": 0.5771136283874512,
+      "learning_rate": 0.0001951357806652507,
+      "loss": 0.772,
+      "step": 1724
+    },
+    {
+      "epoch": 0.1268475623207589,
+      "grad_norm": 1.017773151397705,
+      "learning_rate": 0.00019512844048296483,
+      "loss": 0.7848,
+      "step": 1725
+    },
+    {
+      "epoch": 0.12692109713949556,
+      "grad_norm": 0.6834537982940674,
+      "learning_rate": 0.00019512109490489344,
+      "loss": 0.7719,
+      "step": 1726
+    },
+    {
+      "epoch": 0.1269946319582322,
+      "grad_norm": 1.0678815841674805,
+      "learning_rate": 0.00019511374393145313,
+      "loss": 0.7323,
+      "step": 1727
+    },
+    {
+      "epoch": 0.1270681667769689,
+      "grad_norm": 1.0056872367858887,
+      "learning_rate": 0.0001951063875630609,
+      "loss": 0.7415,
+      "step": 1728
+    },
+    {
+      "epoch": 0.12714170159570556,
+      "grad_norm": 0.6046265959739685,
+      "learning_rate": 0.00019509902580013399,
+      "loss": 0.7747,
+      "step": 1729
+    },
+    {
+      "epoch": 0.12721523641444224,
+      "grad_norm": 0.6998772025108337,
+      "learning_rate": 0.00019509165864308998,
+      "loss": 0.7332,
+      "step": 1730
+    },
+    {
+      "epoch": 0.12728877123317892,
+      "grad_norm": 0.9207208752632141,
+      "learning_rate": 0.00019508428609234672,
+      "loss": 0.7589,
+      "step": 1731
+    },
+    {
+      "epoch": 0.1273623060519156,
+      "grad_norm": 0.8462632894515991,
+      "learning_rate": 0.0001950769081483224,
+      "loss": 0.751,
+      "step": 1732
+    },
+    {
+      "epoch": 0.12743584087065227,
+      "grad_norm": 0.5885737538337708,
+      "learning_rate": 0.00019506952481143555,
+      "loss": 0.7545,
+      "step": 1733
+    },
+    {
+      "epoch": 0.12750937568938892,
+      "grad_norm": 1.0281561613082886,
+      "learning_rate": 0.00019506213608210492,
+      "loss": 0.7817,
+      "step": 1734
+    },
+    {
+      "epoch": 0.1275829105081256,
+      "grad_norm": 0.7124502062797546,
+      "learning_rate": 0.00019505474196074957,
+      "loss": 0.7497,
+      "step": 1735
+    },
+    {
+      "epoch": 0.12765644532686227,
+      "grad_norm": 0.8069841861724854,
+      "learning_rate": 0.00019504734244778896,
+      "loss": 0.7555,
+      "step": 1736
+    },
+    {
+      "epoch": 0.12772998014559894,
+      "grad_norm": 0.9870133399963379,
+      "learning_rate": 0.0001950399375436428,
+      "loss": 0.7583,
+      "step": 1737
+    },
+    {
+      "epoch": 0.12780351496433562,
+      "grad_norm": 0.8712988495826721,
+      "learning_rate": 0.00019503252724873107,
+      "loss": 0.7715,
+      "step": 1738
+    },
+    {
+      "epoch": 0.1278770497830723,
+      "grad_norm": 0.6297889947891235,
+      "learning_rate": 0.0001950251115634741,
+      "loss": 0.7859,
+      "step": 1739
+    },
+    {
+      "epoch": 0.12795058460180897,
+      "grad_norm": 0.5477202534675598,
+      "learning_rate": 0.0001950176904882925,
+      "loss": 0.7381,
+      "step": 1740
+    },
+    {
+      "epoch": 0.12802411942054562,
+      "grad_norm": 1.139739990234375,
+      "learning_rate": 0.00019501026402360725,
+      "loss": 0.7239,
+      "step": 1741
+    },
+    {
+      "epoch": 0.1280976542392823,
+      "grad_norm": 1.087947130203247,
+      "learning_rate": 0.00019500283216983952,
+      "loss": 0.7172,
+      "step": 1742
+    },
+    {
+      "epoch": 0.12817118905801897,
+      "grad_norm": 0.5017962455749512,
+      "learning_rate": 0.00019499539492741092,
+      "loss": 0.7262,
+      "step": 1743
+    },
+    {
+      "epoch": 0.12824472387675565,
+      "grad_norm": 0.8045006990432739,
+      "learning_rate": 0.00019498795229674326,
+      "loss": 0.7539,
+      "step": 1744
+    },
+    {
+      "epoch": 0.12831825869549232,
+      "grad_norm": 0.915061891078949,
+      "learning_rate": 0.0001949805042782587,
+      "loss": 0.7209,
+      "step": 1745
+    },
+    {
+      "epoch": 0.128391793514229,
+      "grad_norm": 0.5888285636901855,
+      "learning_rate": 0.00019497305087237972,
+      "loss": 0.7383,
+      "step": 1746
+    },
+    {
+      "epoch": 0.12846532833296567,
+      "grad_norm": 0.6382765769958496,
+      "learning_rate": 0.00019496559207952903,
+      "loss": 0.7329,
+      "step": 1747
+    },
+    {
+      "epoch": 0.12853886315170232,
+      "grad_norm": 0.7936418652534485,
+      "learning_rate": 0.00019495812790012976,
+      "loss": 0.7656,
+      "step": 1748
+    },
+    {
+      "epoch": 0.128612397970439,
+      "grad_norm": 0.6652594804763794,
+      "learning_rate": 0.00019495065833460525,
+      "loss": 0.7479,
+      "step": 1749
+    },
+    {
+      "epoch": 0.12868593278917567,
+      "grad_norm": 0.49818679690361023,
+      "learning_rate": 0.00019494318338337917,
+      "loss": 0.7252,
+      "step": 1750
+    },
+    {
+      "epoch": 0.12875946760791235,
+      "grad_norm": 0.5918554663658142,
+      "learning_rate": 0.00019493570304687558,
+      "loss": 0.7466,
+      "step": 1751
+    },
+    {
+      "epoch": 0.12883300242664902,
+      "grad_norm": 0.7673982381820679,
+      "learning_rate": 0.0001949282173255187,
+      "loss": 0.7621,
+      "step": 1752
+    },
+    {
+      "epoch": 0.1289065372453857,
+      "grad_norm": 0.9725474119186401,
+      "learning_rate": 0.00019492072621973315,
+      "loss": 0.7478,
+      "step": 1753
+    },
+    {
+      "epoch": 0.12898007206412238,
+      "grad_norm": 1.132180094718933,
+      "learning_rate": 0.00019491322972994384,
+      "loss": 0.7501,
+      "step": 1754
+    },
+    {
+      "epoch": 0.12905360688285902,
+      "grad_norm": 0.5836314558982849,
+      "learning_rate": 0.00019490572785657597,
+      "loss": 0.7672,
+      "step": 1755
+    },
+    {
+      "epoch": 0.1291271417015957,
+      "grad_norm": 0.7588699460029602,
+      "learning_rate": 0.00019489822060005506,
+      "loss": 0.7363,
+      "step": 1756
+    },
+    {
+      "epoch": 0.12920067652033237,
+      "grad_norm": 1.0236645936965942,
+      "learning_rate": 0.00019489070796080692,
+      "loss": 0.7358,
+      "step": 1757
+    },
+    {
+      "epoch": 0.12927421133906905,
+      "grad_norm": 0.6299167275428772,
+      "learning_rate": 0.0001948831899392577,
+      "loss": 0.7346,
+      "step": 1758
+    },
+    {
+      "epoch": 0.12934774615780573,
+      "grad_norm": 0.7498030066490173,
+      "learning_rate": 0.00019487566653583378,
+      "loss": 0.7216,
+      "step": 1759
+    },
+    {
+      "epoch": 0.1294212809765424,
+      "grad_norm": 0.9118340611457825,
+      "learning_rate": 0.00019486813775096198,
+      "loss": 0.7245,
+      "step": 1760
+    },
+    {
+      "epoch": 0.12949481579527908,
+      "grad_norm": 1.0518486499786377,
+      "learning_rate": 0.00019486060358506926,
+      "loss": 0.7401,
+      "step": 1761
+    },
+    {
+      "epoch": 0.12956835061401573,
+      "grad_norm": 0.9552172422409058,
+      "learning_rate": 0.00019485306403858296,
+      "loss": 0.7161,
+      "step": 1762
+    },
+    {
+      "epoch": 0.1296418854327524,
+      "grad_norm": 0.6864818334579468,
+      "learning_rate": 0.00019484551911193082,
+      "loss": 0.6968,
+      "step": 1763
+    },
+    {
+      "epoch": 0.12971542025148908,
+      "grad_norm": 0.7652767896652222,
+      "learning_rate": 0.0001948379688055407,
+      "loss": 0.7265,
+      "step": 1764
+    },
+    {
+      "epoch": 0.12978895507022575,
+      "grad_norm": 1.254724383354187,
+      "learning_rate": 0.00019483041311984094,
+      "loss": 0.7349,
+      "step": 1765
+    },
+    {
+      "epoch": 0.12986248988896243,
+      "grad_norm": 0.5217794179916382,
+      "learning_rate": 0.00019482285205526005,
+      "loss": 0.7134,
+      "step": 1766
+    },
+    {
+      "epoch": 0.1299360247076991,
+      "grad_norm": 1.2283852100372314,
+      "learning_rate": 0.00019481528561222691,
+      "loss": 0.6786,
+      "step": 1767
+    },
+    {
+      "epoch": 0.13000955952643578,
+      "grad_norm": 0.7905964851379395,
+      "learning_rate": 0.00019480771379117072,
+      "loss": 0.6951,
+      "step": 1768
+    },
+    {
+      "epoch": 0.13008309434517243,
+      "grad_norm": 0.603315532207489,
+      "learning_rate": 0.00019480013659252095,
+      "loss": 0.6697,
+      "step": 1769
+    },
+    {
+      "epoch": 0.1301566291639091,
+      "grad_norm": 0.5920364856719971,
+      "learning_rate": 0.0001947925540167074,
+      "loss": 0.7282,
+      "step": 1770
+    },
+    {
+      "epoch": 0.13023016398264578,
+      "grad_norm": 0.6949764490127563,
+      "learning_rate": 0.00019478496606416013,
+      "loss": 0.6919,
+      "step": 1771
+    },
+    {
+      "epoch": 0.13030369880138246,
+      "grad_norm": 1.1755118370056152,
+      "learning_rate": 0.00019477737273530953,
+      "loss": 0.7293,
+      "step": 1772
+    },
+    {
+      "epoch": 0.13037723362011913,
+      "grad_norm": 1.2203205823898315,
+      "learning_rate": 0.00019476977403058636,
+      "loss": 0.7065,
+      "step": 1773
+    },
+    {
+      "epoch": 0.1304507684388558,
+      "grad_norm": 0.6201728582382202,
+      "learning_rate": 0.0001947621699504216,
+      "loss": 0.7147,
+      "step": 1774
+    },
+    {
+      "epoch": 0.13052430325759248,
+      "grad_norm": 1.577972173690796,
+      "learning_rate": 0.00019475456049524654,
+      "loss": 0.6939,
+      "step": 1775
+    },
+    {
+      "epoch": 0.13059783807632913,
+      "grad_norm": 0.6166728138923645,
+      "learning_rate": 0.0001947469456654928,
+      "loss": 0.692,
+      "step": 1776
+    },
+    {
+      "epoch": 0.1306713728950658,
+      "grad_norm": 1.5228432416915894,
+      "learning_rate": 0.00019473932546159234,
+      "loss": 0.7323,
+      "step": 1777
+    },
+    {
+      "epoch": 0.13074490771380248,
+      "grad_norm": 0.6862233877182007,
+      "learning_rate": 0.00019473169988397731,
+      "loss": 0.6959,
+      "step": 1778
+    },
+    {
+      "epoch": 0.13081844253253916,
+      "grad_norm": 1.5647475719451904,
+      "learning_rate": 0.00019472406893308034,
+      "loss": 0.6973,
+      "step": 1779
+    },
+    {
+      "epoch": 0.13089197735127583,
+      "grad_norm": 0.7381709218025208,
+      "learning_rate": 0.00019471643260933418,
+      "loss": 0.7071,
+      "step": 1780
+    },
+    {
+      "epoch": 0.1309655121700125,
+      "grad_norm": 1.1969512701034546,
+      "learning_rate": 0.000194708790913172,
+      "loss": 0.6877,
+      "step": 1781
+    },
+    {
+      "epoch": 0.13103904698874919,
+      "grad_norm": 0.8015069365501404,
+      "learning_rate": 0.00019470114384502728,
+      "loss": 0.676,
+      "step": 1782
+    },
+    {
+      "epoch": 0.13111258180748583,
+      "grad_norm": 0.7068674564361572,
+      "learning_rate": 0.00019469349140533373,
+      "loss": 0.7131,
+      "step": 1783
+    },
+    {
+      "epoch": 0.1311861166262225,
+      "grad_norm": 0.732627272605896,
+      "learning_rate": 0.0001946858335945254,
+      "loss": 0.6824,
+      "step": 1784
+    },
+    {
+      "epoch": 0.13125965144495919,
+      "grad_norm": 0.6682825684547424,
+      "learning_rate": 0.00019467817041303668,
+      "loss": 0.7171,
+      "step": 1785
+    },
+    {
+      "epoch": 0.13133318626369586,
+      "grad_norm": 1.0662728548049927,
+      "learning_rate": 0.0001946705018613022,
+      "loss": 0.722,
+      "step": 1786
+    },
+    {
+      "epoch": 0.13140672108243254,
+      "grad_norm": 1.0373941659927368,
+      "learning_rate": 0.00019466282793975694,
+      "loss": 0.7016,
+      "step": 1787
+    },
+    {
+      "epoch": 0.1314802559011692,
+      "grad_norm": 0.6554132103919983,
+      "learning_rate": 0.0001946551486488362,
+      "loss": 0.6908,
+      "step": 1788
+    },
+    {
+      "epoch": 0.1315537907199059,
+      "grad_norm": 1.0959566831588745,
+      "learning_rate": 0.00019464746398897552,
+      "loss": 0.7124,
+      "step": 1789
+    },
+    {
+      "epoch": 0.13162732553864254,
+      "grad_norm": 0.6204771995544434,
+      "learning_rate": 0.0001946397739606108,
+      "loss": 0.6845,
+      "step": 1790
+    },
+    {
+      "epoch": 0.1317008603573792,
+      "grad_norm": 0.6984374523162842,
+      "learning_rate": 0.0001946320785641782,
+      "loss": 0.6816,
+      "step": 1791
+    },
+    {
+      "epoch": 0.1317743951761159,
+      "grad_norm": 0.7033853530883789,
+      "learning_rate": 0.00019462437780011426,
+      "loss": 0.6897,
+      "step": 1792
+    },
+    {
+      "epoch": 0.13184792999485256,
+      "grad_norm": 0.73503178358078,
+      "learning_rate": 0.00019461667166885578,
+      "loss": 0.6959,
+      "step": 1793
+    },
+    {
+      "epoch": 0.13192146481358924,
+      "grad_norm": 0.5896041989326477,
+      "learning_rate": 0.00019460896017083979,
+      "loss": 0.6695,
+      "step": 1794
+    },
+    {
+      "epoch": 0.13199499963232592,
+      "grad_norm": 0.7088213562965393,
+      "learning_rate": 0.00019460124330650374,
+      "loss": 0.6723,
+      "step": 1795
+    },
+    {
+      "epoch": 0.1320685344510626,
+      "grad_norm": 0.5740508437156677,
+      "learning_rate": 0.00019459352107628532,
+      "loss": 0.6731,
+      "step": 1796
+    },
+    {
+      "epoch": 0.13214206926979924,
+      "grad_norm": 0.9552329182624817,
+      "learning_rate": 0.0001945857934806226,
+      "loss": 0.7234,
+      "step": 1797
+    },
+    {
+      "epoch": 0.13221560408853592,
+      "grad_norm": 0.8026865720748901,
+      "learning_rate": 0.0001945780605199538,
+      "loss": 0.7076,
+      "step": 1798
+    },
+    {
+      "epoch": 0.1322891389072726,
+      "grad_norm": 0.894896924495697,
+      "learning_rate": 0.00019457032219471763,
+      "loss": 0.6596,
+      "step": 1799
+    },
+    {
+      "epoch": 0.13236267372600927,
+      "grad_norm": 1.0590513944625854,
+      "learning_rate": 0.00019456257850535295,
+      "loss": 0.7106,
+      "step": 1800
+    },
+    {
+      "epoch": 0.13243620854474594,
+      "grad_norm": 1.0827540159225464,
+      "learning_rate": 0.00019455482945229905,
+      "loss": 0.6582,
+      "step": 1801
+    },
+    {
+      "epoch": 0.13250974336348262,
+      "grad_norm": 0.6141829490661621,
+      "learning_rate": 0.00019454707503599542,
+      "loss": 0.7173,
+      "step": 1802
+    },
+    {
+      "epoch": 0.1325832781822193,
+      "grad_norm": 1.002894401550293,
+      "learning_rate": 0.00019453931525688194,
+      "loss": 0.7485,
+      "step": 1803
+    },
+    {
+      "epoch": 0.13265681300095594,
+      "grad_norm": 1.0377345085144043,
+      "learning_rate": 0.00019453155011539872,
+      "loss": 0.6908,
+      "step": 1804
+    },
+    {
+      "epoch": 0.13273034781969262,
+      "grad_norm": 0.6541041731834412,
+      "learning_rate": 0.00019452377961198618,
+      "loss": 0.6892,
+      "step": 1805
+    },
+    {
+      "epoch": 0.1328038826384293,
+      "grad_norm": 0.5425859689712524,
+      "learning_rate": 0.00019451600374708513,
+      "loss": 0.6742,
+      "step": 1806
+    },
+    {
+      "epoch": 0.13287741745716597,
+      "grad_norm": 0.607164740562439,
+      "learning_rate": 0.0001945082225211366,
+      "loss": 0.7117,
+      "step": 1807
+    },
+    {
+      "epoch": 0.13295095227590265,
+      "grad_norm": 0.5848445296287537,
+      "learning_rate": 0.00019450043593458196,
+      "loss": 0.6982,
+      "step": 1808
+    },
+    {
+      "epoch": 0.13302448709463932,
+      "grad_norm": 0.652161180973053,
+      "learning_rate": 0.00019449264398786284,
+      "loss": 0.6761,
+      "step": 1809
+    },
+    {
+      "epoch": 0.133098021913376,
+      "grad_norm": 0.8452996015548706,
+      "learning_rate": 0.00019448484668142127,
+      "loss": 0.7221,
+      "step": 1810
+    },
+    {
+      "epoch": 0.13317155673211264,
+      "grad_norm": 1.3601144552230835,
+      "learning_rate": 0.00019447704401569945,
+      "loss": 0.6596,
+      "step": 1811
+    },
+    {
+      "epoch": 0.13324509155084932,
+      "grad_norm": 0.6048308610916138,
+      "learning_rate": 0.00019446923599114,
+      "loss": 0.6574,
+      "step": 1812
+    },
+    {
+      "epoch": 0.133318626369586,
+      "grad_norm": 1.222616195678711,
+      "learning_rate": 0.00019446142260818577,
+      "loss": 0.6756,
+      "step": 1813
+    },
+    {
+      "epoch": 0.13339216118832267,
+      "grad_norm": 0.7290657758712769,
+      "learning_rate": 0.00019445360386727997,
+      "loss": 0.64,
+      "step": 1814
+    },
+    {
+      "epoch": 0.13346569600705935,
+      "grad_norm": 0.6757525205612183,
+      "learning_rate": 0.00019444577976886608,
+      "loss": 0.7205,
+      "step": 1815
+    },
+    {
+      "epoch": 0.13353923082579602,
+      "grad_norm": 0.7393860816955566,
+      "learning_rate": 0.0001944379503133879,
+      "loss": 0.6853,
+      "step": 1816
+    },
+    {
+      "epoch": 0.1336127656445327,
+      "grad_norm": 0.6107260584831238,
+      "learning_rate": 0.0001944301155012895,
+      "loss": 0.6992,
+      "step": 1817
+    },
+    {
+      "epoch": 0.13368630046326935,
+      "grad_norm": 1.1067112684249878,
+      "learning_rate": 0.0001944222753330153,
+      "loss": 0.6873,
+      "step": 1818
+    },
+    {
+      "epoch": 0.13375983528200602,
+      "grad_norm": 0.9028744101524353,
+      "learning_rate": 0.00019441442980900998,
+      "loss": 0.701,
+      "step": 1819
+    },
+    {
+      "epoch": 0.1338333701007427,
+      "grad_norm": 0.7045875787734985,
+      "learning_rate": 0.00019440657892971859,
+      "loss": 0.6885,
+      "step": 1820
+    },
+    {
+      "epoch": 0.13390690491947937,
+      "grad_norm": 0.8709607124328613,
+      "learning_rate": 0.00019439872269558638,
+      "loss": 0.6871,
+      "step": 1821
+    },
+    {
+      "epoch": 0.13398043973821605,
+      "grad_norm": 1.1543354988098145,
+      "learning_rate": 0.00019439086110705902,
+      "loss": 0.7191,
+      "step": 1822
+    },
+    {
+      "epoch": 0.13405397455695273,
+      "grad_norm": 0.7575855255126953,
+      "learning_rate": 0.00019438299416458241,
+      "loss": 0.6879,
+      "step": 1823
+    },
+    {
+      "epoch": 0.1341275093756894,
+      "grad_norm": 0.6522961258888245,
+      "learning_rate": 0.00019437512186860274,
+      "loss": 0.7164,
+      "step": 1824
+    },
+    {
+      "epoch": 0.13420104419442605,
+      "grad_norm": 0.78129643201828,
+      "learning_rate": 0.00019436724421956658,
+      "loss": 0.6733,
+      "step": 1825
+    },
+    {
+      "epoch": 0.13427457901316273,
+      "grad_norm": 0.7831710577011108,
+      "learning_rate": 0.00019435936121792071,
+      "loss": 0.6702,
+      "step": 1826
+    },
+    {
+      "epoch": 0.1343481138318994,
+      "grad_norm": 0.7312570214271545,
+      "learning_rate": 0.0001943514728641123,
+      "loss": 0.686,
+      "step": 1827
+    },
+    {
+      "epoch": 0.13442164865063608,
+      "grad_norm": 0.7675382494926453,
+      "learning_rate": 0.0001943435791585888,
+      "loss": 0.706,
+      "step": 1828
+    },
+    {
+      "epoch": 0.13449518346937275,
+      "grad_norm": 0.600424587726593,
+      "learning_rate": 0.0001943356801017979,
+      "loss": 0.7131,
+      "step": 1829
+    },
+    {
+      "epoch": 0.13456871828810943,
+      "grad_norm": 0.7247167825698853,
+      "learning_rate": 0.00019432777569418767,
+      "loss": 0.7314,
+      "step": 1830
+    },
+    {
+      "epoch": 0.1346422531068461,
+      "grad_norm": 0.8456330299377441,
+      "learning_rate": 0.00019431986593620648,
+      "loss": 0.7113,
+      "step": 1831
+    },
+    {
+      "epoch": 0.13471578792558275,
+      "grad_norm": 1.6428226232528687,
+      "learning_rate": 0.00019431195082830294,
+      "loss": 0.7363,
+      "step": 1832
+    },
+    {
+      "epoch": 0.13478932274431943,
+      "grad_norm": 0.9136975407600403,
+      "learning_rate": 0.000194304030370926,
+      "loss": 0.6704,
+      "step": 1833
+    },
+    {
+      "epoch": 0.1348628575630561,
+      "grad_norm": 0.5414283275604248,
+      "learning_rate": 0.00019429610456452495,
+      "loss": 0.6907,
+      "step": 1834
+    },
+    {
+      "epoch": 0.13493639238179278,
+      "grad_norm": 0.6430392265319824,
+      "learning_rate": 0.00019428817340954931,
+      "loss": 0.6697,
+      "step": 1835
+    },
+    {
+      "epoch": 0.13500992720052946,
+      "grad_norm": 1.2050249576568604,
+      "learning_rate": 0.000194280236906449,
+      "loss": 0.6996,
+      "step": 1836
+    },
+    {
+      "epoch": 0.13508346201926613,
+      "grad_norm": 1.3290040493011475,
+      "learning_rate": 0.00019427229505567415,
+      "loss": 0.7117,
+      "step": 1837
+    },
+    {
+      "epoch": 0.1351569968380028,
+      "grad_norm": 0.5310449600219727,
+      "learning_rate": 0.00019426434785767524,
+      "loss": 0.7484,
+      "step": 1838
+    },
+    {
+      "epoch": 0.13523053165673946,
+      "grad_norm": 1.1104438304901123,
+      "learning_rate": 0.00019425639531290304,
+      "loss": 0.6858,
+      "step": 1839
+    },
+    {
+      "epoch": 0.13530406647547613,
+      "grad_norm": 0.7570423483848572,
+      "learning_rate": 0.00019424843742180857,
+      "loss": 0.655,
+      "step": 1840
+    },
+    {
+      "epoch": 0.1353776012942128,
+      "grad_norm": 0.6908128261566162,
+      "learning_rate": 0.0001942404741848433,
+      "loss": 0.682,
+      "step": 1841
+    },
+    {
+      "epoch": 0.13545113611294948,
+      "grad_norm": 0.7511933445930481,
+      "learning_rate": 0.00019423250560245893,
+      "loss": 0.6738,
+      "step": 1842
+    },
+    {
+      "epoch": 0.13552467093168616,
+      "grad_norm": 0.7712101936340332,
+      "learning_rate": 0.00019422453167510734,
+      "loss": 0.687,
+      "step": 1843
+    },
+    {
+      "epoch": 0.13559820575042283,
+      "grad_norm": 0.8200168013572693,
+      "learning_rate": 0.0001942165524032409,
+      "loss": 0.7166,
+      "step": 1844
+    },
+    {
+      "epoch": 0.1356717405691595,
+      "grad_norm": 1.5752986669540405,
+      "learning_rate": 0.00019420856778731214,
+      "loss": 0.7235,
+      "step": 1845
+    },
+    {
+      "epoch": 0.13574527538789616,
+      "grad_norm": 0.7058979272842407,
+      "learning_rate": 0.00019420057782777402,
+      "loss": 0.7301,
+      "step": 1846
+    },
+    {
+      "epoch": 0.13581881020663283,
+      "grad_norm": 0.9429430365562439,
+      "learning_rate": 0.0001941925825250797,
+      "loss": 0.6693,
+      "step": 1847
+    },
+    {
+      "epoch": 0.1358923450253695,
+      "grad_norm": 1.3372761011123657,
+      "learning_rate": 0.00019418458187968273,
+      "loss": 0.6847,
+      "step": 1848
+    },
+    {
+      "epoch": 0.13596587984410619,
+      "grad_norm": 0.599082350730896,
+      "learning_rate": 0.00019417657589203684,
+      "loss": 0.6989,
+      "step": 1849
+    },
+    {
+      "epoch": 0.13603941466284286,
+      "grad_norm": 1.4431345462799072,
+      "learning_rate": 0.00019416856456259618,
+      "loss": 0.6786,
+      "step": 1850
+    },
+    {
+      "epoch": 0.13611294948157954,
+      "grad_norm": 0.9219923615455627,
+      "learning_rate": 0.00019416054789181517,
+      "loss": 0.6987,
+      "step": 1851
+    },
+    {
+      "epoch": 0.1361864843003162,
+      "grad_norm": 0.5698516368865967,
+      "learning_rate": 0.00019415252588014852,
+      "loss": 0.67,
+      "step": 1852
+    },
+    {
+      "epoch": 0.13626001911905286,
+      "grad_norm": 0.868013858795166,
+      "learning_rate": 0.00019414449852805121,
+      "loss": 0.675,
+      "step": 1853
+    },
+    {
+      "epoch": 0.13633355393778954,
+      "grad_norm": 1.4987112283706665,
+      "learning_rate": 0.0001941364658359786,
+      "loss": 0.6971,
+      "step": 1854
+    },
+    {
+      "epoch": 0.1364070887565262,
+      "grad_norm": 0.7418074011802673,
+      "learning_rate": 0.00019412842780438633,
+      "loss": 0.6615,
+      "step": 1855
+    },
+    {
+      "epoch": 0.1364806235752629,
+      "grad_norm": 1.3866549730300903,
+      "learning_rate": 0.00019412038443373026,
+      "loss": 0.6921,
+      "step": 1856
+    },
+    {
+      "epoch": 0.13655415839399956,
+      "grad_norm": 1.7340675592422485,
+      "learning_rate": 0.00019411233572446665,
+      "loss": 0.7015,
+      "step": 1857
+    },
+    {
+      "epoch": 0.13662769321273624,
+      "grad_norm": 1.2221713066101074,
+      "learning_rate": 0.00019410428167705205,
+      "loss": 0.6688,
+      "step": 1858
+    },
+    {
+      "epoch": 0.13670122803147292,
+      "grad_norm": 2.8067970275878906,
+      "learning_rate": 0.0001940962222919433,
+      "loss": 0.6999,
+      "step": 1859
+    },
+    {
+      "epoch": 0.13677476285020956,
+      "grad_norm": 2.9688732624053955,
+      "learning_rate": 0.0001940881575695975,
+      "loss": 0.6613,
+      "step": 1860
+    },
+    {
+      "epoch": 0.13684829766894624,
+      "grad_norm": 0.6525237560272217,
+      "learning_rate": 0.00019408008751047213,
+      "loss": 0.6682,
+      "step": 1861
+    },
+    {
+      "epoch": 0.13692183248768292,
+      "grad_norm": 2.5926918983459473,
+      "learning_rate": 0.00019407201211502488,
+      "loss": 0.6887,
+      "step": 1862
+    },
+    {
+      "epoch": 0.1369953673064196,
+      "grad_norm": 1.722714900970459,
+      "learning_rate": 0.00019406393138371385,
+      "loss": 0.6959,
+      "step": 1863
+    },
+    {
+      "epoch": 0.13706890212515627,
+      "grad_norm": 2.0459702014923096,
+      "learning_rate": 0.00019405584531699733,
+      "loss": 0.6735,
+      "step": 1864
+    },
+    {
+      "epoch": 0.13714243694389294,
+      "grad_norm": 2.4482171535491943,
+      "learning_rate": 0.00019404775391533403,
+      "loss": 0.694,
+      "step": 1865
+    },
+    {
+      "epoch": 0.13721597176262962,
+      "grad_norm": 0.6610161066055298,
+      "learning_rate": 0.0001940396571791829,
+      "loss": 0.6619,
+      "step": 1866
+    },
+    {
+      "epoch": 0.13728950658136627,
+      "grad_norm": 2.0398833751678467,
+      "learning_rate": 0.00019403155510900312,
+      "loss": 0.6768,
+      "step": 1867
+    },
+    {
+      "epoch": 0.13736304140010294,
+      "grad_norm": 0.9665430784225464,
+      "learning_rate": 0.0001940234477052543,
+      "loss": 0.6834,
+      "step": 1868
+    },
+    {
+      "epoch": 0.13743657621883962,
+      "grad_norm": 2.519984483718872,
+      "learning_rate": 0.00019401533496839633,
+      "loss": 0.6935,
+      "step": 1869
+    },
+    {
+      "epoch": 0.1375101110375763,
+      "grad_norm": 2.4907233715057373,
+      "learning_rate": 0.00019400721689888934,
+      "loss": 0.699,
+      "step": 1870
+    },
+    {
+      "epoch": 0.13758364585631297,
+      "grad_norm": 0.9714620113372803,
+      "learning_rate": 0.00019399909349719375,
+      "loss": 0.6536,
+      "step": 1871
+    },
+    {
+      "epoch": 0.13765718067504965,
+      "grad_norm": 1.6245673894882202,
+      "learning_rate": 0.00019399096476377042,
+      "loss": 0.6845,
+      "step": 1872
+    },
+    {
+      "epoch": 0.13773071549378632,
+      "grad_norm": 1.0567419528961182,
+      "learning_rate": 0.0001939828306990804,
+      "loss": 0.6557,
+      "step": 1873
+    },
+    {
+      "epoch": 0.13780425031252297,
+      "grad_norm": 1.3148672580718994,
+      "learning_rate": 0.00019397469130358498,
+      "loss": 0.6648,
+      "step": 1874
+    },
+    {
+      "epoch": 0.13787778513125964,
+      "grad_norm": 1.2228047847747803,
+      "learning_rate": 0.00019396654657774593,
+      "loss": 0.697,
+      "step": 1875
+    },
+    {
+      "epoch": 0.13795131994999632,
+      "grad_norm": 1.0799962282180786,
+      "learning_rate": 0.0001939583965220252,
+      "loss": 0.6829,
+      "step": 1876
+    },
+    {
+      "epoch": 0.138024854768733,
+      "grad_norm": 1.2830586433410645,
+      "learning_rate": 0.00019395024113688503,
+      "loss": 0.6843,
+      "step": 1877
+    },
+    {
+      "epoch": 0.13809838958746967,
+      "grad_norm": 0.9344533085823059,
+      "learning_rate": 0.00019394208042278807,
+      "loss": 0.6829,
+      "step": 1878
+    },
+    {
+      "epoch": 0.13817192440620635,
+      "grad_norm": 1.8067911863327026,
+      "learning_rate": 0.00019393391438019713,
+      "loss": 0.6839,
+      "step": 1879
+    },
+    {
+      "epoch": 0.13824545922494302,
+      "grad_norm": 1.0270551443099976,
+      "learning_rate": 0.00019392574300957543,
+      "loss": 0.6891,
+      "step": 1880
+    },
+    {
+      "epoch": 0.13831899404367967,
+      "grad_norm": 1.7363615036010742,
+      "learning_rate": 0.00019391756631138648,
+      "loss": 0.669,
+      "step": 1881
+    },
+    {
+      "epoch": 0.13839252886241635,
+      "grad_norm": 1.1869711875915527,
+      "learning_rate": 0.00019390938428609405,
+      "loss": 0.7,
+      "step": 1882
+    },
+    {
+      "epoch": 0.13846606368115302,
+      "grad_norm": 1.9604946374893188,
+      "learning_rate": 0.00019390119693416224,
+      "loss": 0.6731,
+      "step": 1883
+    },
+    {
+      "epoch": 0.1385395984998897,
+      "grad_norm": 1.471333384513855,
+      "learning_rate": 0.00019389300425605546,
+      "loss": 0.7187,
+      "step": 1884
+    },
+    {
+      "epoch": 0.13861313331862637,
+      "grad_norm": 1.8511942625045776,
+      "learning_rate": 0.00019388480625223836,
+      "loss": 0.7048,
+      "step": 1885
+    },
+    {
+      "epoch": 0.13868666813736305,
+      "grad_norm": 1.8079230785369873,
+      "learning_rate": 0.000193876602923176,
+      "loss": 0.6557,
+      "step": 1886
+    },
+    {
+      "epoch": 0.13876020295609973,
+      "grad_norm": 1.2477995157241821,
+      "learning_rate": 0.0001938683942693336,
+      "loss": 0.6431,
+      "step": 1887
+    },
+    {
+      "epoch": 0.13883373777483637,
+      "grad_norm": 1.4960044622421265,
+      "learning_rate": 0.00019386018029117686,
+      "loss": 0.6699,
+      "step": 1888
+    },
+    {
+      "epoch": 0.13890727259357305,
+      "grad_norm": 1.4818781614303589,
+      "learning_rate": 0.00019385196098917163,
+      "loss": 0.6947,
+      "step": 1889
+    },
+    {
+      "epoch": 0.13898080741230973,
+      "grad_norm": 1.0681113004684448,
+      "learning_rate": 0.00019384373636378412,
+      "loss": 0.6744,
+      "step": 1890
+    },
+    {
+      "epoch": 0.1390543422310464,
+      "grad_norm": 1.8317179679870605,
+      "learning_rate": 0.00019383550641548082,
+      "loss": 0.6728,
+      "step": 1891
+    },
+    {
+      "epoch": 0.13912787704978308,
+      "grad_norm": 1.985023021697998,
+      "learning_rate": 0.00019382727114472862,
+      "loss": 0.7013,
+      "step": 1892
+    },
+    {
+      "epoch": 0.13920141186851975,
+      "grad_norm": 1.1823447942733765,
+      "learning_rate": 0.00019381903055199455,
+      "loss": 0.6906,
+      "step": 1893
+    },
+    {
+      "epoch": 0.13927494668725643,
+      "grad_norm": 1.1670078039169312,
+      "learning_rate": 0.00019381078463774607,
+      "loss": 0.6568,
+      "step": 1894
+    },
+    {
+      "epoch": 0.13934848150599308,
+      "grad_norm": 1.5569177865982056,
+      "learning_rate": 0.00019380253340245087,
+      "loss": 0.6489,
+      "step": 1895
+    },
+    {
+      "epoch": 0.13942201632472975,
+      "grad_norm": 1.265486240386963,
+      "learning_rate": 0.00019379427684657698,
+      "loss": 0.6932,
+      "step": 1896
+    },
+    {
+      "epoch": 0.13949555114346643,
+      "grad_norm": 1.7137454748153687,
+      "learning_rate": 0.00019378601497059274,
+      "loss": 0.6505,
+      "step": 1897
+    },
+    {
+      "epoch": 0.1395690859622031,
+      "grad_norm": 1.5732289552688599,
+      "learning_rate": 0.00019377774777496674,
+      "loss": 0.7018,
+      "step": 1898
+    },
+    {
+      "epoch": 0.13964262078093978,
+      "grad_norm": 1.3844990730285645,
+      "learning_rate": 0.00019376947526016794,
+      "loss": 0.6483,
+      "step": 1899
+    },
+    {
+      "epoch": 0.13971615559967646,
+      "grad_norm": 1.487275242805481,
+      "learning_rate": 0.0001937611974266655,
+      "loss": 0.6688,
+      "step": 1900
+    },
+    {
+      "epoch": 0.13978969041841313,
+      "grad_norm": 1.1824309825897217,
+      "learning_rate": 0.00019375291427492904,
+      "loss": 0.6432,
+      "step": 1901
+    },
+    {
+      "epoch": 0.13986322523714978,
+      "grad_norm": 1.2078161239624023,
+      "learning_rate": 0.0001937446258054283,
+      "loss": 0.6375,
+      "step": 1902
+    },
+    {
+      "epoch": 0.13993676005588646,
+      "grad_norm": 1.2669790983200073,
+      "learning_rate": 0.00019373633201863347,
+      "loss": 0.6514,
+      "step": 1903
+    },
+    {
+      "epoch": 0.14001029487462313,
+      "grad_norm": 0.8740893006324768,
+      "learning_rate": 0.00019372803291501498,
+      "loss": 0.6774,
+      "step": 1904
+    },
+    {
+      "epoch": 0.1400838296933598,
+      "grad_norm": 1.4618330001831055,
+      "learning_rate": 0.00019371972849504352,
+      "loss": 0.6694,
+      "step": 1905
+    },
+    {
+      "epoch": 0.14015736451209648,
+      "grad_norm": 0.9152229428291321,
+      "learning_rate": 0.00019371141875919014,
+      "loss": 0.662,
+      "step": 1906
+    },
+    {
+      "epoch": 0.14023089933083316,
+      "grad_norm": 1.4568893909454346,
+      "learning_rate": 0.00019370310370792624,
+      "loss": 0.7009,
+      "step": 1907
+    },
+    {
+      "epoch": 0.14030443414956983,
+      "grad_norm": 1.0957262516021729,
+      "learning_rate": 0.0001936947833417234,
+      "loss": 0.6792,
+      "step": 1908
+    },
+    {
+      "epoch": 0.14037796896830648,
+      "grad_norm": 1.6712473630905151,
+      "learning_rate": 0.00019368645766105352,
+      "loss": 0.6332,
+      "step": 1909
+    },
+    {
+      "epoch": 0.14045150378704316,
+      "grad_norm": 1.4704976081848145,
+      "learning_rate": 0.00019367812666638894,
+      "loss": 0.6829,
+      "step": 1910
+    },
+    {
+      "epoch": 0.14052503860577983,
+      "grad_norm": 1.230023741722107,
+      "learning_rate": 0.00019366979035820214,
+      "loss": 0.6548,
+      "step": 1911
+    },
+    {
+      "epoch": 0.1405985734245165,
+      "grad_norm": 1.0626047849655151,
+      "learning_rate": 0.00019366144873696596,
+      "loss": 0.6351,
+      "step": 1912
+    },
+    {
+      "epoch": 0.14067210824325319,
+      "grad_norm": 1.522433876991272,
+      "learning_rate": 0.0001936531018031536,
+      "loss": 0.6752,
+      "step": 1913
+    },
+    {
+      "epoch": 0.14074564306198986,
+      "grad_norm": 0.980289101600647,
+      "learning_rate": 0.00019364474955723844,
+      "loss": 0.6899,
+      "step": 1914
+    },
+    {
+      "epoch": 0.14081917788072654,
+      "grad_norm": 1.8819890022277832,
+      "learning_rate": 0.00019363639199969425,
+      "loss": 0.6728,
+      "step": 1915
+    },
+    {
+      "epoch": 0.14089271269946319,
+      "grad_norm": 1.2429685592651367,
+      "learning_rate": 0.0001936280291309951,
+      "loss": 0.6455,
+      "step": 1916
+    },
+    {
+      "epoch": 0.14096624751819986,
+      "grad_norm": 1.6478451490402222,
+      "learning_rate": 0.00019361966095161534,
+      "loss": 0.6628,
+      "step": 1917
+    },
+    {
+      "epoch": 0.14103978233693654,
+      "grad_norm": 1.5852465629577637,
+      "learning_rate": 0.00019361128746202958,
+      "loss": 0.6511,
+      "step": 1918
+    },
+    {
+      "epoch": 0.1411133171556732,
+      "grad_norm": 1.527496337890625,
+      "learning_rate": 0.00019360290866271286,
+      "loss": 0.6504,
+      "step": 1919
+    },
+    {
+      "epoch": 0.1411868519744099,
+      "grad_norm": 1.2857038974761963,
+      "learning_rate": 0.0001935945245541403,
+      "loss": 0.6747,
+      "step": 1920
+    },
+    {
+      "epoch": 0.14126038679314656,
+      "grad_norm": 1.4707939624786377,
+      "learning_rate": 0.00019358613513678757,
+      "loss": 0.6981,
+      "step": 1921
+    },
+    {
+      "epoch": 0.14133392161188324,
+      "grad_norm": 1.3555622100830078,
+      "learning_rate": 0.00019357774041113053,
+      "loss": 0.6861,
+      "step": 1922
+    },
+    {
+      "epoch": 0.1414074564306199,
+      "grad_norm": 1.7556695938110352,
+      "learning_rate": 0.00019356934037764525,
+      "loss": 0.6559,
+      "step": 1923
+    },
+    {
+      "epoch": 0.14148099124935656,
+      "grad_norm": 1.816557765007019,
+      "learning_rate": 0.00019356093503680824,
+      "loss": 0.679,
+      "step": 1924
+    },
+    {
+      "epoch": 0.14155452606809324,
+      "grad_norm": 1.1277501583099365,
+      "learning_rate": 0.00019355252438909627,
+      "loss": 0.65,
+      "step": 1925
+    },
+    {
+      "epoch": 0.14162806088682992,
+      "grad_norm": 1.168056845664978,
+      "learning_rate": 0.0001935441084349864,
+      "loss": 0.6582,
+      "step": 1926
+    },
+    {
+      "epoch": 0.1417015957055666,
+      "grad_norm": 1.1983609199523926,
+      "learning_rate": 0.00019353568717495593,
+      "loss": 0.6645,
+      "step": 1927
+    },
+    {
+      "epoch": 0.14177513052430327,
+      "grad_norm": 1.0412496328353882,
+      "learning_rate": 0.0001935272606094826,
+      "loss": 0.6694,
+      "step": 1928
+    },
+    {
+      "epoch": 0.14184866534303994,
+      "grad_norm": 1.5185648202896118,
+      "learning_rate": 0.00019351882873904438,
+      "loss": 0.6931,
+      "step": 1929
+    },
+    {
+      "epoch": 0.1419222001617766,
+      "grad_norm": 1.3279929161071777,
+      "learning_rate": 0.00019351039156411944,
+      "loss": 0.6873,
+      "step": 1930
+    },
+    {
+      "epoch": 0.14199573498051327,
+      "grad_norm": 1.4379587173461914,
+      "learning_rate": 0.00019350194908518642,
+      "loss": 0.6978,
+      "step": 1931
+    },
+    {
+      "epoch": 0.14206926979924994,
+      "grad_norm": 1.0025296211242676,
+      "learning_rate": 0.0001934935013027242,
+      "loss": 0.6898,
+      "step": 1932
+    },
+    {
+      "epoch": 0.14214280461798662,
+      "grad_norm": 1.8637322187423706,
+      "learning_rate": 0.00019348504821721187,
+      "loss": 0.6606,
+      "step": 1933
+    },
+    {
+      "epoch": 0.1422163394367233,
+      "grad_norm": 1.491957664489746,
+      "learning_rate": 0.000193476589829129,
+      "loss": 0.6718,
+      "step": 1934
+    },
+    {
+      "epoch": 0.14228987425545997,
+      "grad_norm": 1.4069849252700806,
+      "learning_rate": 0.00019346812613895526,
+      "loss": 0.6621,
+      "step": 1935
+    },
+    {
+      "epoch": 0.14236340907419665,
+      "grad_norm": 1.2662701606750488,
+      "learning_rate": 0.0001934596571471708,
+      "loss": 0.6739,
+      "step": 1936
+    },
+    {
+      "epoch": 0.1424369438929333,
+      "grad_norm": 1.1209158897399902,
+      "learning_rate": 0.00019345118285425594,
+      "loss": 0.6528,
+      "step": 1937
+    },
+    {
+      "epoch": 0.14251047871166997,
+      "grad_norm": 1.0474029779434204,
+      "learning_rate": 0.00019344270326069135,
+      "loss": 0.6463,
+      "step": 1938
+    },
+    {
+      "epoch": 0.14258401353040664,
+      "grad_norm": 1.6811782121658325,
+      "learning_rate": 0.00019343421836695807,
+      "loss": 0.6896,
+      "step": 1939
+    },
+    {
+      "epoch": 0.14265754834914332,
+      "grad_norm": 1.4078545570373535,
+      "learning_rate": 0.00019342572817353725,
+      "loss": 0.6796,
+      "step": 1940
+    },
+    {
+      "epoch": 0.14273108316788,
+      "grad_norm": 1.6542562246322632,
+      "learning_rate": 0.00019341723268091057,
+      "loss": 0.6716,
+      "step": 1941
+    },
+    {
+      "epoch": 0.14280461798661667,
+      "grad_norm": 1.4729952812194824,
+      "learning_rate": 0.00019340873188955986,
+      "loss": 0.6436,
+      "step": 1942
+    },
+    {
+      "epoch": 0.14287815280535335,
+      "grad_norm": 1.4922895431518555,
+      "learning_rate": 0.00019340022579996728,
+      "loss": 0.653,
+      "step": 1943
+    },
+    {
+      "epoch": 0.14295168762409,
+      "grad_norm": 1.4589099884033203,
+      "learning_rate": 0.00019339171441261533,
+      "loss": 0.6519,
+      "step": 1944
+    },
+    {
+      "epoch": 0.14302522244282667,
+      "grad_norm": 1.1470392942428589,
+      "learning_rate": 0.00019338319772798683,
+      "loss": 0.6324,
+      "step": 1945
+    },
+    {
+      "epoch": 0.14309875726156335,
+      "grad_norm": 1.1291236877441406,
+      "learning_rate": 0.00019337467574656472,
+      "loss": 0.6784,
+      "step": 1946
+    },
+    {
+      "epoch": 0.14317229208030002,
+      "grad_norm": 1.4124726057052612,
+      "learning_rate": 0.00019336614846883254,
+      "loss": 0.6864,
+      "step": 1947
+    },
+    {
+      "epoch": 0.1432458268990367,
+      "grad_norm": 1.3456716537475586,
+      "learning_rate": 0.00019335761589527387,
+      "loss": 0.6403,
+      "step": 1948
+    },
+    {
+      "epoch": 0.14331936171777337,
+      "grad_norm": 1.3948644399642944,
+      "learning_rate": 0.00019334907802637266,
+      "loss": 0.6861,
+      "step": 1949
+    },
+    {
+      "epoch": 0.14339289653651005,
+      "grad_norm": 1.3900080919265747,
+      "learning_rate": 0.00019334053486261325,
+      "loss": 0.6851,
+      "step": 1950
+    },
+    {
+      "epoch": 0.1434664313552467,
+      "grad_norm": 1.1498348712921143,
+      "learning_rate": 0.00019333198640448021,
+      "loss": 0.6543,
+      "step": 1951
+    },
+    {
+      "epoch": 0.14353996617398337,
+      "grad_norm": 1.1199016571044922,
+      "learning_rate": 0.00019332343265245837,
+      "loss": 0.6301,
+      "step": 1952
+    },
+    {
+      "epoch": 0.14361350099272005,
+      "grad_norm": 1.5158535242080688,
+      "learning_rate": 0.00019331487360703295,
+      "loss": 0.6672,
+      "step": 1953
+    },
+    {
+      "epoch": 0.14368703581145673,
+      "grad_norm": 1.1664992570877075,
+      "learning_rate": 0.00019330630926868946,
+      "loss": 0.6263,
+      "step": 1954
+    },
+    {
+      "epoch": 0.1437605706301934,
+      "grad_norm": 1.695796251296997,
+      "learning_rate": 0.0001932977396379136,
+      "loss": 0.6143,
+      "step": 1955
+    },
+    {
+      "epoch": 0.14383410544893008,
+      "grad_norm": 1.3502144813537598,
+      "learning_rate": 0.0001932891647151915,
+      "loss": 0.6641,
+      "step": 1956
+    },
+    {
+      "epoch": 0.14390764026766675,
+      "grad_norm": 1.2197703123092651,
+      "learning_rate": 0.0001932805845010095,
+      "loss": 0.6458,
+      "step": 1957
+    },
+    {
+      "epoch": 0.1439811750864034,
+      "grad_norm": 1.24602210521698,
+      "learning_rate": 0.00019327199899585434,
+      "loss": 0.6247,
+      "step": 1958
+    },
+    {
+      "epoch": 0.14405470990514008,
+      "grad_norm": 1.1714051961898804,
+      "learning_rate": 0.00019326340820021294,
+      "loss": 0.6688,
+      "step": 1959
+    },
+    {
+      "epoch": 0.14412824472387675,
+      "grad_norm": 0.9701496362686157,
+      "learning_rate": 0.0001932548121145726,
+      "loss": 0.6328,
+      "step": 1960
+    },
+    {
+      "epoch": 0.14420177954261343,
+      "grad_norm": 1.4254060983657837,
+      "learning_rate": 0.00019324621073942092,
+      "loss": 0.6684,
+      "step": 1961
+    },
+    {
+      "epoch": 0.1442753143613501,
+      "grad_norm": 1.1160047054290771,
+      "learning_rate": 0.00019323760407524576,
+      "loss": 0.6639,
+      "step": 1962
+    },
+    {
+      "epoch": 0.14434884918008678,
+      "grad_norm": 1.587978482246399,
+      "learning_rate": 0.0001932289921225353,
+      "loss": 0.6924,
+      "step": 1963
+    },
+    {
+      "epoch": 0.14442238399882346,
+      "grad_norm": 1.151463270187378,
+      "learning_rate": 0.00019322037488177801,
+      "loss": 0.6331,
+      "step": 1964
+    },
+    {
+      "epoch": 0.1444959188175601,
+      "grad_norm": 1.3673588037490845,
+      "learning_rate": 0.00019321175235346267,
+      "loss": 0.6586,
+      "step": 1965
+    },
+    {
+      "epoch": 0.14456945363629678,
+      "grad_norm": 1.035649061203003,
+      "learning_rate": 0.00019320312453807836,
+      "loss": 0.6515,
+      "step": 1966
+    },
+    {
+      "epoch": 0.14464298845503346,
+      "grad_norm": 1.4371039867401123,
+      "learning_rate": 0.0001931944914361145,
+      "loss": 0.6626,
+      "step": 1967
+    },
+    {
+      "epoch": 0.14471652327377013,
+      "grad_norm": 0.9067654013633728,
+      "learning_rate": 0.00019318585304806073,
+      "loss": 0.625,
+      "step": 1968
+    },
+    {
+      "epoch": 0.1447900580925068,
+      "grad_norm": 1.3499383926391602,
+      "learning_rate": 0.000193177209374407,
+      "loss": 0.6678,
+      "step": 1969
+    },
+    {
+      "epoch": 0.14486359291124348,
+      "grad_norm": 0.9081353545188904,
+      "learning_rate": 0.00019316856041564365,
+      "loss": 0.641,
+      "step": 1970
+    },
+    {
+      "epoch": 0.14493712772998016,
+      "grad_norm": 1.5444923639297485,
+      "learning_rate": 0.00019315990617226118,
+      "loss": 0.6359,
+      "step": 1971
+    },
+    {
+      "epoch": 0.1450106625487168,
+      "grad_norm": 1.1230003833770752,
+      "learning_rate": 0.00019315124664475057,
+      "loss": 0.6209,
+      "step": 1972
+    },
+    {
+      "epoch": 0.14508419736745348,
+      "grad_norm": 1.635380506515503,
+      "learning_rate": 0.00019314258183360293,
+      "loss": 0.6119,
+      "step": 1973
+    },
+    {
+      "epoch": 0.14515773218619016,
+      "grad_norm": 1.600946068763733,
+      "learning_rate": 0.00019313391173930974,
+      "loss": 0.6587,
+      "step": 1974
+    },
+    {
+      "epoch": 0.14523126700492683,
+      "grad_norm": 1.1104121208190918,
+      "learning_rate": 0.0001931252363623628,
+      "loss": 0.6331,
+      "step": 1975
+    },
+    {
+      "epoch": 0.1453048018236635,
+      "grad_norm": 0.9078276753425598,
+      "learning_rate": 0.00019311655570325415,
+      "loss": 0.6461,
+      "step": 1976
+    },
+    {
+      "epoch": 0.14537833664240019,
+      "grad_norm": 1.1584452390670776,
+      "learning_rate": 0.0001931078697624762,
+      "loss": 0.6484,
+      "step": 1977
+    },
+    {
+      "epoch": 0.14545187146113686,
+      "grad_norm": 0.7045577764511108,
+      "learning_rate": 0.00019309917854052163,
+      "loss": 0.6585,
+      "step": 1978
+    },
+    {
+      "epoch": 0.1455254062798735,
+      "grad_norm": 1.2014377117156982,
+      "learning_rate": 0.00019309048203788341,
+      "loss": 0.675,
+      "step": 1979
+    },
+    {
+      "epoch": 0.14559894109861019,
+      "grad_norm": 0.5838351845741272,
+      "learning_rate": 0.0001930817802550548,
+      "loss": 0.6443,
+      "step": 1980
+    },
+    {
+      "epoch": 0.14567247591734686,
+      "grad_norm": 1.3402981758117676,
+      "learning_rate": 0.00019307307319252936,
+      "loss": 0.6251,
+      "step": 1981
+    },
+    {
+      "epoch": 0.14574601073608354,
+      "grad_norm": 0.6302463412284851,
+      "learning_rate": 0.00019306436085080102,
+      "loss": 0.6486,
+      "step": 1982
+    },
+    {
+      "epoch": 0.1458195455548202,
+      "grad_norm": 1.4268893003463745,
+      "learning_rate": 0.00019305564323036388,
+      "loss": 0.6376,
+      "step": 1983
+    },
+    {
+      "epoch": 0.1458930803735569,
+      "grad_norm": 0.7537228465080261,
+      "learning_rate": 0.00019304692033171245,
+      "loss": 0.6365,
+      "step": 1984
+    },
+    {
+      "epoch": 0.14596661519229356,
+      "grad_norm": 1.5324152708053589,
+      "learning_rate": 0.00019303819215534153,
+      "loss": 0.6068,
+      "step": 1985
+    },
+    {
+      "epoch": 0.1460401500110302,
+      "grad_norm": 1.1278643608093262,
+      "learning_rate": 0.00019302945870174615,
+      "loss": 0.6452,
+      "step": 1986
+    },
+    {
+      "epoch": 0.1461136848297669,
+      "grad_norm": 1.565880537033081,
+      "learning_rate": 0.0001930207199714217,
+      "loss": 0.6099,
+      "step": 1987
+    },
+    {
+      "epoch": 0.14618721964850356,
+      "grad_norm": 1.3076791763305664,
+      "learning_rate": 0.00019301197596486384,
+      "loss": 0.6484,
+      "step": 1988
+    },
+    {
+      "epoch": 0.14626075446724024,
+      "grad_norm": 1.3162649869918823,
+      "learning_rate": 0.00019300322668256854,
+      "loss": 0.6431,
+      "step": 1989
+    },
+    {
+      "epoch": 0.14633428928597692,
+      "grad_norm": 1.3310368061065674,
+      "learning_rate": 0.00019299447212503207,
+      "loss": 0.6526,
+      "step": 1990
+    },
+    {
+      "epoch": 0.1464078241047136,
+      "grad_norm": 1.2591893672943115,
+      "learning_rate": 0.00019298571229275102,
+      "loss": 0.6299,
+      "step": 1991
+    },
+    {
+      "epoch": 0.14648135892345027,
+      "grad_norm": 1.1943774223327637,
+      "learning_rate": 0.00019297694718622223,
+      "loss": 0.6371,
+      "step": 1992
+    },
+    {
+      "epoch": 0.14655489374218691,
+      "grad_norm": 1.196664810180664,
+      "learning_rate": 0.0001929681768059429,
+      "loss": 0.6726,
+      "step": 1993
+    },
+    {
+      "epoch": 0.1466284285609236,
+      "grad_norm": 1.2544530630111694,
+      "learning_rate": 0.00019295940115241042,
+      "loss": 0.5913,
+      "step": 1994
+    },
+    {
+      "epoch": 0.14670196337966027,
+      "grad_norm": 0.7166563868522644,
+      "learning_rate": 0.0001929506202261226,
+      "loss": 0.6398,
+      "step": 1995
+    },
+    {
+      "epoch": 0.14677549819839694,
+      "grad_norm": 1.396506905555725,
+      "learning_rate": 0.00019294183402757752,
+      "loss": 0.6614,
+      "step": 1996
+    },
+    {
+      "epoch": 0.14684903301713362,
+      "grad_norm": 0.7974982857704163,
+      "learning_rate": 0.00019293304255727355,
+      "loss": 0.6252,
+      "step": 1997
+    },
+    {
+      "epoch": 0.1469225678358703,
+      "grad_norm": 0.967223048210144,
+      "learning_rate": 0.0001929242458157093,
+      "loss": 0.6523,
+      "step": 1998
+    },
+    {
+      "epoch": 0.14699610265460697,
+      "grad_norm": 1.3762701749801636,
+      "learning_rate": 0.00019291544380338376,
+      "loss": 0.6375,
+      "step": 1999
+    },
+    {
+      "epoch": 0.14706963747334362,
+      "grad_norm": 0.8087677955627441,
+      "learning_rate": 0.00019290663652079623,
+      "loss": 0.6434,
+      "step": 2000
+    },
+    {
+      "epoch": 0.1471431722920803,
+      "grad_norm": 1.816511869430542,
+      "learning_rate": 0.00019289782396844621,
+      "loss": 0.652,
+      "step": 2001
+    },
+    {
+      "epoch": 0.14721670711081697,
+      "grad_norm": 1.2142229080200195,
+      "learning_rate": 0.00019288900614683357,
+      "loss": 0.6208,
+      "step": 2002
+    },
+    {
+      "epoch": 0.14729024192955364,
+      "grad_norm": 2.33131742477417,
+      "learning_rate": 0.00019288018305645848,
+      "loss": 0.6243,
+      "step": 2003
+    },
+    {
+      "epoch": 0.14736377674829032,
+      "grad_norm": 2.3979601860046387,
+      "learning_rate": 0.00019287135469782136,
+      "loss": 0.6253,
+      "step": 2004
+    },
+    {
+      "epoch": 0.147437311567027,
+      "grad_norm": 0.8409308791160583,
+      "learning_rate": 0.00019286252107142301,
+      "loss": 0.6214,
+      "step": 2005
+    },
+    {
+      "epoch": 0.14751084638576367,
+      "grad_norm": 1.9376274347305298,
+      "learning_rate": 0.00019285368217776444,
+      "loss": 0.6415,
+      "step": 2006
+    },
+    {
+      "epoch": 0.14758438120450032,
+      "grad_norm": 1.1347136497497559,
+      "learning_rate": 0.00019284483801734707,
+      "loss": 0.6291,
+      "step": 2007
+    },
+    {
+      "epoch": 0.147657916023237,
+      "grad_norm": 1.4847608804702759,
+      "learning_rate": 0.0001928359885906725,
+      "loss": 0.6726,
+      "step": 2008
+    },
+    {
+      "epoch": 0.14773145084197367,
+      "grad_norm": 1.4031270742416382,
+      "learning_rate": 0.00019282713389824263,
+      "loss": 0.6581,
+      "step": 2009
+    },
+    {
+      "epoch": 0.14780498566071035,
+      "grad_norm": 1.342368245124817,
+      "learning_rate": 0.0001928182739405598,
+      "loss": 0.6302,
+      "step": 2010
+    },
+    {
+      "epoch": 0.14787852047944702,
+      "grad_norm": 1.2550064325332642,
+      "learning_rate": 0.0001928094087181265,
+      "loss": 0.6392,
+      "step": 2011
+    },
+    {
+      "epoch": 0.1479520552981837,
+      "grad_norm": 1.236204743385315,
+      "learning_rate": 0.00019280053823144563,
+      "loss": 0.6373,
+      "step": 2012
+    },
+    {
+      "epoch": 0.14802559011692037,
+      "grad_norm": 0.9144555330276489,
+      "learning_rate": 0.00019279166248102026,
+      "loss": 0.6319,
+      "step": 2013
+    },
+    {
+      "epoch": 0.14809912493565702,
+      "grad_norm": 0.9581331014633179,
+      "learning_rate": 0.00019278278146735387,
+      "loss": 0.6146,
+      "step": 2014
+    },
+    {
+      "epoch": 0.1481726597543937,
+      "grad_norm": 1.0511705875396729,
+      "learning_rate": 0.0001927738951909502,
+      "loss": 0.62,
+      "step": 2015
+    },
+    {
+      "epoch": 0.14824619457313037,
+      "grad_norm": 0.8784657120704651,
+      "learning_rate": 0.0001927650036523133,
+      "loss": 0.6003,
+      "step": 2016
+    },
+    {
+      "epoch": 0.14831972939186705,
+      "grad_norm": 1.1746193170547485,
+      "learning_rate": 0.0001927561068519475,
+      "loss": 0.6202,
+      "step": 2017
+    },
+    {
+      "epoch": 0.14839326421060373,
+      "grad_norm": 0.7868152856826782,
+      "learning_rate": 0.0001927472047903574,
+      "loss": 0.6333,
+      "step": 2018
+    },
+    {
+      "epoch": 0.1484667990293404,
+      "grad_norm": 1.2505329847335815,
+      "learning_rate": 0.00019273829746804797,
+      "loss": 0.6188,
+      "step": 2019
+    },
+    {
+      "epoch": 0.14854033384807708,
+      "grad_norm": 0.7318904399871826,
+      "learning_rate": 0.00019272938488552446,
+      "loss": 0.6405,
+      "step": 2020
+    },
+    {
+      "epoch": 0.14861386866681373,
+      "grad_norm": 1.4727258682250977,
+      "learning_rate": 0.00019272046704329237,
+      "loss": 0.6551,
+      "step": 2021
+    },
+    {
+      "epoch": 0.1486874034855504,
+      "grad_norm": 0.966797947883606,
+      "learning_rate": 0.00019271154394185752,
+      "loss": 0.6407,
+      "step": 2022
+    },
+    {
+      "epoch": 0.14876093830428708,
+      "grad_norm": 1.7617501020431519,
+      "learning_rate": 0.00019270261558172605,
+      "loss": 0.657,
+      "step": 2023
+    },
+    {
+      "epoch": 0.14883447312302375,
+      "grad_norm": 1.3796648979187012,
+      "learning_rate": 0.00019269368196340441,
+      "loss": 0.6264,
+      "step": 2024
+    },
+    {
+      "epoch": 0.14890800794176043,
+      "grad_norm": 1.2927241325378418,
+      "learning_rate": 0.0001926847430873993,
+      "loss": 0.6163,
+      "step": 2025
+    },
+    {
+      "epoch": 0.1489815427604971,
+      "grad_norm": 1.1493231058120728,
+      "learning_rate": 0.00019267579895421773,
+      "loss": 0.6122,
+      "step": 2026
+    },
+    {
+      "epoch": 0.14905507757923378,
+      "grad_norm": 1.51605224609375,
+      "learning_rate": 0.0001926668495643671,
+      "loss": 0.6187,
+      "step": 2027
+    },
+    {
+      "epoch": 0.14912861239797043,
+      "grad_norm": 1.0358598232269287,
+      "learning_rate": 0.0001926578949183549,
+      "loss": 0.5839,
+      "step": 2028
+    },
+    {
+      "epoch": 0.1492021472167071,
+      "grad_norm": 1.7470245361328125,
+      "learning_rate": 0.00019264893501668917,
+      "loss": 0.6155,
+      "step": 2029
+    },
+    {
+      "epoch": 0.14927568203544378,
+      "grad_norm": 1.3408371210098267,
+      "learning_rate": 0.00019263996985987802,
+      "loss": 0.6103,
+      "step": 2030
+    },
+    {
+      "epoch": 0.14934921685418046,
+      "grad_norm": 1.508851408958435,
+      "learning_rate": 0.00019263099944843003,
+      "loss": 0.6605,
+      "step": 2031
+    },
+    {
+      "epoch": 0.14942275167291713,
+      "grad_norm": 1.3566837310791016,
+      "learning_rate": 0.00019262202378285398,
+      "loss": 0.5995,
+      "step": 2032
+    },
+    {
+      "epoch": 0.1494962864916538,
+      "grad_norm": 1.4621951580047607,
+      "learning_rate": 0.00019261304286365905,
+      "loss": 0.6351,
+      "step": 2033
+    },
+    {
+      "epoch": 0.14956982131039048,
+      "grad_norm": 1.0156264305114746,
+      "learning_rate": 0.00019260405669135455,
+      "loss": 0.6295,
+      "step": 2034
+    },
+    {
+      "epoch": 0.14964335612912713,
+      "grad_norm": 1.5580650568008423,
+      "learning_rate": 0.00019259506526645023,
+      "loss": 0.6216,
+      "step": 2035
+    },
+    {
+      "epoch": 0.1497168909478638,
+      "grad_norm": 0.9025373458862305,
+      "learning_rate": 0.00019258606858945607,
+      "loss": 0.6159,
+      "step": 2036
+    },
+    {
+      "epoch": 0.14979042576660048,
+      "grad_norm": 1.6752684116363525,
+      "learning_rate": 0.0001925770666608824,
+      "loss": 0.6083,
+      "step": 2037
+    },
+    {
+      "epoch": 0.14986396058533716,
+      "grad_norm": 1.1833577156066895,
+      "learning_rate": 0.00019256805948123982,
+      "loss": 0.6193,
+      "step": 2038
+    },
+    {
+      "epoch": 0.14993749540407383,
+      "grad_norm": 1.678662896156311,
+      "learning_rate": 0.0001925590470510392,
+      "loss": 0.6049,
+      "step": 2039
+    },
+    {
+      "epoch": 0.1500110302228105,
+      "grad_norm": 1.498367190361023,
+      "learning_rate": 0.00019255002937079176,
+      "loss": 0.605,
+      "step": 2040
+    },
+    {
+      "epoch": 0.15008456504154719,
+      "grad_norm": 1.245611548423767,
+      "learning_rate": 0.00019254100644100895,
+      "loss": 0.6089,
+      "step": 2041
+    },
+    {
+      "epoch": 0.15015809986028383,
+      "grad_norm": 1.0867477655410767,
+      "learning_rate": 0.00019253197826220263,
+      "loss": 0.6555,
+      "step": 2042
+    },
+    {
+      "epoch": 0.1502316346790205,
+      "grad_norm": 1.4593067169189453,
+      "learning_rate": 0.00019252294483488488,
+      "loss": 0.6369,
+      "step": 2043
+    },
+    {
+      "epoch": 0.15030516949775719,
+      "grad_norm": 1.2390090227127075,
+      "learning_rate": 0.000192513906159568,
+      "loss": 0.6489,
+      "step": 2044
+    },
+    {
+      "epoch": 0.15037870431649386,
+      "grad_norm": 1.5560941696166992,
+      "learning_rate": 0.00019250486223676476,
+      "loss": 0.5725,
+      "step": 2045
+    },
+    {
+      "epoch": 0.15045223913523054,
+      "grad_norm": 1.442752718925476,
+      "learning_rate": 0.0001924958130669881,
+      "loss": 0.5958,
+      "step": 2046
+    },
+    {
+      "epoch": 0.1505257739539672,
+      "grad_norm": 1.04511559009552,
+      "learning_rate": 0.0001924867586507513,
+      "loss": 0.6176,
+      "step": 2047
+    },
+    {
+      "epoch": 0.1505993087727039,
+      "grad_norm": 1.163986325263977,
+      "learning_rate": 0.00019247769898856793,
+      "loss": 0.6046,
+      "step": 2048
+    },
+    {
+      "epoch": 0.15067284359144054,
+      "grad_norm": 1.1305537223815918,
+      "learning_rate": 0.00019246863408095192,
+      "loss": 0.6202,
+      "step": 2049
+    },
+    {
+      "epoch": 0.1507463784101772,
+      "grad_norm": 0.8695442080497742,
+      "learning_rate": 0.00019245956392841739,
+      "loss": 0.6692,
+      "step": 2050
+    },
+    {
+      "epoch": 0.1508199132289139,
+      "grad_norm": 1.3421311378479004,
+      "learning_rate": 0.00019245048853147877,
+      "loss": 0.625,
+      "step": 2051
+    },
+    {
+      "epoch": 0.15089344804765056,
+      "grad_norm": 0.9166889190673828,
+      "learning_rate": 0.0001924414078906509,
+      "loss": 0.6546,
+      "step": 2052
+    },
+    {
+      "epoch": 0.15096698286638724,
+      "grad_norm": 1.7107834815979004,
+      "learning_rate": 0.00019243232200644883,
+      "loss": 0.6055,
+      "step": 2053
+    },
+    {
+      "epoch": 0.15104051768512392,
+      "grad_norm": 1.3685253858566284,
+      "learning_rate": 0.0001924232308793879,
+      "loss": 0.6786,
+      "step": 2054
+    },
+    {
+      "epoch": 0.1511140525038606,
+      "grad_norm": 1.6002603769302368,
+      "learning_rate": 0.0001924141345099838,
+      "loss": 0.6158,
+      "step": 2055
+    },
+    {
+      "epoch": 0.15118758732259724,
+      "grad_norm": 1.4885495901107788,
+      "learning_rate": 0.00019240503289875242,
+      "loss": 0.6179,
+      "step": 2056
+    },
+    {
+      "epoch": 0.15126112214133391,
+      "grad_norm": 1.106736183166504,
+      "learning_rate": 0.00019239592604621008,
+      "loss": 0.5985,
+      "step": 2057
+    },
+    {
+      "epoch": 0.1513346569600706,
+      "grad_norm": 1.1397817134857178,
+      "learning_rate": 0.0001923868139528733,
+      "loss": 0.5629,
+      "step": 2058
+    },
+    {
+      "epoch": 0.15140819177880727,
+      "grad_norm": 1.4271053075790405,
+      "learning_rate": 0.00019237769661925892,
+      "loss": 0.6171,
+      "step": 2059
+    },
+    {
+      "epoch": 0.15148172659754394,
+      "grad_norm": 1.210062026977539,
+      "learning_rate": 0.00019236857404588412,
+      "loss": 0.6473,
+      "step": 2060
+    },
+    {
+      "epoch": 0.15155526141628062,
+      "grad_norm": 1.4140828847885132,
+      "learning_rate": 0.00019235944623326634,
+      "loss": 0.6128,
+      "step": 2061
+    },
+    {
+      "epoch": 0.1516287962350173,
+      "grad_norm": 1.4777034521102905,
+      "learning_rate": 0.00019235031318192325,
+      "loss": 0.6048,
+      "step": 2062
+    },
+    {
+      "epoch": 0.15170233105375394,
+      "grad_norm": 1.1547625064849854,
+      "learning_rate": 0.00019234117489237294,
+      "loss": 0.6237,
+      "step": 2063
+    },
+    {
+      "epoch": 0.15177586587249062,
+      "grad_norm": 0.9958681464195251,
+      "learning_rate": 0.00019233203136513377,
+      "loss": 0.6235,
+      "step": 2064
+    },
+    {
+      "epoch": 0.1518494006912273,
+      "grad_norm": 1.5366793870925903,
+      "learning_rate": 0.0001923228826007243,
+      "loss": 0.6073,
+      "step": 2065
+    },
+    {
+      "epoch": 0.15192293550996397,
+      "grad_norm": 1.3143858909606934,
+      "learning_rate": 0.00019231372859966354,
+      "loss": 0.609,
+      "step": 2066
+    },
+    {
+      "epoch": 0.15199647032870064,
+      "grad_norm": 1.483809232711792,
+      "learning_rate": 0.00019230456936247066,
+      "loss": 0.6176,
+      "step": 2067
+    },
+    {
+      "epoch": 0.15207000514743732,
+      "grad_norm": 1.4432401657104492,
+      "learning_rate": 0.00019229540488966519,
+      "loss": 0.6044,
+      "step": 2068
+    },
+    {
+      "epoch": 0.152143539966174,
+      "grad_norm": 1.0988813638687134,
+      "learning_rate": 0.00019228623518176692,
+      "loss": 0.6202,
+      "step": 2069
+    },
+    {
+      "epoch": 0.15221707478491064,
+      "grad_norm": 1.1079120635986328,
+      "learning_rate": 0.00019227706023929603,
+      "loss": 0.6006,
+      "step": 2070
+    },
+    {
+      "epoch": 0.15229060960364732,
+      "grad_norm": 1.291090726852417,
+      "learning_rate": 0.0001922678800627729,
+      "loss": 0.618,
+      "step": 2071
+    },
+    {
+      "epoch": 0.152364144422384,
+      "grad_norm": 1.0627764463424683,
+      "learning_rate": 0.00019225869465271824,
+      "loss": 0.6186,
+      "step": 2072
+    },
+    {
+      "epoch": 0.15243767924112067,
+      "grad_norm": 1.4403389692306519,
+      "learning_rate": 0.00019224950400965302,
+      "loss": 0.6063,
+      "step": 2073
+    },
+    {
+      "epoch": 0.15251121405985735,
+      "grad_norm": 1.2926967144012451,
+      "learning_rate": 0.0001922403081340986,
+      "loss": 0.617,
+      "step": 2074
+    },
+    {
+      "epoch": 0.15258474887859402,
+      "grad_norm": 1.2764631509780884,
+      "learning_rate": 0.00019223110702657658,
+      "loss": 0.5981,
+      "step": 2075
+    },
+    {
+      "epoch": 0.1526582836973307,
+      "grad_norm": 1.2088043689727783,
+      "learning_rate": 0.00019222190068760882,
+      "loss": 0.6047,
+      "step": 2076
+    },
+    {
+      "epoch": 0.15273181851606735,
+      "grad_norm": 1.2906941175460815,
+      "learning_rate": 0.00019221268911771752,
+      "loss": 0.6133,
+      "step": 2077
+    },
+    {
+      "epoch": 0.15280535333480402,
+      "grad_norm": 0.9755882620811462,
+      "learning_rate": 0.00019220347231742516,
+      "loss": 0.6086,
+      "step": 2078
+    },
+    {
+      "epoch": 0.1528788881535407,
+      "grad_norm": 1.279767632484436,
+      "learning_rate": 0.00019219425028725457,
+      "loss": 0.5872,
+      "step": 2079
+    },
+    {
+      "epoch": 0.15295242297227737,
+      "grad_norm": 1.0299097299575806,
+      "learning_rate": 0.0001921850230277288,
+      "loss": 0.6094,
+      "step": 2080
+    },
+    {
+      "epoch": 0.15302595779101405,
+      "grad_norm": 1.7425670623779297,
+      "learning_rate": 0.00019217579053937124,
+      "loss": 0.6473,
+      "step": 2081
+    },
+    {
+      "epoch": 0.15309949260975073,
+      "grad_norm": 1.5757924318313599,
+      "learning_rate": 0.00019216655282270557,
+      "loss": 0.6123,
+      "step": 2082
+    },
+    {
+      "epoch": 0.1531730274284874,
+      "grad_norm": 1.0624752044677734,
+      "learning_rate": 0.00019215730987825574,
+      "loss": 0.5766,
+      "step": 2083
+    },
+    {
+      "epoch": 0.15324656224722405,
+      "grad_norm": 1.1421533823013306,
+      "learning_rate": 0.00019214806170654605,
+      "loss": 0.6225,
+      "step": 2084
+    },
+    {
+      "epoch": 0.15332009706596073,
+      "grad_norm": 1.2640854120254517,
+      "learning_rate": 0.00019213880830810104,
+      "loss": 0.612,
+      "step": 2085
+    },
+    {
+      "epoch": 0.1533936318846974,
+      "grad_norm": 0.9683430194854736,
+      "learning_rate": 0.0001921295496834456,
+      "loss": 0.5555,
+      "step": 2086
+    },
+    {
+      "epoch": 0.15346716670343408,
+      "grad_norm": 1.6607552766799927,
+      "learning_rate": 0.00019212028583310488,
+      "loss": 0.61,
+      "step": 2087
+    },
+    {
+      "epoch": 0.15354070152217075,
+      "grad_norm": 1.4791414737701416,
+      "learning_rate": 0.0001921110167576043,
+      "loss": 0.6203,
+      "step": 2088
+    },
+    {
+      "epoch": 0.15361423634090743,
+      "grad_norm": 1.3145577907562256,
+      "learning_rate": 0.00019210174245746963,
+      "loss": 0.6001,
+      "step": 2089
+    },
+    {
+      "epoch": 0.1536877711596441,
+      "grad_norm": 1.0585238933563232,
+      "learning_rate": 0.00019209246293322698,
+      "loss": 0.5749,
+      "step": 2090
+    },
+    {
+      "epoch": 0.15376130597838075,
+      "grad_norm": 1.2738804817199707,
+      "learning_rate": 0.00019208317818540261,
+      "loss": 0.6262,
+      "step": 2091
+    },
+    {
+      "epoch": 0.15383484079711743,
+      "grad_norm": 1.1807081699371338,
+      "learning_rate": 0.00019207388821452321,
+      "loss": 0.5949,
+      "step": 2092
+    },
+    {
+      "epoch": 0.1539083756158541,
+      "grad_norm": 1.6652098894119263,
+      "learning_rate": 0.00019206459302111568,
+      "loss": 0.6199,
+      "step": 2093
+    },
+    {
+      "epoch": 0.15398191043459078,
+      "grad_norm": 1.3402431011199951,
+      "learning_rate": 0.0001920552926057073,
+      "loss": 0.6142,
+      "step": 2094
+    },
+    {
+      "epoch": 0.15405544525332746,
+      "grad_norm": 1.300307273864746,
+      "learning_rate": 0.00019204598696882555,
+      "loss": 0.6184,
+      "step": 2095
+    },
+    {
+      "epoch": 0.15412898007206413,
+      "grad_norm": 1.1726100444793701,
+      "learning_rate": 0.0001920366761109983,
+      "loss": 0.5973,
+      "step": 2096
+    },
+    {
+      "epoch": 0.1542025148908008,
+      "grad_norm": 1.359039545059204,
+      "learning_rate": 0.00019202736003275363,
+      "loss": 0.607,
+      "step": 2097
+    },
+    {
+      "epoch": 0.15427604970953746,
+      "grad_norm": 1.2337024211883545,
+      "learning_rate": 0.00019201803873462,
+      "loss": 0.5866,
+      "step": 2098
+    },
+    {
+      "epoch": 0.15434958452827413,
+      "grad_norm": 1.399442195892334,
+      "learning_rate": 0.0001920087122171261,
+      "loss": 0.6109,
+      "step": 2099
+    },
+    {
+      "epoch": 0.1544231193470108,
+      "grad_norm": 1.1953675746917725,
+      "learning_rate": 0.0001919993804808009,
+      "loss": 0.606,
+      "step": 2100
+    },
+    {
+      "epoch": 0.15449665416574748,
+      "grad_norm": 1.234302282333374,
+      "learning_rate": 0.0001919900435261738,
+      "loss": 0.5841,
+      "step": 2101
+    },
+    {
+      "epoch": 0.15457018898448416,
+      "grad_norm": 1.2199605703353882,
+      "learning_rate": 0.00019198070135377433,
+      "loss": 0.5986,
+      "step": 2102
+    },
+    {
+      "epoch": 0.15464372380322083,
+      "grad_norm": 1.2612601518630981,
+      "learning_rate": 0.0001919713539641324,
+      "loss": 0.6405,
+      "step": 2103
+    },
+    {
+      "epoch": 0.1547172586219575,
+      "grad_norm": 1.027937889099121,
+      "learning_rate": 0.00019196200135777823,
+      "loss": 0.6221,
+      "step": 2104
+    },
+    {
+      "epoch": 0.15479079344069416,
+      "grad_norm": 1.5683932304382324,
+      "learning_rate": 0.0001919526435352423,
+      "loss": 0.6722,
+      "step": 2105
+    },
+    {
+      "epoch": 0.15486432825943083,
+      "grad_norm": 1.3107693195343018,
+      "learning_rate": 0.00019194328049705535,
+      "loss": 0.6159,
+      "step": 2106
+    },
+    {
+      "epoch": 0.1549378630781675,
+      "grad_norm": 1.297667145729065,
+      "learning_rate": 0.00019193391224374854,
+      "loss": 0.6069,
+      "step": 2107
+    },
+    {
+      "epoch": 0.15501139789690419,
+      "grad_norm": 1.3455016613006592,
+      "learning_rate": 0.00019192453877585317,
+      "loss": 0.6181,
+      "step": 2108
+    },
+    {
+      "epoch": 0.15508493271564086,
+      "grad_norm": 1.0626004934310913,
+      "learning_rate": 0.00019191516009390097,
+      "loss": 0.6061,
+      "step": 2109
+    },
+    {
+      "epoch": 0.15515846753437754,
+      "grad_norm": 1.107873797416687,
+      "learning_rate": 0.00019190577619842387,
+      "loss": 0.6336,
+      "step": 2110
+    },
+    {
+      "epoch": 0.1552320023531142,
+      "grad_norm": 1.3925905227661133,
+      "learning_rate": 0.00019189638708995416,
+      "loss": 0.6202,
+      "step": 2111
+    },
+    {
+      "epoch": 0.15530553717185086,
+      "grad_norm": 1.273090124130249,
+      "learning_rate": 0.00019188699276902443,
+      "loss": 0.6139,
+      "step": 2112
+    },
+    {
+      "epoch": 0.15537907199058754,
+      "grad_norm": 1.5439510345458984,
+      "learning_rate": 0.00019187759323616743,
+      "loss": 0.6584,
+      "step": 2113
+    },
+    {
+      "epoch": 0.1554526068093242,
+      "grad_norm": 1.4256669282913208,
+      "learning_rate": 0.00019186818849191642,
+      "loss": 0.6271,
+      "step": 2114
+    },
+    {
+      "epoch": 0.1555261416280609,
+      "grad_norm": 0.9905490875244141,
+      "learning_rate": 0.0001918587785368048,
+      "loss": 0.6517,
+      "step": 2115
+    },
+    {
+      "epoch": 0.15559967644679756,
+      "grad_norm": 0.8915784358978271,
+      "learning_rate": 0.00019184936337136632,
+      "loss": 0.6466,
+      "step": 2116
+    },
+    {
+      "epoch": 0.15567321126553424,
+      "grad_norm": 1.0180715322494507,
+      "learning_rate": 0.00019183994299613503,
+      "loss": 0.6045,
+      "step": 2117
+    },
+    {
+      "epoch": 0.15574674608427092,
+      "grad_norm": 0.6573257446289062,
+      "learning_rate": 0.00019183051741164522,
+      "loss": 0.6417,
+      "step": 2118
+    },
+    {
+      "epoch": 0.15582028090300756,
+      "grad_norm": 1.032867193222046,
+      "learning_rate": 0.00019182108661843156,
+      "loss": 0.6398,
+      "step": 2119
+    },
+    {
+      "epoch": 0.15589381572174424,
+      "grad_norm": 0.629179835319519,
+      "learning_rate": 0.00019181165061702896,
+      "loss": 0.6651,
+      "step": 2120
+    },
+    {
+      "epoch": 0.15596735054048091,
+      "grad_norm": 1.0843836069107056,
+      "learning_rate": 0.00019180220940797268,
+      "loss": 0.6539,
+      "step": 2121
+    },
+    {
+      "epoch": 0.1560408853592176,
+      "grad_norm": 0.6069067120552063,
+      "learning_rate": 0.00019179276299179817,
+      "loss": 0.6342,
+      "step": 2122
+    },
+    {
+      "epoch": 0.15611442017795427,
+      "grad_norm": 0.7789467573165894,
+      "learning_rate": 0.00019178331136904126,
+      "loss": 0.6836,
+      "step": 2123
+    },
+    {
+      "epoch": 0.15618795499669094,
+      "grad_norm": 0.5800456404685974,
+      "learning_rate": 0.0001917738545402381,
+      "loss": 0.655,
+      "step": 2124
+    },
+    {
+      "epoch": 0.15626148981542762,
+      "grad_norm": 0.7173088192939758,
+      "learning_rate": 0.00019176439250592504,
+      "loss": 0.6766,
+      "step": 2125
+    },
+    {
+      "epoch": 0.15633502463416427,
+      "grad_norm": 0.6396276354789734,
+      "learning_rate": 0.00019175492526663881,
+      "loss": 0.6493,
+      "step": 2126
+    },
+    {
+      "epoch": 0.15640855945290094,
+      "grad_norm": 0.6834299564361572,
+      "learning_rate": 0.00019174545282291634,
+      "loss": 0.6448,
+      "step": 2127
+    },
+    {
+      "epoch": 0.15648209427163762,
+      "grad_norm": 0.8776350617408752,
+      "learning_rate": 0.00019173597517529501,
+      "loss": 0.6282,
+      "step": 2128
+    },
+    {
+      "epoch": 0.1565556290903743,
+      "grad_norm": 0.6228664517402649,
+      "learning_rate": 0.0001917264923243123,
+      "loss": 0.6624,
+      "step": 2129
+    },
+    {
+      "epoch": 0.15662916390911097,
+      "grad_norm": 0.5540098547935486,
+      "learning_rate": 0.0001917170042705062,
+      "loss": 0.6373,
+      "step": 2130
+    },
+    {
+      "epoch": 0.15670269872784764,
+      "grad_norm": 0.5611093044281006,
+      "learning_rate": 0.0001917075110144148,
+      "loss": 0.6503,
+      "step": 2131
+    },
+    {
+      "epoch": 0.15677623354658432,
+      "grad_norm": 0.5990686416625977,
+      "learning_rate": 0.0001916980125565766,
+      "loss": 0.6541,
+      "step": 2132
+    },
+    {
+      "epoch": 0.15684976836532097,
+      "grad_norm": 0.5697470903396606,
+      "learning_rate": 0.00019168850889753037,
+      "loss": 0.6219,
+      "step": 2133
+    },
+    {
+      "epoch": 0.15692330318405764,
+      "grad_norm": 0.6302569508552551,
+      "learning_rate": 0.00019167900003781515,
+      "loss": 0.6376,
+      "step": 2134
+    },
+    {
+      "epoch": 0.15699683800279432,
+      "grad_norm": 0.5192708373069763,
+      "learning_rate": 0.00019166948597797027,
+      "loss": 0.6114,
+      "step": 2135
+    },
+    {
+      "epoch": 0.157070372821531,
+      "grad_norm": 0.6720625758171082,
+      "learning_rate": 0.0001916599667185354,
+      "loss": 0.6309,
+      "step": 2136
+    },
+    {
+      "epoch": 0.15714390764026767,
+      "grad_norm": 0.633891224861145,
+      "learning_rate": 0.00019165044226005052,
+      "loss": 0.6273,
+      "step": 2137
+    },
+    {
+      "epoch": 0.15721744245900435,
+      "grad_norm": 0.5748793482780457,
+      "learning_rate": 0.0001916409126030558,
+      "loss": 0.6444,
+      "step": 2138
+    },
+    {
+      "epoch": 0.15729097727774102,
+      "grad_norm": 0.9179320335388184,
+      "learning_rate": 0.0001916313777480918,
+      "loss": 0.6283,
+      "step": 2139
+    },
+    {
+      "epoch": 0.15736451209647767,
+      "grad_norm": 1.140636920928955,
+      "learning_rate": 0.0001916218376956994,
+      "loss": 0.6373,
+      "step": 2140
+    },
+    {
+      "epoch": 0.15743804691521435,
+      "grad_norm": 0.9402908086776733,
+      "learning_rate": 0.00019161229244641961,
+      "loss": 0.6672,
+      "step": 2141
+    },
+    {
+      "epoch": 0.15751158173395102,
+      "grad_norm": 0.8982166647911072,
+      "learning_rate": 0.00019160274200079392,
+      "loss": 0.6358,
+      "step": 2142
+    },
+    {
+      "epoch": 0.1575851165526877,
+      "grad_norm": 0.6174270510673523,
+      "learning_rate": 0.00019159318635936408,
+      "loss": 0.6333,
+      "step": 2143
+    },
+    {
+      "epoch": 0.15765865137142437,
+      "grad_norm": 1.4778465032577515,
+      "learning_rate": 0.00019158362552267202,
+      "loss": 0.6609,
+      "step": 2144
+    },
+    {
+      "epoch": 0.15773218619016105,
+      "grad_norm": 1.3314608335494995,
+      "learning_rate": 0.00019157405949126006,
+      "loss": 0.6535,
+      "step": 2145
+    },
+    {
+      "epoch": 0.15780572100889773,
+      "grad_norm": 0.9024116396903992,
+      "learning_rate": 0.0001915644882656708,
+      "loss": 0.6585,
+      "step": 2146
+    },
+    {
+      "epoch": 0.15787925582763437,
+      "grad_norm": 2.2677083015441895,
+      "learning_rate": 0.00019155491184644715,
+      "loss": 0.6635,
+      "step": 2147
+    },
+    {
+      "epoch": 0.15795279064637105,
+      "grad_norm": 1.9821138381958008,
+      "learning_rate": 0.0001915453302341323,
+      "loss": 0.6757,
+      "step": 2148
+    },
+    {
+      "epoch": 0.15802632546510773,
+      "grad_norm": 1.5982760190963745,
+      "learning_rate": 0.00019153574342926967,
+      "loss": 0.6669,
+      "step": 2149
+    },
+    {
+      "epoch": 0.1580998602838444,
+      "grad_norm": 1.868851661682129,
+      "learning_rate": 0.0001915261514324031,
+      "loss": 0.6889,
+      "step": 2150
+    },
+    {
+      "epoch": 0.15817339510258108,
+      "grad_norm": 1.1529220342636108,
+      "learning_rate": 0.0001915165542440766,
+      "loss": 0.6345,
+      "step": 2151
+    },
+    {
+      "epoch": 0.15824692992131775,
+      "grad_norm": 1.3319082260131836,
+      "learning_rate": 0.00019150695186483457,
+      "loss": 0.6498,
+      "step": 2152
+    },
+    {
+      "epoch": 0.15832046474005443,
+      "grad_norm": 1.1484408378601074,
+      "learning_rate": 0.00019149734429522167,
+      "loss": 0.6185,
+      "step": 2153
+    },
+    {
+      "epoch": 0.15839399955879108,
+      "grad_norm": 1.0978519916534424,
+      "learning_rate": 0.0001914877315357828,
+      "loss": 0.691,
+      "step": 2154
+    },
+    {
+      "epoch": 0.15846753437752775,
+      "grad_norm": 1.1299211978912354,
+      "learning_rate": 0.0001914781135870633,
+      "loss": 0.6098,
+      "step": 2155
+    },
+    {
+      "epoch": 0.15854106919626443,
+      "grad_norm": 0.8983652591705322,
+      "learning_rate": 0.00019146849044960864,
+      "loss": 0.6396,
+      "step": 2156
+    },
+    {
+      "epoch": 0.1586146040150011,
+      "grad_norm": 1.2938261032104492,
+      "learning_rate": 0.00019145886212396467,
+      "loss": 0.636,
+      "step": 2157
+    },
+    {
+      "epoch": 0.15868813883373778,
+      "grad_norm": 0.8432628512382507,
+      "learning_rate": 0.00019144922861067754,
+      "loss": 0.6483,
+      "step": 2158
+    },
+    {
+      "epoch": 0.15876167365247446,
+      "grad_norm": 0.9158918857574463,
+      "learning_rate": 0.0001914395899102936,
+      "loss": 0.61,
+      "step": 2159
+    },
+    {
+      "epoch": 0.15883520847121113,
+      "grad_norm": 0.7635400295257568,
+      "learning_rate": 0.00019142994602335965,
+      "loss": 0.6366,
+      "step": 2160
+    },
+    {
+      "epoch": 0.15890874328994778,
+      "grad_norm": 0.8548750877380371,
+      "learning_rate": 0.00019142029695042267,
+      "loss": 0.6247,
+      "step": 2161
+    },
+    {
+      "epoch": 0.15898227810868445,
+      "grad_norm": 0.614342451095581,
+      "learning_rate": 0.00019141064269203,
+      "loss": 0.64,
+      "step": 2162
+    },
+    {
+      "epoch": 0.15905581292742113,
+      "grad_norm": 0.9320647716522217,
+      "learning_rate": 0.00019140098324872917,
+      "loss": 0.6378,
+      "step": 2163
+    },
+    {
+      "epoch": 0.1591293477461578,
+      "grad_norm": 0.6568562388420105,
+      "learning_rate": 0.00019139131862106813,
+      "loss": 0.6279,
+      "step": 2164
+    },
+    {
+      "epoch": 0.15920288256489448,
+      "grad_norm": 0.6443667411804199,
+      "learning_rate": 0.00019138164880959501,
+      "loss": 0.6208,
+      "step": 2165
+    },
+    {
+      "epoch": 0.15927641738363116,
+      "grad_norm": 0.6028559803962708,
+      "learning_rate": 0.00019137197381485836,
+      "loss": 0.6418,
+      "step": 2166
+    },
+    {
+      "epoch": 0.15934995220236783,
+      "grad_norm": 0.6664873361587524,
+      "learning_rate": 0.0001913622936374069,
+      "loss": 0.6507,
+      "step": 2167
+    },
+    {
+      "epoch": 0.15942348702110448,
+      "grad_norm": 0.7387083172798157,
+      "learning_rate": 0.00019135260827778978,
+      "loss": 0.638,
+      "step": 2168
+    },
+    {
+      "epoch": 0.15949702183984116,
+      "grad_norm": 0.8267252445220947,
+      "learning_rate": 0.0001913429177365563,
+      "loss": 0.6353,
+      "step": 2169
+    },
+    {
+      "epoch": 0.15957055665857783,
+      "grad_norm": 0.6043718457221985,
+      "learning_rate": 0.00019133322201425607,
+      "loss": 0.6255,
+      "step": 2170
+    },
+    {
+      "epoch": 0.1596440914773145,
+      "grad_norm": 0.7126490473747253,
+      "learning_rate": 0.00019132352111143915,
+      "loss": 0.6613,
+      "step": 2171
+    },
+    {
+      "epoch": 0.15971762629605118,
+      "grad_norm": 1.044957160949707,
+      "learning_rate": 0.00019131381502865573,
+      "loss": 0.6074,
+      "step": 2172
+    },
+    {
+      "epoch": 0.15979116111478786,
+      "grad_norm": 1.1037592887878418,
+      "learning_rate": 0.00019130410376645633,
+      "loss": 0.6284,
+      "step": 2173
+    },
+    {
+      "epoch": 0.15986469593352454,
+      "grad_norm": 0.5753788948059082,
+      "learning_rate": 0.00019129438732539184,
+      "loss": 0.575,
+      "step": 2174
+    },
+    {
+      "epoch": 0.15993823075226118,
+      "grad_norm": 1.0623217821121216,
+      "learning_rate": 0.00019128466570601335,
+      "loss": 0.6302,
+      "step": 2175
+    },
+    {
+      "epoch": 0.16001176557099786,
+      "grad_norm": 1.2378612756729126,
+      "learning_rate": 0.00019127493890887224,
+      "loss": 0.6216,
+      "step": 2176
+    },
+    {
+      "epoch": 0.16008530038973454,
+      "grad_norm": 0.5903173089027405,
+      "learning_rate": 0.00019126520693452033,
+      "loss": 0.681,
+      "step": 2177
+    },
+    {
+      "epoch": 0.1601588352084712,
+      "grad_norm": 1.1613059043884277,
+      "learning_rate": 0.00019125546978350952,
+      "loss": 0.6228,
+      "step": 2178
+    },
+    {
+      "epoch": 0.1602323700272079,
+      "grad_norm": 0.7827247381210327,
+      "learning_rate": 0.00019124572745639218,
+      "loss": 0.6243,
+      "step": 2179
+    },
+    {
+      "epoch": 0.16030590484594456,
+      "grad_norm": 0.6669643521308899,
+      "learning_rate": 0.0001912359799537209,
+      "loss": 0.6158,
+      "step": 2180
+    },
+    {
+      "epoch": 0.16037943966468124,
+      "grad_norm": 0.8932924270629883,
+      "learning_rate": 0.0001912262272760485,
+      "loss": 0.6192,
+      "step": 2181
+    },
+    {
+      "epoch": 0.1604529744834179,
+      "grad_norm": 0.7368266582489014,
+      "learning_rate": 0.00019121646942392827,
+      "loss": 0.6296,
+      "step": 2182
+    },
+    {
+      "epoch": 0.16052650930215456,
+      "grad_norm": 0.634384274482727,
+      "learning_rate": 0.0001912067063979136,
+      "loss": 0.6297,
+      "step": 2183
+    },
+    {
+      "epoch": 0.16060004412089124,
+      "grad_norm": 0.6204035878181458,
+      "learning_rate": 0.0001911969381985583,
+      "loss": 0.6177,
+      "step": 2184
+    },
+    {
+      "epoch": 0.16067357893962791,
+      "grad_norm": 0.6050305366516113,
+      "learning_rate": 0.00019118716482641643,
+      "loss": 0.6451,
+      "step": 2185
+    },
+    {
+      "epoch": 0.1607471137583646,
+      "grad_norm": 0.7035597562789917,
+      "learning_rate": 0.0001911773862820423,
+      "loss": 0.6252,
+      "step": 2186
+    },
+    {
+      "epoch": 0.16082064857710127,
+      "grad_norm": 0.8586496114730835,
+      "learning_rate": 0.00019116760256599063,
+      "loss": 0.6039,
+      "step": 2187
+    },
+    {
+      "epoch": 0.16089418339583794,
+      "grad_norm": 0.9056009650230408,
+      "learning_rate": 0.00019115781367881631,
+      "loss": 0.6322,
+      "step": 2188
+    },
+    {
+      "epoch": 0.1609677182145746,
+      "grad_norm": 0.7366215586662292,
+      "learning_rate": 0.0001911480196210746,
+      "loss": 0.6264,
+      "step": 2189
+    },
+    {
+      "epoch": 0.16104125303331127,
+      "grad_norm": 0.6286762952804565,
+      "learning_rate": 0.00019113822039332106,
+      "loss": 0.6131,
+      "step": 2190
+    },
+    {
+      "epoch": 0.16111478785204794,
+      "grad_norm": 1.1564215421676636,
+      "learning_rate": 0.00019112841599611144,
+      "loss": 0.626,
+      "step": 2191
+    },
+    {
+      "epoch": 0.16118832267078462,
+      "grad_norm": 0.6435374021530151,
+      "learning_rate": 0.0001911186064300019,
+      "loss": 0.6667,
+      "step": 2192
+    },
+    {
+      "epoch": 0.1612618574895213,
+      "grad_norm": 1.2930060625076294,
+      "learning_rate": 0.00019110879169554883,
+      "loss": 0.6035,
+      "step": 2193
+    },
+    {
+      "epoch": 0.16133539230825797,
+      "grad_norm": 1.1795870065689087,
+      "learning_rate": 0.00019109897179330897,
+      "loss": 0.6473,
+      "step": 2194
+    },
+    {
+      "epoch": 0.16140892712699464,
+      "grad_norm": 0.6027973890304565,
+      "learning_rate": 0.00019108914672383927,
+      "loss": 0.6149,
+      "step": 2195
+    },
+    {
+      "epoch": 0.1614824619457313,
+      "grad_norm": 1.0513123273849487,
+      "learning_rate": 0.00019107931648769703,
+      "loss": 0.6326,
+      "step": 2196
+    },
+    {
+      "epoch": 0.16155599676446797,
+      "grad_norm": 1.4592509269714355,
+      "learning_rate": 0.00019106948108543986,
+      "loss": 0.6271,
+      "step": 2197
+    },
+    {
+      "epoch": 0.16162953158320464,
+      "grad_norm": 0.8664681911468506,
+      "learning_rate": 0.0001910596405176256,
+      "loss": 0.6391,
+      "step": 2198
+    },
+    {
+      "epoch": 0.16170306640194132,
+      "grad_norm": 2.528028726577759,
+      "learning_rate": 0.00019104979478481245,
+      "loss": 0.6859,
+      "step": 2199
+    },
+    {
+      "epoch": 0.161776601220678,
+      "grad_norm": 1.758297085762024,
+      "learning_rate": 0.00019103994388755882,
+      "loss": 0.6762,
+      "step": 2200
+    },
+    {
+      "epoch": 0.16185013603941467,
+      "grad_norm": 1.5180413722991943,
+      "learning_rate": 0.0001910300878264235,
+      "loss": 0.6524,
+      "step": 2201
+    },
+    {
+      "epoch": 0.16192367085815135,
+      "grad_norm": 1.0148382186889648,
+      "learning_rate": 0.00019102022660196557,
+      "loss": 0.6617,
+      "step": 2202
+    },
+    {
+      "epoch": 0.161997205676888,
+      "grad_norm": 1.736011028289795,
+      "learning_rate": 0.0001910103602147443,
+      "loss": 0.6671,
+      "step": 2203
+    },
+    {
+      "epoch": 0.16207074049562467,
+      "grad_norm": 0.7359936833381653,
+      "learning_rate": 0.00019100048866531934,
+      "loss": 0.661,
+      "step": 2204
+    },
+    {
+      "epoch": 0.16214427531436135,
+      "grad_norm": 0.9722472429275513,
+      "learning_rate": 0.00019099061195425066,
+      "loss": 0.664,
+      "step": 2205
+    },
+    {
+      "epoch": 0.16221781013309802,
+      "grad_norm": 1.1769371032714844,
+      "learning_rate": 0.0001909807300820984,
+      "loss": 0.6607,
+      "step": 2206
+    },
+    {
+      "epoch": 0.1622913449518347,
+      "grad_norm": 0.9886273145675659,
+      "learning_rate": 0.00019097084304942315,
+      "loss": 0.6509,
+      "step": 2207
+    },
+    {
+      "epoch": 0.16236487977057137,
+      "grad_norm": 0.7259057760238647,
+      "learning_rate": 0.00019096095085678567,
+      "loss": 0.6755,
+      "step": 2208
+    },
+    {
+      "epoch": 0.16243841458930805,
+      "grad_norm": 0.6283203959465027,
+      "learning_rate": 0.00019095105350474704,
+      "loss": 0.6533,
+      "step": 2209
+    },
+    {
+      "epoch": 0.1625119494080447,
+      "grad_norm": 0.5870376229286194,
+      "learning_rate": 0.00019094115099386872,
+      "loss": 0.6339,
+      "step": 2210
+    },
+    {
+      "epoch": 0.16258548422678137,
+      "grad_norm": 0.6920521855354309,
+      "learning_rate": 0.0001909312433247123,
+      "loss": 0.6357,
+      "step": 2211
+    },
+    {
+      "epoch": 0.16265901904551805,
+      "grad_norm": 0.7739456295967102,
+      "learning_rate": 0.0001909213304978398,
+      "loss": 0.7012,
+      "step": 2212
+    },
+    {
+      "epoch": 0.16273255386425473,
+      "grad_norm": 1.2596499919891357,
+      "learning_rate": 0.0001909114125138135,
+      "loss": 0.6715,
+      "step": 2213
+    },
+    {
+      "epoch": 0.1628060886829914,
+      "grad_norm": 1.0600128173828125,
+      "learning_rate": 0.00019090148937319593,
+      "loss": 0.6426,
+      "step": 2214
+    },
+    {
+      "epoch": 0.16287962350172808,
+      "grad_norm": 0.7162318825721741,
+      "learning_rate": 0.00019089156107654994,
+      "loss": 0.6334,
+      "step": 2215
+    },
+    {
+      "epoch": 0.16295315832046475,
+      "grad_norm": 0.9141923785209656,
+      "learning_rate": 0.0001908816276244387,
+      "loss": 0.652,
+      "step": 2216
+    },
+    {
+      "epoch": 0.1630266931392014,
+      "grad_norm": 0.9664901494979858,
+      "learning_rate": 0.00019087168901742562,
+      "loss": 0.6557,
+      "step": 2217
+    },
+    {
+      "epoch": 0.16310022795793808,
+      "grad_norm": 0.617310643196106,
+      "learning_rate": 0.0001908617452560744,
+      "loss": 0.6406,
+      "step": 2218
+    },
+    {
+      "epoch": 0.16317376277667475,
+      "grad_norm": 0.6342182755470276,
+      "learning_rate": 0.00019085179634094918,
+      "loss": 0.633,
+      "step": 2219
+    },
+    {
+      "epoch": 0.16324729759541143,
+      "grad_norm": 0.8543694019317627,
+      "learning_rate": 0.00019084184227261412,
+      "loss": 0.6304,
+      "step": 2220
+    },
+    {
+      "epoch": 0.1633208324141481,
+      "grad_norm": 1.2320103645324707,
+      "learning_rate": 0.00019083188305163393,
+      "loss": 0.665,
+      "step": 2221
+    },
+    {
+      "epoch": 0.16339436723288478,
+      "grad_norm": 1.4839578866958618,
+      "learning_rate": 0.0001908219186785735,
+      "loss": 0.6078,
+      "step": 2222
+    },
+    {
+      "epoch": 0.16346790205162146,
+      "grad_norm": 0.5664721727371216,
+      "learning_rate": 0.00019081194915399792,
+      "loss": 0.6442,
+      "step": 2223
+    },
+    {
+      "epoch": 0.1635414368703581,
+      "grad_norm": 1.3387855291366577,
+      "learning_rate": 0.00019080197447847281,
+      "loss": 0.6242,
+      "step": 2224
+    },
+    {
+      "epoch": 0.16361497168909478,
+      "grad_norm": 1.0854068994522095,
+      "learning_rate": 0.00019079199465256384,
+      "loss": 0.6316,
+      "step": 2225
+    },
+    {
+      "epoch": 0.16368850650783145,
+      "grad_norm": 0.6048920750617981,
+      "learning_rate": 0.00019078200967683718,
+      "loss": 0.6381,
+      "step": 2226
+    },
+    {
+      "epoch": 0.16376204132656813,
+      "grad_norm": 0.6663655042648315,
+      "learning_rate": 0.00019077201955185907,
+      "loss": 0.6286,
+      "step": 2227
+    },
+    {
+      "epoch": 0.1638355761453048,
+      "grad_norm": 0.714576244354248,
+      "learning_rate": 0.00019076202427819623,
+      "loss": 0.6335,
+      "step": 2228
+    },
+    {
+      "epoch": 0.16390911096404148,
+      "grad_norm": 0.6018645763397217,
+      "learning_rate": 0.00019075202385641557,
+      "loss": 0.6501,
+      "step": 2229
+    },
+    {
+      "epoch": 0.16398264578277816,
+      "grad_norm": 1.2483736276626587,
+      "learning_rate": 0.00019074201828708437,
+      "loss": 0.6415,
+      "step": 2230
+    },
+    {
+      "epoch": 0.1640561806015148,
+      "grad_norm": 1.1382880210876465,
+      "learning_rate": 0.0001907320075707701,
+      "loss": 0.6472,
+      "step": 2231
+    },
+    {
+      "epoch": 0.16412971542025148,
+      "grad_norm": 0.5671947002410889,
+      "learning_rate": 0.00019072199170804064,
+      "loss": 0.6273,
+      "step": 2232
+    },
+    {
+      "epoch": 0.16420325023898816,
+      "grad_norm": 0.865458071231842,
+      "learning_rate": 0.00019071197069946404,
+      "loss": 0.6482,
+      "step": 2233
+    },
+    {
+      "epoch": 0.16427678505772483,
+      "grad_norm": 0.9757769107818604,
+      "learning_rate": 0.00019070194454560876,
+      "loss": 0.6592,
+      "step": 2234
+    },
+    {
+      "epoch": 0.1643503198764615,
+      "grad_norm": 1.308665156364441,
+      "learning_rate": 0.00019069191324704344,
+      "loss": 0.6404,
+      "step": 2235
+    },
+    {
+      "epoch": 0.16442385469519818,
+      "grad_norm": 1.0625808238983154,
+      "learning_rate": 0.00019068187680433712,
+      "loss": 0.6523,
+      "step": 2236
+    },
+    {
+      "epoch": 0.16449738951393486,
+      "grad_norm": 1.2617169618606567,
+      "learning_rate": 0.000190671835218059,
+      "loss": 0.6465,
+      "step": 2237
+    },
+    {
+      "epoch": 0.1645709243326715,
+      "grad_norm": 0.8140657544136047,
+      "learning_rate": 0.00019066178848877873,
+      "loss": 0.6714,
+      "step": 2238
+    },
+    {
+      "epoch": 0.16464445915140818,
+      "grad_norm": 0.9638645052909851,
+      "learning_rate": 0.00019065173661706613,
+      "loss": 0.6583,
+      "step": 2239
+    },
+    {
+      "epoch": 0.16471799397014486,
+      "grad_norm": 2.212385892868042,
+      "learning_rate": 0.00019064167960349134,
+      "loss": 0.6586,
+      "step": 2240
+    },
+    {
+      "epoch": 0.16479152878888154,
+      "grad_norm": 1.3571721315383911,
+      "learning_rate": 0.00019063161744862486,
+      "loss": 0.6431,
+      "step": 2241
+    },
+    {
+      "epoch": 0.1648650636076182,
+      "grad_norm": 3.0443625450134277,
+      "learning_rate": 0.0001906215501530374,
+      "loss": 0.6738,
+      "step": 2242
+    },
+    {
+      "epoch": 0.1649385984263549,
+      "grad_norm": 3.2277917861938477,
+      "learning_rate": 0.00019061147771729992,
+      "loss": 0.6541,
+      "step": 2243
+    },
+    {
+      "epoch": 0.16501213324509156,
+      "grad_norm": 0.8904592990875244,
+      "learning_rate": 0.00019060140014198385,
+      "loss": 0.6487,
+      "step": 2244
+    },
+    {
+      "epoch": 0.1650856680638282,
+      "grad_norm": 2.289583206176758,
+      "learning_rate": 0.0001905913174276607,
+      "loss": 0.6611,
+      "step": 2245
+    },
+    {
+      "epoch": 0.1651592028825649,
+      "grad_norm": 1.7087868452072144,
+      "learning_rate": 0.00019058122957490246,
+      "loss": 0.663,
+      "step": 2246
+    },
+    {
+      "epoch": 0.16523273770130156,
+      "grad_norm": 1.5161769390106201,
+      "learning_rate": 0.00019057113658428127,
+      "loss": 0.6191,
+      "step": 2247
+    },
+    {
+      "epoch": 0.16530627252003824,
+      "grad_norm": 1.6709160804748535,
+      "learning_rate": 0.00019056103845636963,
+      "loss": 0.6405,
+      "step": 2248
+    },
+    {
+      "epoch": 0.16537980733877491,
+      "grad_norm": 1.4936836957931519,
+      "learning_rate": 0.00019055093519174032,
+      "loss": 0.629,
+      "step": 2249
+    },
+    {
+      "epoch": 0.1654533421575116,
+      "grad_norm": 1.4548832178115845,
+      "learning_rate": 0.00019054082679096642,
+      "loss": 0.6544,
+      "step": 2250
+    },
+    {
+      "epoch": 0.16552687697624827,
+      "grad_norm": 1.6587779521942139,
+      "learning_rate": 0.00019053071325462125,
+      "loss": 0.6411,
+      "step": 2251
+    },
+    {
+      "epoch": 0.16560041179498491,
+      "grad_norm": 1.0973225831985474,
+      "learning_rate": 0.00019052059458327849,
+      "loss": 0.6574,
+      "step": 2252
+    },
+    {
+      "epoch": 0.1656739466137216,
+      "grad_norm": 2.0125889778137207,
+      "learning_rate": 0.00019051047077751207,
+      "loss": 0.629,
+      "step": 2253
+    },
+    {
+      "epoch": 0.16574748143245827,
+      "grad_norm": 1.6537948846817017,
+      "learning_rate": 0.00019050034183789626,
+      "loss": 0.6363,
+      "step": 2254
+    },
+    {
+      "epoch": 0.16582101625119494,
+      "grad_norm": 1.7242481708526611,
+      "learning_rate": 0.0001904902077650055,
+      "loss": 0.6389,
+      "step": 2255
+    },
+    {
+      "epoch": 0.16589455106993162,
+      "grad_norm": 2.086482286453247,
+      "learning_rate": 0.00019048006855941466,
+      "loss": 0.6258,
+      "step": 2256
+    },
+    {
+      "epoch": 0.1659680858886683,
+      "grad_norm": 0.982549786567688,
+      "learning_rate": 0.0001904699242216989,
+      "loss": 0.6394,
+      "step": 2257
+    },
+    {
+      "epoch": 0.16604162070740497,
+      "grad_norm": 1.5269120931625366,
+      "learning_rate": 0.00019045977475243354,
+      "loss": 0.6519,
+      "step": 2258
+    },
+    {
+      "epoch": 0.16611515552614162,
+      "grad_norm": 1.0441293716430664,
+      "learning_rate": 0.00019044962015219425,
+      "loss": 0.603,
+      "step": 2259
+    },
+    {
+      "epoch": 0.1661886903448783,
+      "grad_norm": 1.033138394355774,
+      "learning_rate": 0.0001904394604215571,
+      "loss": 0.6232,
+      "step": 2260
+    },
+    {
+      "epoch": 0.16626222516361497,
+      "grad_norm": 1.031362771987915,
+      "learning_rate": 0.00019042929556109826,
+      "loss": 0.6144,
+      "step": 2261
+    },
+    {
+      "epoch": 0.16633575998235164,
+      "grad_norm": 0.7078335881233215,
+      "learning_rate": 0.0001904191255713944,
+      "loss": 0.6363,
+      "step": 2262
+    },
+    {
+      "epoch": 0.16640929480108832,
+      "grad_norm": 1.053062915802002,
+      "learning_rate": 0.00019040895045302226,
+      "loss": 0.6272,
+      "step": 2263
+    },
+    {
+      "epoch": 0.166482829619825,
+      "grad_norm": 0.8137714862823486,
+      "learning_rate": 0.00019039877020655908,
+      "loss": 0.6195,
+      "step": 2264
+    },
+    {
+      "epoch": 0.16655636443856167,
+      "grad_norm": 0.9323541522026062,
+      "learning_rate": 0.0001903885848325822,
+      "loss": 0.6044,
+      "step": 2265
+    },
+    {
+      "epoch": 0.16662989925729832,
+      "grad_norm": 1.2368093729019165,
+      "learning_rate": 0.00019037839433166945,
+      "loss": 0.6021,
+      "step": 2266
+    },
+    {
+      "epoch": 0.166703434076035,
+      "grad_norm": 0.9028075933456421,
+      "learning_rate": 0.00019036819870439878,
+      "loss": 0.6712,
+      "step": 2267
+    },
+    {
+      "epoch": 0.16677696889477167,
+      "grad_norm": 1.6174015998840332,
+      "learning_rate": 0.0001903579979513485,
+      "loss": 0.6435,
+      "step": 2268
+    },
+    {
+      "epoch": 0.16685050371350835,
+      "grad_norm": 1.1119295358657837,
+      "learning_rate": 0.00019034779207309724,
+      "loss": 0.609,
+      "step": 2269
+    },
+    {
+      "epoch": 0.16692403853224502,
+      "grad_norm": 1.7432574033737183,
+      "learning_rate": 0.00019033758107022384,
+      "loss": 0.6594,
+      "step": 2270
+    },
+    {
+      "epoch": 0.1669975733509817,
+      "grad_norm": 1.2717353105545044,
+      "learning_rate": 0.0001903273649433075,
+      "loss": 0.5976,
+      "step": 2271
+    },
+    {
+      "epoch": 0.16707110816971837,
+      "grad_norm": 1.1819802522659302,
+      "learning_rate": 0.00019031714369292772,
+      "loss": 0.6531,
+      "step": 2272
+    },
+    {
+      "epoch": 0.16714464298845502,
+      "grad_norm": 1.1084383726119995,
+      "learning_rate": 0.00019030691731966424,
+      "loss": 0.6295,
+      "step": 2273
+    },
+    {
+      "epoch": 0.1672181778071917,
+      "grad_norm": 1.1625977754592896,
+      "learning_rate": 0.00019029668582409707,
+      "loss": 0.6317,
+      "step": 2274
+    },
+    {
+      "epoch": 0.16729171262592837,
+      "grad_norm": 0.9348340630531311,
+      "learning_rate": 0.0001902864492068066,
+      "loss": 0.6196,
+      "step": 2275
+    },
+    {
+      "epoch": 0.16736524744466505,
+      "grad_norm": 1.013386607170105,
+      "learning_rate": 0.00019027620746837345,
+      "loss": 0.6316,
+      "step": 2276
+    },
+    {
+      "epoch": 0.16743878226340173,
+      "grad_norm": 0.6879242658615112,
+      "learning_rate": 0.00019026596060937854,
+      "loss": 0.6236,
+      "step": 2277
+    },
+    {
+      "epoch": 0.1675123170821384,
+      "grad_norm": 1.0798563957214355,
+      "learning_rate": 0.0001902557086304031,
+      "loss": 0.6085,
+      "step": 2278
+    },
+    {
+      "epoch": 0.16758585190087508,
+      "grad_norm": 1.1731269359588623,
+      "learning_rate": 0.00019024545153202858,
+      "loss": 0.6396,
+      "step": 2279
+    },
+    {
+      "epoch": 0.16765938671961172,
+      "grad_norm": 0.8066635727882385,
+      "learning_rate": 0.00019023518931483683,
+      "loss": 0.643,
+      "step": 2280
+    },
+    {
+      "epoch": 0.1677329215383484,
+      "grad_norm": 0.8768815398216248,
+      "learning_rate": 0.00019022492197940993,
+      "loss": 0.6406,
+      "step": 2281
+    },
+    {
+      "epoch": 0.16780645635708508,
+      "grad_norm": 1.6082412004470825,
+      "learning_rate": 0.0001902146495263302,
+      "loss": 0.6198,
+      "step": 2282
+    },
+    {
+      "epoch": 0.16787999117582175,
+      "grad_norm": 0.9743713140487671,
+      "learning_rate": 0.00019020437195618037,
+      "loss": 0.6244,
+      "step": 2283
+    },
+    {
+      "epoch": 0.16795352599455843,
+      "grad_norm": 2.371201992034912,
+      "learning_rate": 0.00019019408926954334,
+      "loss": 0.6294,
+      "step": 2284
+    },
+    {
+      "epoch": 0.1680270608132951,
+      "grad_norm": 2.4810516834259033,
+      "learning_rate": 0.0001901838014670024,
+      "loss": 0.6356,
+      "step": 2285
+    },
+    {
+      "epoch": 0.16810059563203178,
+      "grad_norm": 0.9760524034500122,
+      "learning_rate": 0.00019017350854914108,
+      "loss": 0.6332,
+      "step": 2286
+    },
+    {
+      "epoch": 0.16817413045076843,
+      "grad_norm": 1.9391226768493652,
+      "learning_rate": 0.00019016321051654316,
+      "loss": 0.6489,
+      "step": 2287
+    },
+    {
+      "epoch": 0.1682476652695051,
+      "grad_norm": 1.273955225944519,
+      "learning_rate": 0.00019015290736979282,
+      "loss": 0.6274,
+      "step": 2288
+    },
+    {
+      "epoch": 0.16832120008824178,
+      "grad_norm": 2.0481326580047607,
+      "learning_rate": 0.0001901425991094744,
+      "loss": 0.6227,
+      "step": 2289
+    },
+    {
+      "epoch": 0.16839473490697845,
+      "grad_norm": 2.010760545730591,
+      "learning_rate": 0.00019013228573617265,
+      "loss": 0.6238,
+      "step": 2290
+    },
+    {
+      "epoch": 0.16846826972571513,
+      "grad_norm": 0.9975161552429199,
+      "learning_rate": 0.0001901219672504725,
+      "loss": 0.6308,
+      "step": 2291
+    },
+    {
+      "epoch": 0.1685418045444518,
+      "grad_norm": 1.7535901069641113,
+      "learning_rate": 0.0001901116436529593,
+      "loss": 0.6131,
+      "step": 2292
+    },
+    {
+      "epoch": 0.16861533936318848,
+      "grad_norm": 1.576683759689331,
+      "learning_rate": 0.00019010131494421857,
+      "loss": 0.615,
+      "step": 2293
+    },
+    {
+      "epoch": 0.16868887418192513,
+      "grad_norm": 0.9800629615783691,
+      "learning_rate": 0.00019009098112483615,
+      "loss": 0.6187,
+      "step": 2294
+    },
+    {
+      "epoch": 0.1687624090006618,
+      "grad_norm": 1.2104239463806152,
+      "learning_rate": 0.00019008064219539817,
+      "loss": 0.5972,
+      "step": 2295
+    },
+    {
+      "epoch": 0.16883594381939848,
+      "grad_norm": 0.7332221269607544,
+      "learning_rate": 0.00019007029815649114,
+      "loss": 0.6405,
+      "step": 2296
+    },
+    {
+      "epoch": 0.16890947863813516,
+      "grad_norm": 1.301347255706787,
+      "learning_rate": 0.00019005994900870177,
+      "loss": 0.5994,
+      "step": 2297
+    },
+    {
+      "epoch": 0.16898301345687183,
+      "grad_norm": 0.7993969321250916,
+      "learning_rate": 0.000190049594752617,
+      "loss": 0.619,
+      "step": 2298
+    },
+    {
+      "epoch": 0.1690565482756085,
+      "grad_norm": 1.0129244327545166,
+      "learning_rate": 0.0001900392353888242,
+      "loss": 0.6326,
+      "step": 2299
+    },
+    {
+      "epoch": 0.16913008309434518,
+      "grad_norm": 0.8818065524101257,
+      "learning_rate": 0.00019002887091791094,
+      "loss": 0.6311,
+      "step": 2300
+    },
+    {
+      "epoch": 0.16920361791308183,
+      "grad_norm": 0.769138514995575,
+      "learning_rate": 0.00019001850134046512,
+      "loss": 0.5838,
+      "step": 2301
+    },
+    {
+      "epoch": 0.1692771527318185,
+      "grad_norm": 0.7338100075721741,
+      "learning_rate": 0.00019000812665707494,
+      "loss": 0.6143,
+      "step": 2302
+    },
+    {
+      "epoch": 0.16935068755055518,
+      "grad_norm": 0.8327420949935913,
+      "learning_rate": 0.00018999774686832876,
+      "loss": 0.6341,
+      "step": 2303
+    },
+    {
+      "epoch": 0.16942422236929186,
+      "grad_norm": 1.4155508279800415,
+      "learning_rate": 0.00018998736197481543,
+      "loss": 0.6129,
+      "step": 2304
+    },
+    {
+      "epoch": 0.16949775718802854,
+      "grad_norm": 0.869440495967865,
+      "learning_rate": 0.00018997697197712396,
+      "loss": 0.6196,
+      "step": 2305
+    },
+    {
+      "epoch": 0.1695712920067652,
+      "grad_norm": 1.254402756690979,
+      "learning_rate": 0.00018996657687584366,
+      "loss": 0.6354,
+      "step": 2306
+    },
+    {
+      "epoch": 0.1696448268255019,
+      "grad_norm": 1.4192092418670654,
+      "learning_rate": 0.00018995617667156423,
+      "loss": 0.6195,
+      "step": 2307
+    },
+    {
+      "epoch": 0.16971836164423854,
+      "grad_norm": 0.9442095756530762,
+      "learning_rate": 0.00018994577136487547,
+      "loss": 0.6084,
+      "step": 2308
+    },
+    {
+      "epoch": 0.1697918964629752,
+      "grad_norm": 2.9396746158599854,
+      "learning_rate": 0.00018993536095636765,
+      "loss": 0.7037,
+      "step": 2309
+    },
+    {
+      "epoch": 0.1698654312817119,
+      "grad_norm": 1.4431856870651245,
+      "learning_rate": 0.00018992494544663126,
+      "loss": 0.618,
+      "step": 2310
+    },
+    {
+      "epoch": 0.16993896610044856,
+      "grad_norm": 2.062535047531128,
+      "learning_rate": 0.00018991452483625704,
+      "loss": 0.6515,
+      "step": 2311
+    },
+    {
+      "epoch": 0.17001250091918524,
+      "grad_norm": 1.7103254795074463,
+      "learning_rate": 0.0001899040991258361,
+      "loss": 0.6455,
+      "step": 2312
+    },
+    {
+      "epoch": 0.17008603573792191,
+      "grad_norm": 1.691036581993103,
+      "learning_rate": 0.0001898936683159598,
+      "loss": 0.6185,
+      "step": 2313
+    },
+    {
+      "epoch": 0.1701595705566586,
+      "grad_norm": 1.527390480041504,
+      "learning_rate": 0.00018988323240721973,
+      "loss": 0.6297,
+      "step": 2314
+    },
+    {
+      "epoch": 0.17023310537539524,
+      "grad_norm": 1.482363224029541,
+      "learning_rate": 0.00018987279140020787,
+      "loss": 0.627,
+      "step": 2315
+    },
+    {
+      "epoch": 0.17030664019413191,
+      "grad_norm": 0.93297278881073,
+      "learning_rate": 0.00018986234529551642,
+      "loss": 0.659,
+      "step": 2316
+    },
+    {
+      "epoch": 0.1703801750128686,
+      "grad_norm": 1.7181750535964966,
+      "learning_rate": 0.0001898518940937379,
+      "loss": 0.6227,
+      "step": 2317
+    },
+    {
+      "epoch": 0.17045370983160527,
+      "grad_norm": 1.2104041576385498,
+      "learning_rate": 0.00018984143779546514,
+      "loss": 0.6275,
+      "step": 2318
+    },
+    {
+      "epoch": 0.17052724465034194,
+      "grad_norm": 2.253628730773926,
+      "learning_rate": 0.00018983097640129124,
+      "loss": 0.6076,
+      "step": 2319
+    },
+    {
+      "epoch": 0.17060077946907862,
+      "grad_norm": 2.3287172317504883,
+      "learning_rate": 0.0001898205099118095,
+      "loss": 0.6568,
+      "step": 2320
+    },
+    {
+      "epoch": 0.1706743142878153,
+      "grad_norm": 0.9878631830215454,
+      "learning_rate": 0.0001898100383276137,
+      "loss": 0.6209,
+      "step": 2321
+    },
+    {
+      "epoch": 0.17074784910655194,
+      "grad_norm": 1.8716177940368652,
+      "learning_rate": 0.0001897995616492977,
+      "loss": 0.6013,
+      "step": 2322
+    },
+    {
+      "epoch": 0.17082138392528862,
+      "grad_norm": 1.1976152658462524,
+      "learning_rate": 0.00018978907987745585,
+      "loss": 0.6325,
+      "step": 2323
+    },
+    {
+      "epoch": 0.1708949187440253,
+      "grad_norm": 2.5173821449279785,
+      "learning_rate": 0.0001897785930126826,
+      "loss": 0.6206,
+      "step": 2324
+    },
+    {
+      "epoch": 0.17096845356276197,
+      "grad_norm": 2.579288959503174,
+      "learning_rate": 0.00018976810105557278,
+      "loss": 0.6814,
+      "step": 2325
+    },
+    {
+      "epoch": 0.17104198838149864,
+      "grad_norm": 1.460637092590332,
+      "learning_rate": 0.00018975760400672157,
+      "loss": 0.6265,
+      "step": 2326
+    },
+    {
+      "epoch": 0.17111552320023532,
+      "grad_norm": 2.197322368621826,
+      "learning_rate": 0.00018974710186672432,
+      "loss": 0.6413,
+      "step": 2327
+    },
+    {
+      "epoch": 0.171189058018972,
+      "grad_norm": 1.6501250267028809,
+      "learning_rate": 0.00018973659463617675,
+      "loss": 0.654,
+      "step": 2328
+    },
+    {
+      "epoch": 0.17126259283770864,
+      "grad_norm": 1.9298173189163208,
+      "learning_rate": 0.00018972608231567484,
+      "loss": 0.6785,
+      "step": 2329
+    },
+    {
+      "epoch": 0.17133612765644532,
+      "grad_norm": 1.9188318252563477,
+      "learning_rate": 0.00018971556490581483,
+      "loss": 0.6679,
+      "step": 2330
+    },
+    {
+      "epoch": 0.171409662475182,
+      "grad_norm": 1.164028525352478,
+      "learning_rate": 0.0001897050424071933,
+      "loss": 0.6439,
+      "step": 2331
+    },
+    {
+      "epoch": 0.17148319729391867,
+      "grad_norm": 1.4542196989059448,
+      "learning_rate": 0.0001896945148204071,
+      "loss": 0.6482,
+      "step": 2332
+    },
+    {
+      "epoch": 0.17155673211265535,
+      "grad_norm": 1.097310185432434,
+      "learning_rate": 0.00018968398214605335,
+      "loss": 0.6486,
+      "step": 2333
+    },
+    {
+      "epoch": 0.17163026693139202,
+      "grad_norm": 1.3830466270446777,
+      "learning_rate": 0.0001896734443847295,
+      "loss": 0.652,
+      "step": 2334
+    },
+    {
+      "epoch": 0.1717038017501287,
+      "grad_norm": 0.9925356507301331,
+      "learning_rate": 0.0001896629015370332,
+      "loss": 0.633,
+      "step": 2335
+    },
+    {
+      "epoch": 0.17177733656886535,
+      "grad_norm": 1.3839046955108643,
+      "learning_rate": 0.00018965235360356256,
+      "loss": 0.6372,
+      "step": 2336
+    },
+    {
+      "epoch": 0.17185087138760202,
+      "grad_norm": 0.9662085175514221,
+      "learning_rate": 0.0001896418005849158,
+      "loss": 0.6636,
+      "step": 2337
+    },
+    {
+      "epoch": 0.1719244062063387,
+      "grad_norm": 1.427117109298706,
+      "learning_rate": 0.00018963124248169152,
+      "loss": 0.6444,
+      "step": 2338
+    },
+    {
+      "epoch": 0.17199794102507537,
+      "grad_norm": 0.8125438094139099,
+      "learning_rate": 0.00018962067929448852,
+      "loss": 0.6853,
+      "step": 2339
+    },
+    {
+      "epoch": 0.17207147584381205,
+      "grad_norm": 1.6870836019515991,
+      "learning_rate": 0.00018961011102390605,
+      "loss": 0.6517,
+      "step": 2340
+    },
+    {
+      "epoch": 0.17214501066254873,
+      "grad_norm": 0.8103761672973633,
+      "learning_rate": 0.00018959953767054352,
+      "loss": 0.6341,
+      "step": 2341
+    },
+    {
+      "epoch": 0.1722185454812854,
+      "grad_norm": 1.5713444948196411,
+      "learning_rate": 0.00018958895923500065,
+      "loss": 0.6655,
+      "step": 2342
+    },
+    {
+      "epoch": 0.17229208030002205,
+      "grad_norm": 1.496880292892456,
+      "learning_rate": 0.00018957837571787742,
+      "loss": 0.6633,
+      "step": 2343
+    },
+    {
+      "epoch": 0.17236561511875872,
+      "grad_norm": 1.8900264501571655,
+      "learning_rate": 0.00018956778711977422,
+      "loss": 0.6846,
+      "step": 2344
+    },
+    {
+      "epoch": 0.1724391499374954,
+      "grad_norm": 2.168450355529785,
+      "learning_rate": 0.00018955719344129163,
+      "loss": 0.6655,
+      "step": 2345
+    },
+    {
+      "epoch": 0.17251268475623208,
+      "grad_norm": 0.9526228308677673,
+      "learning_rate": 0.00018954659468303048,
+      "loss": 0.6592,
+      "step": 2346
+    },
+    {
+      "epoch": 0.17258621957496875,
+      "grad_norm": 1.656417727470398,
+      "learning_rate": 0.000189535990845592,
+      "loss": 0.6536,
+      "step": 2347
+    },
+    {
+      "epoch": 0.17265975439370543,
+      "grad_norm": 0.9767688512802124,
+      "learning_rate": 0.0001895253819295776,
+      "loss": 0.6682,
+      "step": 2348
+    },
+    {
+      "epoch": 0.1727332892124421,
+      "grad_norm": 1.0698906183242798,
+      "learning_rate": 0.0001895147679355891,
+      "loss": 0.6279,
+      "step": 2349
+    },
+    {
+      "epoch": 0.17280682403117875,
+      "grad_norm": 0.9796209335327148,
+      "learning_rate": 0.00018950414886422846,
+      "loss": 0.6504,
+      "step": 2350
+    },
+    {
+      "epoch": 0.17288035884991543,
+      "grad_norm": 1.0260835886001587,
+      "learning_rate": 0.00018949352471609804,
+      "loss": 0.644,
+      "step": 2351
+    },
+    {
+      "epoch": 0.1729538936686521,
+      "grad_norm": 0.9363652467727661,
+      "learning_rate": 0.00018948289549180046,
+      "loss": 0.6588,
+      "step": 2352
+    },
+    {
+      "epoch": 0.17302742848738878,
+      "grad_norm": 1.0535715818405151,
+      "learning_rate": 0.00018947226119193865,
+      "loss": 0.6255,
+      "step": 2353
+    },
+    {
+      "epoch": 0.17310096330612545,
+      "grad_norm": 1.2229914665222168,
+      "learning_rate": 0.0001894616218171157,
+      "loss": 0.6478,
+      "step": 2354
+    },
+    {
+      "epoch": 0.17317449812486213,
+      "grad_norm": 0.9887566566467285,
+      "learning_rate": 0.00018945097736793517,
+      "loss": 0.6399,
+      "step": 2355
+    },
+    {
+      "epoch": 0.1732480329435988,
+      "grad_norm": 1.2738713026046753,
+      "learning_rate": 0.0001894403278450008,
+      "loss": 0.6267,
+      "step": 2356
+    },
+    {
+      "epoch": 0.17332156776233545,
+      "grad_norm": 0.9084752202033997,
+      "learning_rate": 0.00018942967324891668,
+      "loss": 0.6495,
+      "step": 2357
+    },
+    {
+      "epoch": 0.17339510258107213,
+      "grad_norm": 1.068514347076416,
+      "learning_rate": 0.0001894190135802871,
+      "loss": 0.6332,
+      "step": 2358
+    },
+    {
+      "epoch": 0.1734686373998088,
+      "grad_norm": 0.8872542977333069,
+      "learning_rate": 0.00018940834883971665,
+      "loss": 0.6214,
+      "step": 2359
+    },
+    {
+      "epoch": 0.17354217221854548,
+      "grad_norm": 0.6484188437461853,
+      "learning_rate": 0.00018939767902781033,
+      "loss": 0.6304,
+      "step": 2360
+    },
+    {
+      "epoch": 0.17361570703728216,
+      "grad_norm": 0.6814944744110107,
+      "learning_rate": 0.00018938700414517333,
+      "loss": 0.6295,
+      "step": 2361
+    },
+    {
+      "epoch": 0.17368924185601883,
+      "grad_norm": 0.5889917612075806,
+      "learning_rate": 0.0001893763241924111,
+      "loss": 0.6112,
+      "step": 2362
+    },
+    {
+      "epoch": 0.1737627766747555,
+      "grad_norm": 0.685782790184021,
+      "learning_rate": 0.00018936563917012946,
+      "loss": 0.6479,
+      "step": 2363
+    },
+    {
+      "epoch": 0.17383631149349216,
+      "grad_norm": 0.746638298034668,
+      "learning_rate": 0.00018935494907893442,
+      "loss": 0.6434,
+      "step": 2364
+    },
+    {
+      "epoch": 0.17390984631222883,
+      "grad_norm": 1.0362392663955688,
+      "learning_rate": 0.0001893442539194324,
+      "loss": 0.6445,
+      "step": 2365
+    },
+    {
+      "epoch": 0.1739833811309655,
+      "grad_norm": 0.6451114416122437,
+      "learning_rate": 0.00018933355369223,
+      "loss": 0.6518,
+      "step": 2366
+    },
+    {
+      "epoch": 0.17405691594970218,
+      "grad_norm": 0.9554741978645325,
+      "learning_rate": 0.00018932284839793414,
+      "loss": 0.593,
+      "step": 2367
+    },
+    {
+      "epoch": 0.17413045076843886,
+      "grad_norm": 0.800299882888794,
+      "learning_rate": 0.0001893121380371521,
+      "loss": 0.6608,
+      "step": 2368
+    },
+    {
+      "epoch": 0.17420398558717554,
+      "grad_norm": 0.8765491247177124,
+      "learning_rate": 0.0001893014226104913,
+      "loss": 0.6428,
+      "step": 2369
+    },
+    {
+      "epoch": 0.1742775204059122,
+      "grad_norm": 1.1585795879364014,
+      "learning_rate": 0.00018929070211855955,
+      "loss": 0.6398,
+      "step": 2370
+    },
+    {
+      "epoch": 0.17435105522464886,
+      "grad_norm": 0.6409458518028259,
+      "learning_rate": 0.00018927997656196496,
+      "loss": 0.6428,
+      "step": 2371
+    },
+    {
+      "epoch": 0.17442459004338554,
+      "grad_norm": 0.6985546350479126,
+      "learning_rate": 0.0001892692459413159,
+      "loss": 0.6491,
+      "step": 2372
+    },
+    {
+      "epoch": 0.1744981248621222,
+      "grad_norm": 0.7139765024185181,
+      "learning_rate": 0.00018925851025722098,
+      "loss": 0.6265,
+      "step": 2373
+    },
+    {
+      "epoch": 0.1745716596808589,
+      "grad_norm": 0.6486931443214417,
+      "learning_rate": 0.00018924776951028915,
+      "loss": 0.6646,
+      "step": 2374
+    },
+    {
+      "epoch": 0.17464519449959556,
+      "grad_norm": 0.6680175065994263,
+      "learning_rate": 0.00018923702370112966,
+      "loss": 0.6428,
+      "step": 2375
+    },
+    {
+      "epoch": 0.17471872931833224,
+      "grad_norm": 0.9741879105567932,
+      "learning_rate": 0.000189226272830352,
+      "loss": 0.6198,
+      "step": 2376
+    },
+    {
+      "epoch": 0.17479226413706891,
+      "grad_norm": 0.7927713990211487,
+      "learning_rate": 0.000189215516898566,
+      "loss": 0.6267,
+      "step": 2377
+    },
+    {
+      "epoch": 0.17486579895580556,
+      "grad_norm": 0.7359263896942139,
+      "learning_rate": 0.0001892047559063817,
+      "loss": 0.6086,
+      "step": 2378
+    },
+    {
+      "epoch": 0.17493933377454224,
+      "grad_norm": 0.6756446957588196,
+      "learning_rate": 0.00018919398985440953,
+      "loss": 0.6153,
+      "step": 2379
+    },
+    {
+      "epoch": 0.17501286859327891,
+      "grad_norm": 0.5712445378303528,
+      "learning_rate": 0.0001891832187432601,
+      "loss": 0.6606,
+      "step": 2380
+    },
+    {
+      "epoch": 0.1750864034120156,
+      "grad_norm": 0.7253357768058777,
+      "learning_rate": 0.00018917244257354442,
+      "loss": 0.6367,
+      "step": 2381
+    },
+    {
+      "epoch": 0.17515993823075227,
+      "grad_norm": 0.8577306270599365,
+      "learning_rate": 0.00018916166134587368,
+      "loss": 0.6906,
+      "step": 2382
+    },
+    {
+      "epoch": 0.17523347304948894,
+      "grad_norm": 0.738875687122345,
+      "learning_rate": 0.0001891508750608594,
+      "loss": 0.6236,
+      "step": 2383
+    },
+    {
+      "epoch": 0.17530700786822562,
+      "grad_norm": 0.6176818013191223,
+      "learning_rate": 0.00018914008371911337,
+      "loss": 0.6135,
+      "step": 2384
+    },
+    {
+      "epoch": 0.17538054268696227,
+      "grad_norm": 1.2337210178375244,
+      "learning_rate": 0.00018912928732124778,
+      "loss": 0.6195,
+      "step": 2385
+    },
+    {
+      "epoch": 0.17545407750569894,
+      "grad_norm": 0.7826785445213318,
+      "learning_rate": 0.00018911848586787493,
+      "loss": 0.6158,
+      "step": 2386
+    },
+    {
+      "epoch": 0.17552761232443562,
+      "grad_norm": 0.6679283380508423,
+      "learning_rate": 0.0001891076793596075,
+      "loss": 0.674,
+      "step": 2387
+    },
+    {
+      "epoch": 0.1756011471431723,
+      "grad_norm": 0.7006273865699768,
+      "learning_rate": 0.00018909686779705848,
+      "loss": 0.653,
+      "step": 2388
+    },
+    {
+      "epoch": 0.17567468196190897,
+      "grad_norm": 0.7495198845863342,
+      "learning_rate": 0.00018908605118084104,
+      "loss": 0.6378,
+      "step": 2389
+    },
+    {
+      "epoch": 0.17574821678064564,
+      "grad_norm": 0.985997200012207,
+      "learning_rate": 0.0001890752295115688,
+      "loss": 0.6504,
+      "step": 2390
+    },
+    {
+      "epoch": 0.17582175159938232,
+      "grad_norm": 1.1532294750213623,
+      "learning_rate": 0.00018906440278985555,
+      "loss": 0.6216,
+      "step": 2391
+    },
+    {
+      "epoch": 0.17589528641811897,
+      "grad_norm": 0.8955446481704712,
+      "learning_rate": 0.00018905357101631535,
+      "loss": 0.6063,
+      "step": 2392
+    },
+    {
+      "epoch": 0.17596882123685564,
+      "grad_norm": 0.5352664589881897,
+      "learning_rate": 0.00018904273419156263,
+      "loss": 0.6506,
+      "step": 2393
+    },
+    {
+      "epoch": 0.17604235605559232,
+      "grad_norm": 0.7565745115280151,
+      "learning_rate": 0.0001890318923162121,
+      "loss": 0.638,
+      "step": 2394
+    },
+    {
+      "epoch": 0.176115890874329,
+      "grad_norm": 0.9584992527961731,
+      "learning_rate": 0.0001890210453908786,
+      "loss": 0.6591,
+      "step": 2395
+    },
+    {
+      "epoch": 0.17618942569306567,
+      "grad_norm": 1.001641869544983,
+      "learning_rate": 0.0001890101934161775,
+      "loss": 0.6408,
+      "step": 2396
+    },
+    {
+      "epoch": 0.17626296051180235,
+      "grad_norm": 0.9766911268234253,
+      "learning_rate": 0.00018899933639272427,
+      "loss": 0.6461,
+      "step": 2397
+    },
+    {
+      "epoch": 0.17633649533053902,
+      "grad_norm": 0.8393756151199341,
+      "learning_rate": 0.0001889884743211348,
+      "loss": 0.664,
+      "step": 2398
+    },
+    {
+      "epoch": 0.17641003014927567,
+      "grad_norm": 0.5598601698875427,
+      "learning_rate": 0.0001889776072020251,
+      "loss": 0.6264,
+      "step": 2399
+    },
+    {
+      "epoch": 0.17648356496801235,
+      "grad_norm": 1.1888681650161743,
+      "learning_rate": 0.00018896673503601165,
+      "loss": 0.6051,
+      "step": 2400
+    },
+    {
+      "epoch": 0.17655709978674902,
+      "grad_norm": 1.146113634109497,
+      "learning_rate": 0.0001889558578237111,
+      "loss": 0.6101,
+      "step": 2401
+    },
+    {
+      "epoch": 0.1766306346054857,
+      "grad_norm": 0.5168544054031372,
+      "learning_rate": 0.0001889449755657404,
+      "loss": 0.6293,
+      "step": 2402
+    },
+    {
+      "epoch": 0.17670416942422237,
+      "grad_norm": 1.0447853803634644,
+      "learning_rate": 0.00018893408826271683,
+      "loss": 0.6117,
+      "step": 2403
+    },
+    {
+      "epoch": 0.17677770424295905,
+      "grad_norm": 1.0512723922729492,
+      "learning_rate": 0.00018892319591525791,
+      "loss": 0.6097,
+      "step": 2404
+    },
+    {
+      "epoch": 0.17685123906169573,
+      "grad_norm": 0.5859243273735046,
+      "learning_rate": 0.00018891229852398148,
+      "loss": 0.6066,
+      "step": 2405
+    },
+    {
+      "epoch": 0.17692477388043237,
+      "grad_norm": 1.221914529800415,
+      "learning_rate": 0.00018890139608950562,
+      "loss": 0.6334,
+      "step": 2406
+    },
+    {
+      "epoch": 0.17699830869916905,
+      "grad_norm": 0.9196539521217346,
+      "learning_rate": 0.0001888904886124488,
+      "loss": 0.6576,
+      "step": 2407
+    },
+    {
+      "epoch": 0.17707184351790572,
+      "grad_norm": 0.7258049249649048,
+      "learning_rate": 0.00018887957609342964,
+      "loss": 0.6666,
+      "step": 2408
+    },
+    {
+      "epoch": 0.1771453783366424,
+      "grad_norm": 1.5319844484329224,
+      "learning_rate": 0.0001888686585330671,
+      "loss": 0.649,
+      "step": 2409
+    },
+    {
+      "epoch": 0.17721891315537908,
+      "grad_norm": 0.7653617262840271,
+      "learning_rate": 0.0001888577359319805,
+      "loss": 0.6483,
+      "step": 2410
+    },
+    {
+      "epoch": 0.17729244797411575,
+      "grad_norm": 2.168508529663086,
+      "learning_rate": 0.0001888468082907893,
+      "loss": 0.6557,
+      "step": 2411
+    },
+    {
+      "epoch": 0.17736598279285243,
+      "grad_norm": 1.2472912073135376,
+      "learning_rate": 0.0001888358756101134,
+      "loss": 0.6387,
+      "step": 2412
+    },
+    {
+      "epoch": 0.17743951761158908,
+      "grad_norm": 2.377868175506592,
+      "learning_rate": 0.00018882493789057288,
+      "loss": 0.6514,
+      "step": 2413
+    },
+    {
+      "epoch": 0.17751305243032575,
+      "grad_norm": 1.7930500507354736,
+      "learning_rate": 0.00018881399513278813,
+      "loss": 0.6656,
+      "step": 2414
+    },
+    {
+      "epoch": 0.17758658724906243,
+      "grad_norm": 1.5060869455337524,
+      "learning_rate": 0.00018880304733737987,
+      "loss": 0.6506,
+      "step": 2415
+    },
+    {
+      "epoch": 0.1776601220677991,
+      "grad_norm": 1.0525686740875244,
+      "learning_rate": 0.000188792094504969,
+      "loss": 0.6289,
+      "step": 2416
+    },
+    {
+      "epoch": 0.17773365688653578,
+      "grad_norm": 1.6921849250793457,
+      "learning_rate": 0.00018878113663617688,
+      "loss": 0.6343,
+      "step": 2417
+    },
+    {
+      "epoch": 0.17780719170527245,
+      "grad_norm": 1.3939992189407349,
+      "learning_rate": 0.00018877017373162492,
+      "loss": 0.6693,
+      "step": 2418
+    },
+    {
+      "epoch": 0.17788072652400913,
+      "grad_norm": 2.1664798259735107,
+      "learning_rate": 0.0001887592057919351,
+      "loss": 0.6368,
+      "step": 2419
+    },
+    {
+      "epoch": 0.17795426134274578,
+      "grad_norm": 1.9133051633834839,
+      "learning_rate": 0.00018874823281772943,
+      "loss": 0.634,
+      "step": 2420
+    },
+    {
+      "epoch": 0.17802779616148245,
+      "grad_norm": 1.2439926862716675,
+      "learning_rate": 0.0001887372548096303,
+      "loss": 0.6593,
+      "step": 2421
+    },
+    {
+      "epoch": 0.17810133098021913,
+      "grad_norm": 1.1519321203231812,
+      "learning_rate": 0.00018872627176826047,
+      "loss": 0.6427,
+      "step": 2422
+    },
+    {
+      "epoch": 0.1781748657989558,
+      "grad_norm": 1.6940947771072388,
+      "learning_rate": 0.00018871528369424287,
+      "loss": 0.6439,
+      "step": 2423
+    },
+    {
+      "epoch": 0.17824840061769248,
+      "grad_norm": 1.3342238664627075,
+      "learning_rate": 0.00018870429058820076,
+      "loss": 0.6618,
+      "step": 2424
+    },
+    {
+      "epoch": 0.17832193543642916,
+      "grad_norm": 1.748057246208191,
+      "learning_rate": 0.00018869329245075765,
+      "loss": 0.6703,
+      "step": 2425
+    },
+    {
+      "epoch": 0.17839547025516583,
+      "grad_norm": 1.6068716049194336,
+      "learning_rate": 0.00018868228928253743,
+      "loss": 0.6305,
+      "step": 2426
+    },
+    {
+      "epoch": 0.17846900507390248,
+      "grad_norm": 1.6096547842025757,
+      "learning_rate": 0.00018867128108416412,
+      "loss": 0.6272,
+      "step": 2427
+    },
+    {
+      "epoch": 0.17854253989263916,
+      "grad_norm": 1.3447901010513306,
+      "learning_rate": 0.00018866026785626218,
+      "loss": 0.6286,
+      "step": 2428
+    },
+    {
+      "epoch": 0.17861607471137583,
+      "grad_norm": 1.4777121543884277,
+      "learning_rate": 0.0001886492495994563,
+      "loss": 0.6724,
+      "step": 2429
+    },
+    {
+      "epoch": 0.1786896095301125,
+      "grad_norm": 1.2405800819396973,
+      "learning_rate": 0.00018863822631437144,
+      "loss": 0.6245,
+      "step": 2430
+    },
+    {
+      "epoch": 0.17876314434884918,
+      "grad_norm": 1.3386157751083374,
+      "learning_rate": 0.0001886271980016328,
+      "loss": 0.6087,
+      "step": 2431
+    },
+    {
+      "epoch": 0.17883667916758586,
+      "grad_norm": 1.3302925825119019,
+      "learning_rate": 0.00018861616466186597,
+      "loss": 0.6548,
+      "step": 2432
+    },
+    {
+      "epoch": 0.17891021398632254,
+      "grad_norm": 1.4922935962677002,
+      "learning_rate": 0.00018860512629569675,
+      "loss": 0.6473,
+      "step": 2433
+    },
+    {
+      "epoch": 0.17898374880505918,
+      "grad_norm": 1.336517095565796,
+      "learning_rate": 0.00018859408290375126,
+      "loss": 0.6091,
+      "step": 2434
+    },
+    {
+      "epoch": 0.17905728362379586,
+      "grad_norm": 1.428789496421814,
+      "learning_rate": 0.00018858303448665594,
+      "loss": 0.6101,
+      "step": 2435
+    },
+    {
+      "epoch": 0.17913081844253254,
+      "grad_norm": 1.346856713294983,
+      "learning_rate": 0.00018857198104503737,
+      "loss": 0.6398,
+      "step": 2436
+    },
+    {
+      "epoch": 0.1792043532612692,
+      "grad_norm": 1.3490830659866333,
+      "learning_rate": 0.00018856092257952257,
+      "loss": 0.67,
+      "step": 2437
+    },
+    {
+      "epoch": 0.1792778880800059,
+      "grad_norm": 1.345045566558838,
+      "learning_rate": 0.0001885498590907388,
+      "loss": 0.6639,
+      "step": 2438
+    },
+    {
+      "epoch": 0.17935142289874256,
+      "grad_norm": 1.466346263885498,
+      "learning_rate": 0.00018853879057931355,
+      "loss": 0.6501,
+      "step": 2439
+    },
+    {
+      "epoch": 0.17942495771747924,
+      "grad_norm": 1.0011495351791382,
+      "learning_rate": 0.00018852771704587465,
+      "loss": 0.6287,
+      "step": 2440
+    },
+    {
+      "epoch": 0.1794984925362159,
+      "grad_norm": 1.5924280881881714,
+      "learning_rate": 0.00018851663849105024,
+      "loss": 0.6126,
+      "step": 2441
+    },
+    {
+      "epoch": 0.17957202735495256,
+      "grad_norm": 1.3966381549835205,
+      "learning_rate": 0.00018850555491546866,
+      "loss": 0.6038,
+      "step": 2442
+    },
+    {
+      "epoch": 0.17964556217368924,
+      "grad_norm": 1.3580507040023804,
+      "learning_rate": 0.0001884944663197586,
+      "loss": 0.6071,
+      "step": 2443
+    },
+    {
+      "epoch": 0.17971909699242591,
+      "grad_norm": 1.263169527053833,
+      "learning_rate": 0.00018848337270454903,
+      "loss": 0.6491,
+      "step": 2444
+    },
+    {
+      "epoch": 0.1797926318111626,
+      "grad_norm": 1.35139000415802,
+      "learning_rate": 0.0001884722740704692,
+      "loss": 0.665,
+      "step": 2445
+    },
+    {
+      "epoch": 0.17986616662989927,
+      "grad_norm": 0.8954357504844666,
+      "learning_rate": 0.00018846117041814857,
+      "loss": 0.5962,
+      "step": 2446
+    },
+    {
+      "epoch": 0.17993970144863594,
+      "grad_norm": 1.187888503074646,
+      "learning_rate": 0.00018845006174821703,
+      "loss": 0.6191,
+      "step": 2447
+    },
+    {
+      "epoch": 0.1800132362673726,
+      "grad_norm": 0.7289806604385376,
+      "learning_rate": 0.00018843894806130462,
+      "loss": 0.6088,
+      "step": 2448
+    },
+    {
+      "epoch": 0.18008677108610927,
+      "grad_norm": 0.9459788203239441,
+      "learning_rate": 0.00018842782935804176,
+      "loss": 0.6529,
+      "step": 2449
+    },
+    {
+      "epoch": 0.18016030590484594,
+      "grad_norm": 0.6363592743873596,
+      "learning_rate": 0.0001884167056390591,
+      "loss": 0.646,
+      "step": 2450
+    },
+    {
+      "epoch": 0.18023384072358262,
+      "grad_norm": 0.7279635071754456,
+      "learning_rate": 0.00018840557690498754,
+      "loss": 0.6118,
+      "step": 2451
+    },
+    {
+      "epoch": 0.1803073755423193,
+      "grad_norm": 0.6494842171669006,
+      "learning_rate": 0.00018839444315645843,
+      "loss": 0.6055,
+      "step": 2452
+    },
+    {
+      "epoch": 0.18038091036105597,
+      "grad_norm": 0.706833004951477,
+      "learning_rate": 0.0001883833043941032,
+      "loss": 0.6387,
+      "step": 2453
+    },
+    {
+      "epoch": 0.18045444517979264,
+      "grad_norm": 0.6844576597213745,
+      "learning_rate": 0.00018837216061855363,
+      "loss": 0.6049,
+      "step": 2454
+    },
+    {
+      "epoch": 0.1805279799985293,
+      "grad_norm": 0.7195972204208374,
+      "learning_rate": 0.00018836101183044188,
+      "loss": 0.6359,
+      "step": 2455
+    },
+    {
+      "epoch": 0.18060151481726597,
+      "grad_norm": 0.8859819769859314,
+      "learning_rate": 0.00018834985803040027,
+      "loss": 0.6238,
+      "step": 2456
+    },
+    {
+      "epoch": 0.18067504963600264,
+      "grad_norm": 1.2456843852996826,
+      "learning_rate": 0.00018833869921906147,
+      "loss": 0.6514,
+      "step": 2457
+    },
+    {
+      "epoch": 0.18074858445473932,
+      "grad_norm": 0.650600254535675,
+      "learning_rate": 0.00018832753539705846,
+      "loss": 0.6422,
+      "step": 2458
+    },
+    {
+      "epoch": 0.180822119273476,
+      "grad_norm": 0.987863302230835,
+      "learning_rate": 0.0001883163665650244,
+      "loss": 0.6311,
+      "step": 2459
+    },
+    {
+      "epoch": 0.18089565409221267,
+      "grad_norm": 0.8133285641670227,
+      "learning_rate": 0.00018830519272359283,
+      "loss": 0.6211,
+      "step": 2460
+    },
+    {
+      "epoch": 0.18096918891094935,
+      "grad_norm": 0.6207874417304993,
+      "learning_rate": 0.0001882940138733975,
+      "loss": 0.5889,
+      "step": 2461
+    },
+    {
+      "epoch": 0.181042723729686,
+      "grad_norm": 0.5677153468132019,
+      "learning_rate": 0.00018828283001507255,
+      "loss": 0.5868,
+      "step": 2462
+    },
+    {
+      "epoch": 0.18111625854842267,
+      "grad_norm": 0.6169657707214355,
+      "learning_rate": 0.00018827164114925234,
+      "loss": 0.5841,
+      "step": 2463
+    },
+    {
+      "epoch": 0.18118979336715935,
+      "grad_norm": 0.6706321239471436,
+      "learning_rate": 0.00018826044727657148,
+      "loss": 0.6134,
+      "step": 2464
+    },
+    {
+      "epoch": 0.18126332818589602,
+      "grad_norm": 0.640723466873169,
+      "learning_rate": 0.00018824924839766488,
+      "loss": 0.6332,
+      "step": 2465
+    },
+    {
+      "epoch": 0.1813368630046327,
+      "grad_norm": 0.5510254502296448,
+      "learning_rate": 0.0001882380445131678,
+      "loss": 0.6243,
+      "step": 2466
+    },
+    {
+      "epoch": 0.18141039782336937,
+      "grad_norm": 0.9659594297409058,
+      "learning_rate": 0.0001882268356237157,
+      "loss": 0.6274,
+      "step": 2467
+    },
+    {
+      "epoch": 0.18148393264210605,
+      "grad_norm": 0.9006204605102539,
+      "learning_rate": 0.00018821562172994437,
+      "loss": 0.6136,
+      "step": 2468
+    },
+    {
+      "epoch": 0.1815574674608427,
+      "grad_norm": 0.818510115146637,
+      "learning_rate": 0.0001882044028324899,
+      "loss": 0.6428,
+      "step": 2469
+    },
+    {
+      "epoch": 0.18163100227957937,
+      "grad_norm": 0.6160674691200256,
+      "learning_rate": 0.0001881931789319886,
+      "loss": 0.6682,
+      "step": 2470
+    },
+    {
+      "epoch": 0.18170453709831605,
+      "grad_norm": 0.6267377734184265,
+      "learning_rate": 0.00018818195002907713,
+      "loss": 0.6258,
+      "step": 2471
+    },
+    {
+      "epoch": 0.18177807191705272,
+      "grad_norm": 0.8511665463447571,
+      "learning_rate": 0.00018817071612439235,
+      "loss": 0.641,
+      "step": 2472
+    },
+    {
+      "epoch": 0.1818516067357894,
+      "grad_norm": 1.062786340713501,
+      "learning_rate": 0.00018815947721857154,
+      "loss": 0.6611,
+      "step": 2473
+    },
+    {
+      "epoch": 0.18192514155452608,
+      "grad_norm": 1.1451472043991089,
+      "learning_rate": 0.00018814823331225213,
+      "loss": 0.6419,
+      "step": 2474
+    },
+    {
+      "epoch": 0.18199867637326275,
+      "grad_norm": 0.635116696357727,
+      "learning_rate": 0.0001881369844060719,
+      "loss": 0.6522,
+      "step": 2475
+    },
+    {
+      "epoch": 0.1820722111919994,
+      "grad_norm": 1.4851701259613037,
+      "learning_rate": 0.00018812573050066894,
+      "loss": 0.6405,
+      "step": 2476
+    },
+    {
+      "epoch": 0.18214574601073608,
+      "grad_norm": 0.8521588444709778,
+      "learning_rate": 0.0001881144715966815,
+      "loss": 0.5996,
+      "step": 2477
+    },
+    {
+      "epoch": 0.18221928082947275,
+      "grad_norm": 0.602632462978363,
+      "learning_rate": 0.00018810320769474824,
+      "loss": 0.6197,
+      "step": 2478
+    },
+    {
+      "epoch": 0.18229281564820943,
+      "grad_norm": 0.694490909576416,
+      "learning_rate": 0.00018809193879550806,
+      "loss": 0.6242,
+      "step": 2479
+    },
+    {
+      "epoch": 0.1823663504669461,
+      "grad_norm": 0.602527379989624,
+      "learning_rate": 0.00018808066489960017,
+      "loss": 0.6125,
+      "step": 2480
+    },
+    {
+      "epoch": 0.18243988528568278,
+      "grad_norm": 1.1874990463256836,
+      "learning_rate": 0.00018806938600766397,
+      "loss": 0.6721,
+      "step": 2481
+    },
+    {
+      "epoch": 0.18251342010441945,
+      "grad_norm": 1.239303469657898,
+      "learning_rate": 0.0001880581021203393,
+      "loss": 0.6127,
+      "step": 2482
+    },
+    {
+      "epoch": 0.1825869549231561,
+      "grad_norm": 0.7101176381111145,
+      "learning_rate": 0.00018804681323826613,
+      "loss": 0.6654,
+      "step": 2483
+    },
+    {
+      "epoch": 0.18266048974189278,
+      "grad_norm": 0.6113443374633789,
+      "learning_rate": 0.00018803551936208478,
+      "loss": 0.6622,
+      "step": 2484
+    },
+    {
+      "epoch": 0.18273402456062945,
+      "grad_norm": 0.8924757242202759,
+      "learning_rate": 0.00018802422049243587,
+      "loss": 0.6348,
+      "step": 2485
+    },
+    {
+      "epoch": 0.18280755937936613,
+      "grad_norm": 0.8048526048660278,
+      "learning_rate": 0.0001880129166299603,
+      "loss": 0.657,
+      "step": 2486
+    },
+    {
+      "epoch": 0.1828810941981028,
+      "grad_norm": 0.5154417753219604,
+      "learning_rate": 0.00018800160777529917,
+      "loss": 0.6334,
+      "step": 2487
+    },
+    {
+      "epoch": 0.18295462901683948,
+      "grad_norm": 0.8349583745002747,
+      "learning_rate": 0.00018799029392909403,
+      "loss": 0.6395,
+      "step": 2488
+    },
+    {
+      "epoch": 0.18302816383557616,
+      "grad_norm": 1.1897023916244507,
+      "learning_rate": 0.0001879789750919865,
+      "loss": 0.642,
+      "step": 2489
+    },
+    {
+      "epoch": 0.1831016986543128,
+      "grad_norm": 0.7908719778060913,
+      "learning_rate": 0.0001879676512646187,
+      "loss": 0.6098,
+      "step": 2490
+    },
+    {
+      "epoch": 0.18317523347304948,
+      "grad_norm": 0.8853146433830261,
+      "learning_rate": 0.00018795632244763283,
+      "loss": 0.609,
+      "step": 2491
+    },
+    {
+      "epoch": 0.18324876829178616,
+      "grad_norm": 1.5952218770980835,
+      "learning_rate": 0.00018794498864167157,
+      "loss": 0.5813,
+      "step": 2492
+    },
+    {
+      "epoch": 0.18332230311052283,
+      "grad_norm": 0.8653494119644165,
+      "learning_rate": 0.00018793364984737772,
+      "loss": 0.624,
+      "step": 2493
+    },
+    {
+      "epoch": 0.1833958379292595,
+      "grad_norm": 2.66925311088562,
+      "learning_rate": 0.00018792230606539453,
+      "loss": 0.6345,
+      "step": 2494
+    },
+    {
+      "epoch": 0.18346937274799618,
+      "grad_norm": 1.9769691228866577,
+      "learning_rate": 0.00018791095729636526,
+      "loss": 0.6273,
+      "step": 2495
+    },
+    {
+      "epoch": 0.18354290756673286,
+      "grad_norm": 1.7713263034820557,
+      "learning_rate": 0.00018789960354093375,
+      "loss": 0.628,
+      "step": 2496
+    },
+    {
+      "epoch": 0.1836164423854695,
+      "grad_norm": 2.0047659873962402,
+      "learning_rate": 0.00018788824479974397,
+      "loss": 0.6113,
+      "step": 2497
+    },
+    {
+      "epoch": 0.18368997720420618,
+      "grad_norm": 1.175247311592102,
+      "learning_rate": 0.00018787688107344016,
+      "loss": 0.6414,
+      "step": 2498
+    },
+    {
+      "epoch": 0.18376351202294286,
+      "grad_norm": 1.1116169691085815,
+      "learning_rate": 0.00018786551236266695,
+      "loss": 0.5803,
+      "step": 2499
+    },
+    {
+      "epoch": 0.18383704684167954,
+      "grad_norm": 1.7227249145507812,
+      "learning_rate": 0.00018785413866806913,
+      "loss": 0.6249,
+      "step": 2500
+    },
+    {
+      "epoch": 0.1839105816604162,
+      "grad_norm": 1.6032942533493042,
+      "learning_rate": 0.00018784275999029183,
+      "loss": 0.6233,
+      "step": 2501
+    },
+    {
+      "epoch": 0.1839841164791529,
+      "grad_norm": 1.5694438219070435,
+      "learning_rate": 0.0001878313763299805,
+      "loss": 0.607,
+      "step": 2502
+    },
+    {
+      "epoch": 0.18405765129788956,
+      "grad_norm": 1.604749083518982,
+      "learning_rate": 0.00018781998768778083,
+      "loss": 0.6164,
+      "step": 2503
+    },
+    {
+      "epoch": 0.1841311861166262,
+      "grad_norm": 1.1127599477767944,
+      "learning_rate": 0.00018780859406433877,
+      "loss": 0.6158,
+      "step": 2504
+    },
+    {
+      "epoch": 0.1842047209353629,
+      "grad_norm": 1.1234883069992065,
+      "learning_rate": 0.00018779719546030058,
+      "loss": 0.6333,
+      "step": 2505
+    },
+    {
+      "epoch": 0.18427825575409956,
+      "grad_norm": 1.5214883089065552,
+      "learning_rate": 0.00018778579187631283,
+      "loss": 0.6224,
+      "step": 2506
+    },
+    {
+      "epoch": 0.18435179057283624,
+      "grad_norm": 1.0095103979110718,
+      "learning_rate": 0.00018777438331302228,
+      "loss": 0.621,
+      "step": 2507
+    },
+    {
+      "epoch": 0.18442532539157291,
+      "grad_norm": 2.055384874343872,
+      "learning_rate": 0.00018776296977107611,
+      "loss": 0.62,
+      "step": 2508
+    },
+    {
+      "epoch": 0.1844988602103096,
+      "grad_norm": 1.8980218172073364,
+      "learning_rate": 0.00018775155125112165,
+      "loss": 0.6009,
+      "step": 2509
+    },
+    {
+      "epoch": 0.18457239502904627,
+      "grad_norm": 1.1719920635223389,
+      "learning_rate": 0.0001877401277538066,
+      "loss": 0.5846,
+      "step": 2510
+    },
+    {
+      "epoch": 0.1846459298477829,
+      "grad_norm": 1.2263938188552856,
+      "learning_rate": 0.00018772869927977893,
+      "loss": 0.6035,
+      "step": 2511
+    },
+    {
+      "epoch": 0.1847194646665196,
+      "grad_norm": 1.1605405807495117,
+      "learning_rate": 0.00018771726582968685,
+      "loss": 0.6229,
+      "step": 2512
+    },
+    {
+      "epoch": 0.18479299948525627,
+      "grad_norm": 0.7421529293060303,
+      "learning_rate": 0.00018770582740417884,
+      "loss": 0.6152,
+      "step": 2513
+    },
+    {
+      "epoch": 0.18486653430399294,
+      "grad_norm": 1.1941273212432861,
+      "learning_rate": 0.00018769438400390378,
+      "loss": 0.6545,
+      "step": 2514
+    },
+    {
+      "epoch": 0.18494006912272962,
+      "grad_norm": 0.5685896277427673,
+      "learning_rate": 0.00018768293562951069,
+      "loss": 0.6554,
+      "step": 2515
+    },
+    {
+      "epoch": 0.1850136039414663,
+      "grad_norm": 1.0790684223175049,
+      "learning_rate": 0.00018767148228164896,
+      "loss": 0.6169,
+      "step": 2516
+    },
+    {
+      "epoch": 0.18508713876020297,
+      "grad_norm": 0.7626508474349976,
+      "learning_rate": 0.00018766002396096828,
+      "loss": 0.6053,
+      "step": 2517
+    },
+    {
+      "epoch": 0.18516067357893962,
+      "grad_norm": 0.6930738687515259,
+      "learning_rate": 0.00018764856066811846,
+      "loss": 0.6071,
+      "step": 2518
+    },
+    {
+      "epoch": 0.1852342083976763,
+      "grad_norm": 0.9108526110649109,
+      "learning_rate": 0.0001876370924037498,
+      "loss": 0.6079,
+      "step": 2519
+    },
+    {
+      "epoch": 0.18530774321641297,
+      "grad_norm": 0.7088383436203003,
+      "learning_rate": 0.0001876256191685128,
+      "loss": 0.6137,
+      "step": 2520
+    },
+    {
+      "epoch": 0.18538127803514964,
+      "grad_norm": 0.6894176602363586,
+      "learning_rate": 0.00018761414096305816,
+      "loss": 0.5805,
+      "step": 2521
+    },
+    {
+      "epoch": 0.18545481285388632,
+      "grad_norm": 0.6716116666793823,
+      "learning_rate": 0.000187602657788037,
+      "loss": 0.6108,
+      "step": 2522
+    },
+    {
+      "epoch": 0.185528347672623,
+      "grad_norm": 0.8317681550979614,
+      "learning_rate": 0.00018759116964410066,
+      "loss": 0.582,
+      "step": 2523
+    },
+    {
+      "epoch": 0.18560188249135967,
+      "grad_norm": 0.7945351004600525,
+      "learning_rate": 0.0001875796765319007,
+      "loss": 0.6459,
+      "step": 2524
+    },
+    {
+      "epoch": 0.18567541731009632,
+      "grad_norm": 0.5775113701820374,
+      "learning_rate": 0.00018756817845208911,
+      "loss": 0.628,
+      "step": 2525
+    },
+    {
+      "epoch": 0.185748952128833,
+      "grad_norm": 1.0119482278823853,
+      "learning_rate": 0.000187556675405318,
+      "loss": 0.6096,
+      "step": 2526
+    },
+    {
+      "epoch": 0.18582248694756967,
+      "grad_norm": 1.4115809202194214,
+      "learning_rate": 0.00018754516739223983,
+      "loss": 0.5986,
+      "step": 2527
+    },
+    {
+      "epoch": 0.18589602176630635,
+      "grad_norm": 0.609552264213562,
+      "learning_rate": 0.00018753365441350744,
+      "loss": 0.5852,
+      "step": 2528
+    },
+    {
+      "epoch": 0.18596955658504302,
+      "grad_norm": 1.3508355617523193,
+      "learning_rate": 0.0001875221364697737,
+      "loss": 0.5926,
+      "step": 2529
+    },
+    {
+      "epoch": 0.1860430914037797,
+      "grad_norm": 0.8721228837966919,
+      "learning_rate": 0.0001875106135616921,
+      "loss": 0.5952,
+      "step": 2530
+    },
+    {
+      "epoch": 0.18611662622251637,
+      "grad_norm": 0.897666335105896,
+      "learning_rate": 0.00018749908568991612,
+      "loss": 0.5814,
+      "step": 2531
+    },
+    {
+      "epoch": 0.18619016104125302,
+      "grad_norm": 1.7026185989379883,
+      "learning_rate": 0.00018748755285509966,
+      "loss": 0.6315,
+      "step": 2532
+    },
+    {
+      "epoch": 0.1862636958599897,
+      "grad_norm": 0.8176699876785278,
+      "learning_rate": 0.00018747601505789686,
+      "loss": 0.5996,
+      "step": 2533
+    },
+    {
+      "epoch": 0.18633723067872637,
+      "grad_norm": 2.106527328491211,
+      "learning_rate": 0.00018746447229896218,
+      "loss": 0.6061,
+      "step": 2534
+    },
+    {
+      "epoch": 0.18641076549746305,
+      "grad_norm": 1.3225679397583008,
+      "learning_rate": 0.00018745292457895033,
+      "loss": 0.5825,
+      "step": 2535
+    },
+    {
+      "epoch": 0.18648430031619972,
+      "grad_norm": 2.511578321456909,
+      "learning_rate": 0.00018744137189851627,
+      "loss": 0.6134,
+      "step": 2536
+    },
+    {
+      "epoch": 0.1865578351349364,
+      "grad_norm": 2.805063486099243,
+      "learning_rate": 0.00018742981425831537,
+      "loss": 0.6241,
+      "step": 2537
+    },
+    {
+      "epoch": 0.18663136995367308,
+      "grad_norm": 0.953283965587616,
+      "learning_rate": 0.0001874182516590031,
+      "loss": 0.6155,
+      "step": 2538
+    },
+    {
+      "epoch": 0.18670490477240972,
+      "grad_norm": 2.506692886352539,
+      "learning_rate": 0.00018740668410123534,
+      "loss": 0.6108,
+      "step": 2539
+    },
+    {
+      "epoch": 0.1867784395911464,
+      "grad_norm": 1.9597082138061523,
+      "learning_rate": 0.00018739511158566823,
+      "loss": 0.6308,
+      "step": 2540
+    },
+    {
+      "epoch": 0.18685197440988308,
+      "grad_norm": 1.7609615325927734,
+      "learning_rate": 0.00018738353411295816,
+      "loss": 0.5895,
+      "step": 2541
+    },
+    {
+      "epoch": 0.18692550922861975,
+      "grad_norm": 1.8710994720458984,
+      "learning_rate": 0.0001873719516837618,
+      "loss": 0.5739,
+      "step": 2542
+    },
+    {
+      "epoch": 0.18699904404735643,
+      "grad_norm": 1.4223418235778809,
+      "learning_rate": 0.00018736036429873614,
+      "loss": 0.5949,
+      "step": 2543
+    },
+    {
+      "epoch": 0.1870725788660931,
+      "grad_norm": 1.4451613426208496,
+      "learning_rate": 0.00018734877195853843,
+      "loss": 0.608,
+      "step": 2544
+    },
+    {
+      "epoch": 0.18714611368482978,
+      "grad_norm": 1.3869320154190063,
+      "learning_rate": 0.00018733717466382617,
+      "loss": 0.5916,
+      "step": 2545
+    },
+    {
+      "epoch": 0.18721964850356643,
+      "grad_norm": 1.4215123653411865,
+      "learning_rate": 0.00018732557241525722,
+      "loss": 0.6095,
+      "step": 2546
+    },
+    {
+      "epoch": 0.1872931833223031,
+      "grad_norm": 1.0540564060211182,
+      "learning_rate": 0.0001873139652134896,
+      "loss": 0.6308,
+      "step": 2547
+    },
+    {
+      "epoch": 0.18736671814103978,
+      "grad_norm": 1.2577840089797974,
+      "learning_rate": 0.00018730235305918177,
+      "loss": 0.6169,
+      "step": 2548
+    },
+    {
+      "epoch": 0.18744025295977645,
+      "grad_norm": 1.042925238609314,
+      "learning_rate": 0.00018729073595299235,
+      "loss": 0.587,
+      "step": 2549
+    },
+    {
+      "epoch": 0.18751378777851313,
+      "grad_norm": 0.9990264773368835,
+      "learning_rate": 0.00018727911389558024,
+      "loss": 0.5899,
+      "step": 2550
+    },
+    {
+      "epoch": 0.1875873225972498,
+      "grad_norm": 0.9881891012191772,
+      "learning_rate": 0.00018726748688760466,
+      "loss": 0.6209,
+      "step": 2551
+    },
+    {
+      "epoch": 0.18766085741598648,
+      "grad_norm": 0.8893856406211853,
+      "learning_rate": 0.00018725585492972515,
+      "loss": 0.612,
+      "step": 2552
+    },
+    {
+      "epoch": 0.18773439223472313,
+      "grad_norm": 0.8498881459236145,
+      "learning_rate": 0.00018724421802260145,
+      "loss": 0.6,
+      "step": 2553
+    },
+    {
+      "epoch": 0.1878079270534598,
+      "grad_norm": 0.8825186491012573,
+      "learning_rate": 0.00018723257616689367,
+      "loss": 0.6047,
+      "step": 2554
+    },
+    {
+      "epoch": 0.18788146187219648,
+      "grad_norm": 0.5734856724739075,
+      "learning_rate": 0.0001872209293632621,
+      "loss": 0.5866,
+      "step": 2555
+    },
+    {
+      "epoch": 0.18795499669093316,
+      "grad_norm": 0.7695047855377197,
+      "learning_rate": 0.00018720927761236734,
+      "loss": 0.5814,
+      "step": 2556
+    },
+    {
+      "epoch": 0.18802853150966983,
+      "grad_norm": 0.624673068523407,
+      "learning_rate": 0.00018719762091487033,
+      "loss": 0.5946,
+      "step": 2557
+    },
+    {
+      "epoch": 0.1881020663284065,
+      "grad_norm": 0.7544350624084473,
+      "learning_rate": 0.00018718595927143223,
+      "loss": 0.6122,
+      "step": 2558
+    },
+    {
+      "epoch": 0.18817560114714318,
+      "grad_norm": 0.5978958010673523,
+      "learning_rate": 0.0001871742926827145,
+      "loss": 0.6015,
+      "step": 2559
+    },
+    {
+      "epoch": 0.18824913596587983,
+      "grad_norm": 0.6315548419952393,
+      "learning_rate": 0.00018716262114937892,
+      "loss": 0.597,
+      "step": 2560
+    },
+    {
+      "epoch": 0.1883226707846165,
+      "grad_norm": 0.5911280512809753,
+      "learning_rate": 0.00018715094467208749,
+      "loss": 0.6058,
+      "step": 2561
+    },
+    {
+      "epoch": 0.18839620560335318,
+      "grad_norm": 0.6241960525512695,
+      "learning_rate": 0.00018713926325150248,
+      "loss": 0.5839,
+      "step": 2562
+    },
+    {
+      "epoch": 0.18846974042208986,
+      "grad_norm": 0.6566564440727234,
+      "learning_rate": 0.0001871275768882865,
+      "loss": 0.5799,
+      "step": 2563
+    },
+    {
+      "epoch": 0.18854327524082654,
+      "grad_norm": 0.6381354928016663,
+      "learning_rate": 0.00018711588558310241,
+      "loss": 0.6226,
+      "step": 2564
+    },
+    {
+      "epoch": 0.1886168100595632,
+      "grad_norm": 0.609361469745636,
+      "learning_rate": 0.00018710418933661337,
+      "loss": 0.5963,
+      "step": 2565
+    },
+    {
+      "epoch": 0.1886903448782999,
+      "grad_norm": 0.625771164894104,
+      "learning_rate": 0.00018709248814948278,
+      "loss": 0.6017,
+      "step": 2566
+    },
+    {
+      "epoch": 0.18876387969703654,
+      "grad_norm": 0.7270624041557312,
+      "learning_rate": 0.00018708078202237432,
+      "loss": 0.6025,
+      "step": 2567
+    },
+    {
+      "epoch": 0.1888374145157732,
+      "grad_norm": 0.6815110445022583,
+      "learning_rate": 0.00018706907095595202,
+      "loss": 0.6267,
+      "step": 2568
+    },
+    {
+      "epoch": 0.1889109493345099,
+      "grad_norm": 0.5236886143684387,
+      "learning_rate": 0.00018705735495088015,
+      "loss": 0.5942,
+      "step": 2569
+    },
+    {
+      "epoch": 0.18898448415324656,
+      "grad_norm": 0.6847719550132751,
+      "learning_rate": 0.00018704563400782323,
+      "loss": 0.5863,
+      "step": 2570
+    },
+    {
+      "epoch": 0.18905801897198324,
+      "grad_norm": 0.6208638548851013,
+      "learning_rate": 0.00018703390812744607,
+      "loss": 0.5673,
+      "step": 2571
+    },
+    {
+      "epoch": 0.18913155379071991,
+      "grad_norm": 0.9410982131958008,
+      "learning_rate": 0.00018702217731041376,
+      "loss": 0.6222,
+      "step": 2572
+    },
+    {
+      "epoch": 0.1892050886094566,
+      "grad_norm": 1.1718195676803589,
+      "learning_rate": 0.00018701044155739177,
+      "loss": 0.6261,
+      "step": 2573
+    },
+    {
+      "epoch": 0.18927862342819324,
+      "grad_norm": 0.5894442200660706,
+      "learning_rate": 0.00018699870086904565,
+      "loss": 0.5821,
+      "step": 2574
+    },
+    {
+      "epoch": 0.1893521582469299,
+      "grad_norm": 0.8346141576766968,
+      "learning_rate": 0.00018698695524604144,
+      "loss": 0.5856,
+      "step": 2575
+    },
+    {
+      "epoch": 0.1894256930656666,
+      "grad_norm": 0.7329805493354797,
+      "learning_rate": 0.00018697520468904534,
+      "loss": 0.6098,
+      "step": 2576
+    },
+    {
+      "epoch": 0.18949922788440327,
+      "grad_norm": 0.6733750700950623,
+      "learning_rate": 0.00018696344919872382,
+      "loss": 0.6302,
+      "step": 2577
+    },
+    {
+      "epoch": 0.18957276270313994,
+      "grad_norm": 0.7079362273216248,
+      "learning_rate": 0.0001869516887757437,
+      "loss": 0.5878,
+      "step": 2578
+    },
+    {
+      "epoch": 0.18964629752187662,
+      "grad_norm": 1.174472689628601,
+      "learning_rate": 0.000186939923420772,
+      "loss": 0.5934,
+      "step": 2579
+    },
+    {
+      "epoch": 0.1897198323406133,
+      "grad_norm": 1.124983549118042,
+      "learning_rate": 0.00018692815313447612,
+      "loss": 0.6422,
+      "step": 2580
+    },
+    {
+      "epoch": 0.18979336715934994,
+      "grad_norm": 0.6888484954833984,
+      "learning_rate": 0.0001869163779175237,
+      "loss": 0.5881,
+      "step": 2581
+    },
+    {
+      "epoch": 0.18986690197808662,
+      "grad_norm": 0.957916796207428,
+      "learning_rate": 0.00018690459777058252,
+      "loss": 0.6235,
+      "step": 2582
+    },
+    {
+      "epoch": 0.1899404367968233,
+      "grad_norm": 1.3624428510665894,
+      "learning_rate": 0.00018689281269432083,
+      "loss": 0.6013,
+      "step": 2583
+    },
+    {
+      "epoch": 0.19001397161555997,
+      "grad_norm": 0.6880180239677429,
+      "learning_rate": 0.00018688102268940717,
+      "loss": 0.6033,
+      "step": 2584
+    },
+    {
+      "epoch": 0.19008750643429664,
+      "grad_norm": 1.6681435108184814,
+      "learning_rate": 0.0001868692277565102,
+      "loss": 0.6187,
+      "step": 2585
+    },
+    {
+      "epoch": 0.19016104125303332,
+      "grad_norm": 0.797166645526886,
+      "learning_rate": 0.00018685742789629892,
+      "loss": 0.6151,
+      "step": 2586
+    },
+    {
+      "epoch": 0.19023457607177,
+      "grad_norm": 1.7946076393127441,
+      "learning_rate": 0.0001868456231094427,
+      "loss": 0.6008,
+      "step": 2587
+    },
+    {
+      "epoch": 0.19030811089050664,
+      "grad_norm": 1.1425156593322754,
+      "learning_rate": 0.00018683381339661104,
+      "loss": 0.5994,
+      "step": 2588
+    },
+    {
+      "epoch": 0.19038164570924332,
+      "grad_norm": 1.900220274925232,
+      "learning_rate": 0.0001868219987584739,
+      "loss": 0.6116,
+      "step": 2589
+    },
+    {
+      "epoch": 0.19045518052798,
+      "grad_norm": 1.414361834526062,
+      "learning_rate": 0.0001868101791957013,
+      "loss": 0.61,
+      "step": 2590
+    },
+    {
+      "epoch": 0.19052871534671667,
+      "grad_norm": 1.4215556383132935,
+      "learning_rate": 0.0001867983547089638,
+      "loss": 0.6356,
+      "step": 2591
+    },
+    {
+      "epoch": 0.19060225016545335,
+      "grad_norm": 1.1677451133728027,
+      "learning_rate": 0.00018678652529893194,
+      "loss": 0.6012,
+      "step": 2592
+    },
+    {
+      "epoch": 0.19067578498419002,
+      "grad_norm": 1.6508255004882812,
+      "learning_rate": 0.00018677469096627683,
+      "loss": 0.622,
+      "step": 2593
+    },
+    {
+      "epoch": 0.1907493198029267,
+      "grad_norm": 0.9846705794334412,
+      "learning_rate": 0.00018676285171166966,
+      "loss": 0.6076,
+      "step": 2594
+    },
+    {
+      "epoch": 0.19082285462166335,
+      "grad_norm": 1.3708356618881226,
+      "learning_rate": 0.00018675100753578195,
+      "loss": 0.6145,
+      "step": 2595
+    },
+    {
+      "epoch": 0.19089638944040002,
+      "grad_norm": 1.0750364065170288,
+      "learning_rate": 0.00018673915843928557,
+      "loss": 0.5832,
+      "step": 2596
+    },
+    {
+      "epoch": 0.1909699242591367,
+      "grad_norm": 1.5163770914077759,
+      "learning_rate": 0.00018672730442285257,
+      "loss": 0.5838,
+      "step": 2597
+    },
+    {
+      "epoch": 0.19104345907787337,
+      "grad_norm": 1.1183280944824219,
+      "learning_rate": 0.00018671544548715536,
+      "loss": 0.611,
+      "step": 2598
+    },
+    {
+      "epoch": 0.19111699389661005,
+      "grad_norm": 1.257219910621643,
+      "learning_rate": 0.00018670358163286655,
+      "loss": 0.591,
+      "step": 2599
+    },
+    {
+      "epoch": 0.19119052871534672,
+      "grad_norm": 1.0314099788665771,
+      "learning_rate": 0.00018669171286065907,
+      "loss": 0.5903,
+      "step": 2600
+    },
+    {
+      "epoch": 0.1912640635340834,
+      "grad_norm": 1.0134953260421753,
+      "learning_rate": 0.00018667983917120616,
+      "loss": 0.6034,
+      "step": 2601
+    },
+    {
+      "epoch": 0.19133759835282005,
+      "grad_norm": 0.7701963186264038,
+      "learning_rate": 0.00018666796056518133,
+      "loss": 0.5928,
+      "step": 2602
+    },
+    {
+      "epoch": 0.19141113317155672,
+      "grad_norm": 0.9881088137626648,
+      "learning_rate": 0.00018665607704325827,
+      "loss": 0.606,
+      "step": 2603
+    },
+    {
+      "epoch": 0.1914846679902934,
+      "grad_norm": 0.7884853482246399,
+      "learning_rate": 0.00018664418860611109,
+      "loss": 0.5927,
+      "step": 2604
+    },
+    {
+      "epoch": 0.19155820280903008,
+      "grad_norm": 1.0345321893692017,
+      "learning_rate": 0.00018663229525441407,
+      "loss": 0.6175,
+      "step": 2605
+    },
+    {
+      "epoch": 0.19163173762776675,
+      "grad_norm": 0.8919309377670288,
+      "learning_rate": 0.00018662039698884186,
+      "loss": 0.5919,
+      "step": 2606
+    },
+    {
+      "epoch": 0.19170527244650343,
+      "grad_norm": 1.3705370426177979,
+      "learning_rate": 0.0001866084938100693,
+      "loss": 0.6427,
+      "step": 2607
+    },
+    {
+      "epoch": 0.1917788072652401,
+      "grad_norm": 1.0683646202087402,
+      "learning_rate": 0.00018659658571877155,
+      "loss": 0.5668,
+      "step": 2608
+    },
+    {
+      "epoch": 0.19185234208397675,
+      "grad_norm": 1.5266605615615845,
+      "learning_rate": 0.00018658467271562412,
+      "loss": 0.6361,
+      "step": 2609
+    },
+    {
+      "epoch": 0.19192587690271343,
+      "grad_norm": 1.3683440685272217,
+      "learning_rate": 0.00018657275480130268,
+      "loss": 0.6017,
+      "step": 2610
+    },
+    {
+      "epoch": 0.1919994117214501,
+      "grad_norm": 0.9810406565666199,
+      "learning_rate": 0.00018656083197648318,
+      "loss": 0.601,
+      "step": 2611
+    },
+    {
+      "epoch": 0.19207294654018678,
+      "grad_norm": 1.029162049293518,
+      "learning_rate": 0.00018654890424184194,
+      "loss": 0.5991,
+      "step": 2612
+    },
+    {
+      "epoch": 0.19214648135892345,
+      "grad_norm": 0.8589317798614502,
+      "learning_rate": 0.00018653697159805554,
+      "loss": 0.6174,
+      "step": 2613
+    },
+    {
+      "epoch": 0.19222001617766013,
+      "grad_norm": 1.4821535348892212,
+      "learning_rate": 0.00018652503404580079,
+      "loss": 0.5902,
+      "step": 2614
+    },
+    {
+      "epoch": 0.1922935509963968,
+      "grad_norm": 0.5905015468597412,
+      "learning_rate": 0.00018651309158575476,
+      "loss": 0.6037,
+      "step": 2615
+    },
+    {
+      "epoch": 0.19236708581513345,
+      "grad_norm": 1.052579402923584,
+      "learning_rate": 0.0001865011442185949,
+      "loss": 0.5764,
+      "step": 2616
+    },
+    {
+      "epoch": 0.19244062063387013,
+      "grad_norm": 1.1403275728225708,
+      "learning_rate": 0.00018648919194499885,
+      "loss": 0.6052,
+      "step": 2617
+    },
+    {
+      "epoch": 0.1925141554526068,
+      "grad_norm": 0.8102548122406006,
+      "learning_rate": 0.00018647723476564453,
+      "loss": 0.5769,
+      "step": 2618
+    },
+    {
+      "epoch": 0.19258769027134348,
+      "grad_norm": 0.7563402652740479,
+      "learning_rate": 0.00018646527268121024,
+      "loss": 0.5671,
+      "step": 2619
+    },
+    {
+      "epoch": 0.19266122509008016,
+      "grad_norm": 0.7522573471069336,
+      "learning_rate": 0.0001864533056923744,
+      "loss": 0.5808,
+      "step": 2620
+    },
+    {
+      "epoch": 0.19273475990881683,
+      "grad_norm": 0.6415799856185913,
+      "learning_rate": 0.00018644133379981583,
+      "loss": 0.6298,
+      "step": 2621
+    },
+    {
+      "epoch": 0.1928082947275535,
+      "grad_norm": 0.8194668292999268,
+      "learning_rate": 0.00018642935700421357,
+      "loss": 0.5986,
+      "step": 2622
+    },
+    {
+      "epoch": 0.19288182954629016,
+      "grad_norm": 0.8817406892776489,
+      "learning_rate": 0.000186417375306247,
+      "loss": 0.6117,
+      "step": 2623
+    },
+    {
+      "epoch": 0.19295536436502683,
+      "grad_norm": 0.7329156398773193,
+      "learning_rate": 0.00018640538870659567,
+      "loss": 0.6123,
+      "step": 2624
+    },
+    {
+      "epoch": 0.1930288991837635,
+      "grad_norm": 0.647241473197937,
+      "learning_rate": 0.0001863933972059395,
+      "loss": 0.6043,
+      "step": 2625
+    },
+    {
+      "epoch": 0.19310243400250018,
+      "grad_norm": 0.9314565062522888,
+      "learning_rate": 0.00018638140080495872,
+      "loss": 0.6047,
+      "step": 2626
+    },
+    {
+      "epoch": 0.19317596882123686,
+      "grad_norm": 0.9918014407157898,
+      "learning_rate": 0.00018636939950433367,
+      "loss": 0.5707,
+      "step": 2627
+    },
+    {
+      "epoch": 0.19324950363997354,
+      "grad_norm": 0.6554892659187317,
+      "learning_rate": 0.00018635739330474514,
+      "loss": 0.5842,
+      "step": 2628
+    },
+    {
+      "epoch": 0.1933230384587102,
+      "grad_norm": 0.5905621647834778,
+      "learning_rate": 0.00018634538220687415,
+      "loss": 0.6099,
+      "step": 2629
+    },
+    {
+      "epoch": 0.19339657327744686,
+      "grad_norm": 0.7396208047866821,
+      "learning_rate": 0.00018633336621140195,
+      "loss": 0.6108,
+      "step": 2630
+    },
+    {
+      "epoch": 0.19347010809618354,
+      "grad_norm": 0.736533522605896,
+      "learning_rate": 0.0001863213453190101,
+      "loss": 0.6038,
+      "step": 2631
+    },
+    {
+      "epoch": 0.1935436429149202,
+      "grad_norm": 0.5711544156074524,
+      "learning_rate": 0.00018630931953038043,
+      "loss": 0.6193,
+      "step": 2632
+    },
+    {
+      "epoch": 0.1936171777336569,
+      "grad_norm": 0.6944261789321899,
+      "learning_rate": 0.00018629728884619507,
+      "loss": 0.6013,
+      "step": 2633
+    },
+    {
+      "epoch": 0.19369071255239356,
+      "grad_norm": 0.7786366939544678,
+      "learning_rate": 0.00018628525326713647,
+      "loss": 0.6036,
+      "step": 2634
+    },
+    {
+      "epoch": 0.19376424737113024,
+      "grad_norm": 0.6435895562171936,
+      "learning_rate": 0.00018627321279388718,
+      "loss": 0.5625,
+      "step": 2635
+    },
+    {
+      "epoch": 0.19383778218986691,
+      "grad_norm": 0.5853399038314819,
+      "learning_rate": 0.00018626116742713025,
+      "loss": 0.5889,
+      "step": 2636
+    },
+    {
+      "epoch": 0.19391131700860356,
+      "grad_norm": 0.6711776256561279,
+      "learning_rate": 0.00018624911716754884,
+      "loss": 0.6007,
+      "step": 2637
+    },
+    {
+      "epoch": 0.19398485182734024,
+      "grad_norm": 1.0083293914794922,
+      "learning_rate": 0.00018623706201582651,
+      "loss": 0.6036,
+      "step": 2638
+    },
+    {
+      "epoch": 0.1940583866460769,
+      "grad_norm": 1.4460477828979492,
+      "learning_rate": 0.000186225001972647,
+      "loss": 0.6064,
+      "step": 2639
+    },
+    {
+      "epoch": 0.1941319214648136,
+      "grad_norm": 0.765835165977478,
+      "learning_rate": 0.00018621293703869438,
+      "loss": 0.62,
+      "step": 2640
+    },
+    {
+      "epoch": 0.19420545628355027,
+      "grad_norm": 0.7039175033569336,
+      "learning_rate": 0.00018620086721465298,
+      "loss": 0.6136,
+      "step": 2641
+    },
+    {
+      "epoch": 0.19427899110228694,
+      "grad_norm": 0.5521687865257263,
+      "learning_rate": 0.00018618879250120743,
+      "loss": 0.6311,
+      "step": 2642
+    },
+    {
+      "epoch": 0.19435252592102362,
+      "grad_norm": 0.8889257907867432,
+      "learning_rate": 0.0001861767128990426,
+      "loss": 0.5992,
+      "step": 2643
+    },
+    {
+      "epoch": 0.19442606073976026,
+      "grad_norm": 1.1538890600204468,
+      "learning_rate": 0.00018616462840884368,
+      "loss": 0.6086,
+      "step": 2644
+    },
+    {
+      "epoch": 0.19449959555849694,
+      "grad_norm": 0.829679012298584,
+      "learning_rate": 0.0001861525390312961,
+      "loss": 0.6154,
+      "step": 2645
+    },
+    {
+      "epoch": 0.19457313037723362,
+      "grad_norm": 0.6388718485832214,
+      "learning_rate": 0.00018614044476708558,
+      "loss": 0.6472,
+      "step": 2646
+    },
+    {
+      "epoch": 0.1946466651959703,
+      "grad_norm": 0.7009196281433105,
+      "learning_rate": 0.00018612834561689814,
+      "loss": 0.5916,
+      "step": 2647
+    },
+    {
+      "epoch": 0.19472020001470697,
+      "grad_norm": 1.0598026514053345,
+      "learning_rate": 0.00018611624158142002,
+      "loss": 0.6275,
+      "step": 2648
+    },
+    {
+      "epoch": 0.19479373483344364,
+      "grad_norm": 1.1666706800460815,
+      "learning_rate": 0.0001861041326613378,
+      "loss": 0.6305,
+      "step": 2649
+    },
+    {
+      "epoch": 0.19486726965218032,
+      "grad_norm": 0.8397045731544495,
+      "learning_rate": 0.00018609201885733833,
+      "loss": 0.6212,
+      "step": 2650
+    },
+    {
+      "epoch": 0.19494080447091697,
+      "grad_norm": 0.6869182586669922,
+      "learning_rate": 0.0001860799001701087,
+      "loss": 0.6004,
+      "step": 2651
+    },
+    {
+      "epoch": 0.19501433928965364,
+      "grad_norm": 0.5893464684486389,
+      "learning_rate": 0.00018606777660033625,
+      "loss": 0.5934,
+      "step": 2652
+    },
+    {
+      "epoch": 0.19508787410839032,
+      "grad_norm": 0.7054097652435303,
+      "learning_rate": 0.00018605564814870872,
+      "loss": 0.6173,
+      "step": 2653
+    },
+    {
+      "epoch": 0.195161408927127,
+      "grad_norm": 0.8400381803512573,
+      "learning_rate": 0.00018604351481591396,
+      "loss": 0.5941,
+      "step": 2654
+    },
+    {
+      "epoch": 0.19523494374586367,
+      "grad_norm": 1.7146016359329224,
+      "learning_rate": 0.00018603137660264026,
+      "loss": 0.6116,
+      "step": 2655
+    },
+    {
+      "epoch": 0.19530847856460035,
+      "grad_norm": 0.7051020264625549,
+      "learning_rate": 0.00018601923350957606,
+      "loss": 0.5896,
+      "step": 2656
+    },
+    {
+      "epoch": 0.19538201338333702,
+      "grad_norm": 1.3579139709472656,
+      "learning_rate": 0.0001860070855374102,
+      "loss": 0.6064,
+      "step": 2657
+    },
+    {
+      "epoch": 0.19545554820207367,
+      "grad_norm": 1.2967723608016968,
+      "learning_rate": 0.0001859949326868317,
+      "loss": 0.5926,
+      "step": 2658
+    },
+    {
+      "epoch": 0.19552908302081035,
+      "grad_norm": 0.6965769529342651,
+      "learning_rate": 0.0001859827749585298,
+      "loss": 0.604,
+      "step": 2659
+    },
+    {
+      "epoch": 0.19560261783954702,
+      "grad_norm": 2.036062717437744,
+      "learning_rate": 0.00018597061235319418,
+      "loss": 0.6026,
+      "step": 2660
+    },
+    {
+      "epoch": 0.1956761526582837,
+      "grad_norm": 0.7885276675224304,
+      "learning_rate": 0.00018595844487151472,
+      "loss": 0.5941,
+      "step": 2661
+    },
+    {
+      "epoch": 0.19574968747702037,
+      "grad_norm": 2.5202133655548096,
+      "learning_rate": 0.00018594627251418153,
+      "loss": 0.656,
+      "step": 2662
+    },
+    {
+      "epoch": 0.19582322229575705,
+      "grad_norm": 1.5507333278656006,
+      "learning_rate": 0.00018593409528188507,
+      "loss": 0.6212,
+      "step": 2663
+    },
+    {
+      "epoch": 0.19589675711449372,
+      "grad_norm": 2.539299726486206,
+      "learning_rate": 0.00018592191317531604,
+      "loss": 0.6002,
+      "step": 2664
+    },
+    {
+      "epoch": 0.19597029193323037,
+      "grad_norm": 2.5797300338745117,
+      "learning_rate": 0.00018590972619516541,
+      "loss": 0.6447,
+      "step": 2665
+    },
+    {
+      "epoch": 0.19604382675196705,
+      "grad_norm": 1.0408356189727783,
+      "learning_rate": 0.00018589753434212445,
+      "loss": 0.6159,
+      "step": 2666
+    },
+    {
+      "epoch": 0.19611736157070372,
+      "grad_norm": 1.6915240287780762,
+      "learning_rate": 0.00018588533761688473,
+      "loss": 0.642,
+      "step": 2667
+    },
+    {
+      "epoch": 0.1961908963894404,
+      "grad_norm": 1.4477404356002808,
+      "learning_rate": 0.00018587313602013797,
+      "loss": 0.6037,
+      "step": 2668
+    },
+    {
+      "epoch": 0.19626443120817708,
+      "grad_norm": 1.331308126449585,
+      "learning_rate": 0.0001858609295525763,
+      "loss": 0.609,
+      "step": 2669
+    },
+    {
+      "epoch": 0.19633796602691375,
+      "grad_norm": 1.5054805278778076,
+      "learning_rate": 0.00018584871821489211,
+      "loss": 0.5905,
+      "step": 2670
+    },
+    {
+      "epoch": 0.19641150084565043,
+      "grad_norm": 0.7719610929489136,
+      "learning_rate": 0.00018583650200777805,
+      "loss": 0.5965,
+      "step": 2671
+    },
+    {
+      "epoch": 0.19648503566438708,
+      "grad_norm": 1.8635362386703491,
+      "learning_rate": 0.000185824280931927,
+      "loss": 0.6188,
+      "step": 2672
+    },
+    {
+      "epoch": 0.19655857048312375,
+      "grad_norm": 1.0359195470809937,
+      "learning_rate": 0.00018581205498803213,
+      "loss": 0.6151,
+      "step": 2673
+    },
+    {
+      "epoch": 0.19663210530186043,
+      "grad_norm": 2.0720858573913574,
+      "learning_rate": 0.00018579982417678696,
+      "loss": 0.5874,
+      "step": 2674
+    },
+    {
+      "epoch": 0.1967056401205971,
+      "grad_norm": 1.8580702543258667,
+      "learning_rate": 0.0001857875884988852,
+      "loss": 0.6251,
+      "step": 2675
+    },
+    {
+      "epoch": 0.19677917493933378,
+      "grad_norm": 1.4339314699172974,
+      "learning_rate": 0.00018577534795502092,
+      "loss": 0.5912,
+      "step": 2676
+    },
+    {
+      "epoch": 0.19685270975807045,
+      "grad_norm": 0.9470658302307129,
+      "learning_rate": 0.00018576310254588837,
+      "loss": 0.5979,
+      "step": 2677
+    },
+    {
+      "epoch": 0.19692624457680713,
+      "grad_norm": 1.2630304098129272,
+      "learning_rate": 0.00018575085227218212,
+      "loss": 0.641,
+      "step": 2678
+    },
+    {
+      "epoch": 0.19699977939554378,
+      "grad_norm": 0.8351110816001892,
+      "learning_rate": 0.00018573859713459702,
+      "loss": 0.5814,
+      "step": 2679
+    },
+    {
+      "epoch": 0.19707331421428045,
+      "grad_norm": 0.9818311333656311,
+      "learning_rate": 0.00018572633713382822,
+      "loss": 0.5907,
+      "step": 2680
+    },
+    {
+      "epoch": 0.19714684903301713,
+      "grad_norm": 1.3519964218139648,
+      "learning_rate": 0.0001857140722705711,
+      "loss": 0.6097,
+      "step": 2681
+    },
+    {
+      "epoch": 0.1972203838517538,
+      "grad_norm": 0.6550007462501526,
+      "learning_rate": 0.0001857018025455214,
+      "loss": 0.6155,
+      "step": 2682
+    },
+    {
+      "epoch": 0.19729391867049048,
+      "grad_norm": 1.1406787633895874,
+      "learning_rate": 0.00018568952795937495,
+      "loss": 0.6099,
+      "step": 2683
+    },
+    {
+      "epoch": 0.19736745348922716,
+      "grad_norm": 0.8925144076347351,
+      "learning_rate": 0.00018567724851282805,
+      "loss": 0.6126,
+      "step": 2684
+    },
+    {
+      "epoch": 0.19744098830796383,
+      "grad_norm": 1.087206244468689,
+      "learning_rate": 0.00018566496420657723,
+      "loss": 0.6375,
+      "step": 2685
+    },
+    {
+      "epoch": 0.19751452312670048,
+      "grad_norm": 0.9500728845596313,
+      "learning_rate": 0.00018565267504131922,
+      "loss": 0.6107,
+      "step": 2686
+    },
+    {
+      "epoch": 0.19758805794543716,
+      "grad_norm": 0.9598010778427124,
+      "learning_rate": 0.0001856403810177511,
+      "loss": 0.6175,
+      "step": 2687
+    },
+    {
+      "epoch": 0.19766159276417383,
+      "grad_norm": 0.6356861591339111,
+      "learning_rate": 0.00018562808213657013,
+      "loss": 0.6508,
+      "step": 2688
+    },
+    {
+      "epoch": 0.1977351275829105,
+      "grad_norm": 0.9652954339981079,
+      "learning_rate": 0.00018561577839847405,
+      "loss": 0.6118,
+      "step": 2689
+    },
+    {
+      "epoch": 0.19780866240164718,
+      "grad_norm": 1.1374725103378296,
+      "learning_rate": 0.0001856034698041606,
+      "loss": 0.6038,
+      "step": 2690
+    },
+    {
+      "epoch": 0.19788219722038386,
+      "grad_norm": 0.6031773090362549,
+      "learning_rate": 0.00018559115635432804,
+      "loss": 0.6229,
+      "step": 2691
+    },
+    {
+      "epoch": 0.19795573203912054,
+      "grad_norm": 1.1228481531143188,
+      "learning_rate": 0.00018557883804967476,
+      "loss": 0.6265,
+      "step": 2692
+    },
+    {
+      "epoch": 0.19802926685785718,
+      "grad_norm": 0.7712710499763489,
+      "learning_rate": 0.00018556651489089947,
+      "loss": 0.6105,
+      "step": 2693
+    },
+    {
+      "epoch": 0.19810280167659386,
+      "grad_norm": 0.6362004280090332,
+      "learning_rate": 0.00018555418687870115,
+      "loss": 0.6212,
+      "step": 2694
+    },
+    {
+      "epoch": 0.19817633649533054,
+      "grad_norm": 0.823523223400116,
+      "learning_rate": 0.00018554185401377902,
+      "loss": 0.5899,
+      "step": 2695
+    },
+    {
+      "epoch": 0.1982498713140672,
+      "grad_norm": 1.1002023220062256,
+      "learning_rate": 0.00018552951629683274,
+      "loss": 0.6003,
+      "step": 2696
+    },
+    {
+      "epoch": 0.1983234061328039,
+      "grad_norm": 0.9217124581336975,
+      "learning_rate": 0.00018551717372856198,
+      "loss": 0.5966,
+      "step": 2697
+    },
+    {
+      "epoch": 0.19839694095154056,
+      "grad_norm": 0.8850702047348022,
+      "learning_rate": 0.00018550482630966687,
+      "loss": 0.6247,
+      "step": 2698
+    },
+    {
+      "epoch": 0.19847047577027724,
+      "grad_norm": 1.2420393228530884,
+      "learning_rate": 0.0001854924740408478,
+      "loss": 0.6212,
+      "step": 2699
+    },
+    {
+      "epoch": 0.1985440105890139,
+      "grad_norm": 1.1946107149124146,
+      "learning_rate": 0.00018548011692280536,
+      "loss": 0.6293,
+      "step": 2700
+    },
+    {
+      "epoch": 0.19861754540775056,
+      "grad_norm": 0.8282182812690735,
+      "learning_rate": 0.00018546775495624049,
+      "loss": 0.6243,
+      "step": 2701
+    },
+    {
+      "epoch": 0.19869108022648724,
+      "grad_norm": 0.5624870657920837,
+      "learning_rate": 0.00018545538814185436,
+      "loss": 0.5623,
+      "step": 2702
+    },
+    {
+      "epoch": 0.1987646150452239,
+      "grad_norm": 0.8853166699409485,
+      "learning_rate": 0.0001854430164803484,
+      "loss": 0.6004,
+      "step": 2703
+    },
+    {
+      "epoch": 0.1988381498639606,
+      "grad_norm": 0.98769611120224,
+      "learning_rate": 0.00018543063997242444,
+      "loss": 0.607,
+      "step": 2704
+    },
+    {
+      "epoch": 0.19891168468269727,
+      "grad_norm": 0.8825187683105469,
+      "learning_rate": 0.00018541825861878437,
+      "loss": 0.6097,
+      "step": 2705
+    },
+    {
+      "epoch": 0.19898521950143394,
+      "grad_norm": 0.9986087679862976,
+      "learning_rate": 0.00018540587242013055,
+      "loss": 0.5923,
+      "step": 2706
+    },
+    {
+      "epoch": 0.1990587543201706,
+      "grad_norm": 1.813644528388977,
+      "learning_rate": 0.00018539348137716548,
+      "loss": 0.6464,
+      "step": 2707
+    },
+    {
+      "epoch": 0.19913228913890726,
+      "grad_norm": 0.7313733100891113,
+      "learning_rate": 0.00018538108549059204,
+      "loss": 0.6243,
+      "step": 2708
+    },
+    {
+      "epoch": 0.19920582395764394,
+      "grad_norm": 1.108949899673462,
+      "learning_rate": 0.00018536868476111336,
+      "loss": 0.6108,
+      "step": 2709
+    },
+    {
+      "epoch": 0.19927935877638062,
+      "grad_norm": 2.3575363159179688,
+      "learning_rate": 0.00018535627918943274,
+      "loss": 0.5995,
+      "step": 2710
+    },
+    {
+      "epoch": 0.1993528935951173,
+      "grad_norm": 1.336165189743042,
+      "learning_rate": 0.0001853438687762539,
+      "loss": 0.5984,
+      "step": 2711
+    },
+    {
+      "epoch": 0.19942642841385397,
+      "grad_norm": 3.089513063430786,
+      "learning_rate": 0.00018533145352228077,
+      "loss": 0.59,
+      "step": 2712
+    },
+    {
+      "epoch": 0.19949996323259064,
+      "grad_norm": 4.282916069030762,
+      "learning_rate": 0.00018531903342821756,
+      "loss": 0.6549,
+      "step": 2713
+    },
+    {
+      "epoch": 0.1995734980513273,
+      "grad_norm": 0.9941296577453613,
+      "learning_rate": 0.0001853066084947687,
+      "loss": 0.6425,
+      "step": 2714
+    },
+    {
+      "epoch": 0.19964703287006397,
+      "grad_norm": 3.8138771057128906,
+      "learning_rate": 0.00018529417872263897,
+      "loss": 0.6302,
+      "step": 2715
+    },
+    {
+      "epoch": 0.19972056768880064,
+      "grad_norm": 4.88326358795166,
+      "learning_rate": 0.0001852817441125334,
+      "loss": 0.6605,
+      "step": 2716
+    },
+    {
+      "epoch": 0.19979410250753732,
+      "grad_norm": 2.2104556560516357,
+      "learning_rate": 0.00018526930466515734,
+      "loss": 0.6576,
+      "step": 2717
+    },
+    {
+      "epoch": 0.199867637326274,
+      "grad_norm": 3.0793299674987793,
+      "learning_rate": 0.00018525686038121632,
+      "loss": 0.6472,
+      "step": 2718
+    },
+    {
+      "epoch": 0.19994117214501067,
+      "grad_norm": 3.9877707958221436,
+      "learning_rate": 0.00018524441126141622,
+      "loss": 0.6629,
+      "step": 2719
+    },
+    {
+      "epoch": 0.20001470696374735,
+      "grad_norm": 1.618975281715393,
+      "learning_rate": 0.00018523195730646313,
+      "loss": 0.6808,
+      "step": 2720
+    },
+    {
+      "epoch": 0.200088241782484,
+      "grad_norm": 2.023980140686035,
+      "learning_rate": 0.00018521949851706346,
+      "loss": 0.6634,
+      "step": 2721
+    },
+    {
+      "epoch": 0.20016177660122067,
+      "grad_norm": 2.454664468765259,
+      "learning_rate": 0.00018520703489392394,
+      "loss": 0.6313,
+      "step": 2722
+    },
+    {
+      "epoch": 0.20023531141995735,
+      "grad_norm": 1.1909340620040894,
+      "learning_rate": 0.00018519456643775143,
+      "loss": 0.628,
+      "step": 2723
+    },
+    {
+      "epoch": 0.20030884623869402,
+      "grad_norm": 2.1574463844299316,
+      "learning_rate": 0.00018518209314925322,
+      "loss": 0.6042,
+      "step": 2724
+    },
+    {
+      "epoch": 0.2003823810574307,
+      "grad_norm": 1.9951987266540527,
+      "learning_rate": 0.00018516961502913678,
+      "loss": 0.6191,
+      "step": 2725
+    },
+    {
+      "epoch": 0.20045591587616737,
+      "grad_norm": 1.3998901844024658,
+      "learning_rate": 0.00018515713207810995,
+      "loss": 0.5886,
+      "step": 2726
+    },
+    {
+      "epoch": 0.20052945069490405,
+      "grad_norm": 1.3838775157928467,
+      "learning_rate": 0.00018514464429688065,
+      "loss": 0.6341,
+      "step": 2727
+    },
+    {
+      "epoch": 0.2006029855136407,
+      "grad_norm": 1.484343409538269,
+      "learning_rate": 0.0001851321516861573,
+      "loss": 0.642,
+      "step": 2728
+    },
+    {
+      "epoch": 0.20067652033237737,
+      "grad_norm": 1.091775894165039,
+      "learning_rate": 0.00018511965424664845,
+      "loss": 0.602,
+      "step": 2729
+    },
+    {
+      "epoch": 0.20075005515111405,
+      "grad_norm": 1.336108684539795,
+      "learning_rate": 0.000185107151979063,
+      "loss": 0.661,
+      "step": 2730
+    },
+    {
+      "epoch": 0.20082358996985072,
+      "grad_norm": 0.9037541151046753,
+      "learning_rate": 0.00018509464488411005,
+      "loss": 0.5827,
+      "step": 2731
+    },
+    {
+      "epoch": 0.2008971247885874,
+      "grad_norm": 1.4259541034698486,
+      "learning_rate": 0.00018508213296249904,
+      "loss": 0.6098,
+      "step": 2732
+    },
+    {
+      "epoch": 0.20097065960732408,
+      "grad_norm": 1.0449965000152588,
+      "learning_rate": 0.00018506961621493967,
+      "loss": 0.6168,
+      "step": 2733
+    },
+    {
+      "epoch": 0.20104419442606075,
+      "grad_norm": 1.4430785179138184,
+      "learning_rate": 0.0001850570946421419,
+      "loss": 0.6073,
+      "step": 2734
+    },
+    {
+      "epoch": 0.2011177292447974,
+      "grad_norm": 1.1358067989349365,
+      "learning_rate": 0.00018504456824481593,
+      "loss": 0.5998,
+      "step": 2735
+    },
+    {
+      "epoch": 0.20119126406353408,
+      "grad_norm": 1.3810760974884033,
+      "learning_rate": 0.0001850320370236723,
+      "loss": 0.5907,
+      "step": 2736
+    },
+    {
+      "epoch": 0.20126479888227075,
+      "grad_norm": 1.3511619567871094,
+      "learning_rate": 0.00018501950097942182,
+      "loss": 0.6148,
+      "step": 2737
+    },
+    {
+      "epoch": 0.20133833370100743,
+      "grad_norm": 1.5508346557617188,
+      "learning_rate": 0.0001850069601127755,
+      "loss": 0.6365,
+      "step": 2738
+    },
+    {
+      "epoch": 0.2014118685197441,
+      "grad_norm": 1.2557309865951538,
+      "learning_rate": 0.0001849944144244447,
+      "loss": 0.6133,
+      "step": 2739
+    },
+    {
+      "epoch": 0.20148540333848078,
+      "grad_norm": 1.1549961566925049,
+      "learning_rate": 0.000184981863915141,
+      "loss": 0.6145,
+      "step": 2740
+    },
+    {
+      "epoch": 0.20155893815721745,
+      "grad_norm": 0.9326412677764893,
+      "learning_rate": 0.00018496930858557633,
+      "loss": 0.6327,
+      "step": 2741
+    },
+    {
+      "epoch": 0.2016324729759541,
+      "grad_norm": 0.6434475183486938,
+      "learning_rate": 0.00018495674843646277,
+      "loss": 0.6181,
+      "step": 2742
+    },
+    {
+      "epoch": 0.20170600779469078,
+      "grad_norm": 0.8477945327758789,
+      "learning_rate": 0.0001849441834685128,
+      "loss": 0.6439,
+      "step": 2743
+    },
+    {
+      "epoch": 0.20177954261342745,
+      "grad_norm": 0.6484140753746033,
+      "learning_rate": 0.00018493161368243906,
+      "loss": 0.6127,
+      "step": 2744
+    },
+    {
+      "epoch": 0.20185307743216413,
+      "grad_norm": 0.7552365660667419,
+      "learning_rate": 0.00018491903907895461,
+      "loss": 0.6062,
+      "step": 2745
+    },
+    {
+      "epoch": 0.2019266122509008,
+      "grad_norm": 0.8764052391052246,
+      "learning_rate": 0.00018490645965877261,
+      "loss": 0.6412,
+      "step": 2746
+    },
+    {
+      "epoch": 0.20200014706963748,
+      "grad_norm": 0.7289955019950867,
+      "learning_rate": 0.00018489387542260666,
+      "loss": 0.6392,
+      "step": 2747
+    },
+    {
+      "epoch": 0.20207368188837416,
+      "grad_norm": 0.7674228549003601,
+      "learning_rate": 0.00018488128637117045,
+      "loss": 0.6068,
+      "step": 2748
+    },
+    {
+      "epoch": 0.2021472167071108,
+      "grad_norm": 0.9960017204284668,
+      "learning_rate": 0.00018486869250517813,
+      "loss": 0.5892,
+      "step": 2749
+    },
+    {
+      "epoch": 0.20222075152584748,
+      "grad_norm": 0.5643746852874756,
+      "learning_rate": 0.00018485609382534399,
+      "loss": 0.5911,
+      "step": 2750
+    },
+    {
+      "epoch": 0.20229428634458416,
+      "grad_norm": 0.7195028066635132,
+      "learning_rate": 0.00018484349033238265,
+      "loss": 0.5995,
+      "step": 2751
+    },
+    {
+      "epoch": 0.20236782116332083,
+      "grad_norm": 0.7853057980537415,
+      "learning_rate": 0.00018483088202700902,
+      "loss": 0.5874,
+      "step": 2752
+    },
+    {
+      "epoch": 0.2024413559820575,
+      "grad_norm": 0.6539014577865601,
+      "learning_rate": 0.0001848182689099382,
+      "loss": 0.6193,
+      "step": 2753
+    },
+    {
+      "epoch": 0.20251489080079418,
+      "grad_norm": 0.7477661371231079,
+      "learning_rate": 0.0001848056509818857,
+      "loss": 0.6276,
+      "step": 2754
+    },
+    {
+      "epoch": 0.20258842561953086,
+      "grad_norm": 0.6099528670310974,
+      "learning_rate": 0.00018479302824356717,
+      "loss": 0.6059,
+      "step": 2755
+    },
+    {
+      "epoch": 0.2026619604382675,
+      "grad_norm": 0.5932965874671936,
+      "learning_rate": 0.00018478040069569858,
+      "loss": 0.64,
+      "step": 2756
+    },
+    {
+      "epoch": 0.20273549525700418,
+      "grad_norm": 0.595656156539917,
+      "learning_rate": 0.00018476776833899617,
+      "loss": 0.6168,
+      "step": 2757
+    },
+    {
+      "epoch": 0.20280903007574086,
+      "grad_norm": 0.6290574073791504,
+      "learning_rate": 0.00018475513117417652,
+      "loss": 0.6234,
+      "step": 2758
+    },
+    {
+      "epoch": 0.20288256489447754,
+      "grad_norm": 0.6073041558265686,
+      "learning_rate": 0.00018474248920195637,
+      "loss": 0.6071,
+      "step": 2759
+    },
+    {
+      "epoch": 0.2029560997132142,
+      "grad_norm": 0.7035950422286987,
+      "learning_rate": 0.0001847298424230528,
+      "loss": 0.6189,
+      "step": 2760
+    },
+    {
+      "epoch": 0.2030296345319509,
+      "grad_norm": 0.7748203277587891,
+      "learning_rate": 0.00018471719083818316,
+      "loss": 0.6502,
+      "step": 2761
+    },
+    {
+      "epoch": 0.20310316935068756,
+      "grad_norm": 0.7469415068626404,
+      "learning_rate": 0.000184704534448065,
+      "loss": 0.6005,
+      "step": 2762
+    },
+    {
+      "epoch": 0.2031767041694242,
+      "grad_norm": 0.7277822494506836,
+      "learning_rate": 0.0001846918732534163,
+      "loss": 0.6049,
+      "step": 2763
+    },
+    {
+      "epoch": 0.2032502389881609,
+      "grad_norm": 0.7235754132270813,
+      "learning_rate": 0.00018467920725495515,
+      "loss": 0.6141,
+      "step": 2764
+    },
+    {
+      "epoch": 0.20332377380689756,
+      "grad_norm": 0.5846988558769226,
+      "learning_rate": 0.0001846665364534,
+      "loss": 0.6097,
+      "step": 2765
+    },
+    {
+      "epoch": 0.20339730862563424,
+      "grad_norm": 0.5091707110404968,
+      "learning_rate": 0.00018465386084946957,
+      "loss": 0.6153,
+      "step": 2766
+    },
+    {
+      "epoch": 0.2034708434443709,
+      "grad_norm": 0.5590506792068481,
+      "learning_rate": 0.0001846411804438828,
+      "loss": 0.6349,
+      "step": 2767
+    },
+    {
+      "epoch": 0.2035443782631076,
+      "grad_norm": 0.6284058094024658,
+      "learning_rate": 0.00018462849523735896,
+      "loss": 0.5977,
+      "step": 2768
+    },
+    {
+      "epoch": 0.20361791308184427,
+      "grad_norm": 0.8471804261207581,
+      "learning_rate": 0.00018461580523061754,
+      "loss": 0.6092,
+      "step": 2769
+    },
+    {
+      "epoch": 0.2036914479005809,
+      "grad_norm": 1.1226083040237427,
+      "learning_rate": 0.00018460311042437836,
+      "loss": 0.6199,
+      "step": 2770
+    },
+    {
+      "epoch": 0.2037649827193176,
+      "grad_norm": 0.7744565606117249,
+      "learning_rate": 0.0001845904108193615,
+      "loss": 0.5886,
+      "step": 2771
+    },
+    {
+      "epoch": 0.20383851753805426,
+      "grad_norm": 0.6791051626205444,
+      "learning_rate": 0.00018457770641628723,
+      "loss": 0.6196,
+      "step": 2772
+    },
+    {
+      "epoch": 0.20391205235679094,
+      "grad_norm": 0.8685276508331299,
+      "learning_rate": 0.0001845649972158762,
+      "loss": 0.6335,
+      "step": 2773
+    },
+    {
+      "epoch": 0.20398558717552762,
+      "grad_norm": 0.5966765284538269,
+      "learning_rate": 0.00018455228321884927,
+      "loss": 0.5888,
+      "step": 2774
+    },
+    {
+      "epoch": 0.2040591219942643,
+      "grad_norm": 0.7251604795455933,
+      "learning_rate": 0.0001845395644259276,
+      "loss": 0.6039,
+      "step": 2775
+    },
+    {
+      "epoch": 0.20413265681300097,
+      "grad_norm": 0.8381624221801758,
+      "learning_rate": 0.00018452684083783264,
+      "loss": 0.6023,
+      "step": 2776
+    },
+    {
+      "epoch": 0.20420619163173762,
+      "grad_norm": 1.27189040184021,
+      "learning_rate": 0.0001845141124552861,
+      "loss": 0.6029,
+      "step": 2777
+    },
+    {
+      "epoch": 0.2042797264504743,
+      "grad_norm": 0.7759961485862732,
+      "learning_rate": 0.00018450137927900984,
+      "loss": 0.6144,
+      "step": 2778
+    },
+    {
+      "epoch": 0.20435326126921097,
+      "grad_norm": 0.7660143375396729,
+      "learning_rate": 0.0001844886413097262,
+      "loss": 0.6271,
+      "step": 2779
+    },
+    {
+      "epoch": 0.20442679608794764,
+      "grad_norm": 1.7705330848693848,
+      "learning_rate": 0.00018447589854815764,
+      "loss": 0.6165,
+      "step": 2780
+    },
+    {
+      "epoch": 0.20450033090668432,
+      "grad_norm": 0.8070230484008789,
+      "learning_rate": 0.000184463150995027,
+      "loss": 0.6008,
+      "step": 2781
+    },
+    {
+      "epoch": 0.204573865725421,
+      "grad_norm": 1.1774746179580688,
+      "learning_rate": 0.0001844503986510573,
+      "loss": 0.6241,
+      "step": 2782
+    },
+    {
+      "epoch": 0.20464740054415767,
+      "grad_norm": 1.869691252708435,
+      "learning_rate": 0.00018443764151697184,
+      "loss": 0.6199,
+      "step": 2783
+    },
+    {
+      "epoch": 0.20472093536289432,
+      "grad_norm": 0.7466079592704773,
+      "learning_rate": 0.00018442487959349424,
+      "loss": 0.6193,
+      "step": 2784
+    },
+    {
+      "epoch": 0.204794470181631,
+      "grad_norm": 2.214674711227417,
+      "learning_rate": 0.00018441211288134837,
+      "loss": 0.591,
+      "step": 2785
+    },
+    {
+      "epoch": 0.20486800500036767,
+      "grad_norm": 1.4366860389709473,
+      "learning_rate": 0.0001843993413812584,
+      "loss": 0.6102,
+      "step": 2786
+    },
+    {
+      "epoch": 0.20494153981910435,
+      "grad_norm": 3.0760812759399414,
+      "learning_rate": 0.0001843865650939487,
+      "loss": 0.6043,
+      "step": 2787
+    },
+    {
+      "epoch": 0.20501507463784102,
+      "grad_norm": 3.171787738800049,
+      "learning_rate": 0.00018437378402014396,
+      "loss": 0.6236,
+      "step": 2788
+    },
+    {
+      "epoch": 0.2050886094565777,
+      "grad_norm": 0.7676699161529541,
+      "learning_rate": 0.00018436099816056915,
+      "loss": 0.5989,
+      "step": 2789
+    },
+    {
+      "epoch": 0.20516214427531437,
+      "grad_norm": 2.452209949493408,
+      "learning_rate": 0.0001843482075159495,
+      "loss": 0.5952,
+      "step": 2790
+    },
+    {
+      "epoch": 0.20523567909405102,
+      "grad_norm": 2.0733721256256104,
+      "learning_rate": 0.00018433541208701053,
+      "loss": 0.5944,
+      "step": 2791
+    },
+    {
+      "epoch": 0.2053092139127877,
+      "grad_norm": 1.4743577241897583,
+      "learning_rate": 0.00018432261187447795,
+      "loss": 0.5922,
+      "step": 2792
+    },
+    {
+      "epoch": 0.20538274873152437,
+      "grad_norm": 1.585784673690796,
+      "learning_rate": 0.00018430980687907783,
+      "loss": 0.5953,
+      "step": 2793
+    },
+    {
+      "epoch": 0.20545628355026105,
+      "grad_norm": 1.3927419185638428,
+      "learning_rate": 0.0001842969971015365,
+      "loss": 0.5882,
+      "step": 2794
+    },
+    {
+      "epoch": 0.20552981836899772,
+      "grad_norm": 1.0952582359313965,
+      "learning_rate": 0.00018428418254258055,
+      "loss": 0.5871,
+      "step": 2795
+    },
+    {
+      "epoch": 0.2056033531877344,
+      "grad_norm": 1.7112733125686646,
+      "learning_rate": 0.0001842713632029368,
+      "loss": 0.6025,
+      "step": 2796
+    },
+    {
+      "epoch": 0.20567688800647108,
+      "grad_norm": 1.05496084690094,
+      "learning_rate": 0.00018425853908333241,
+      "loss": 0.5927,
+      "step": 2797
+    },
+    {
+      "epoch": 0.20575042282520772,
+      "grad_norm": 2.323965311050415,
+      "learning_rate": 0.00018424571018449472,
+      "loss": 0.5861,
+      "step": 2798
+    },
+    {
+      "epoch": 0.2058239576439444,
+      "grad_norm": 1.9002940654754639,
+      "learning_rate": 0.0001842328765071515,
+      "loss": 0.5652,
+      "step": 2799
+    },
+    {
+      "epoch": 0.20589749246268108,
+      "grad_norm": 1.8309842348098755,
+      "learning_rate": 0.0001842200380520306,
+      "loss": 0.6156,
+      "step": 2800
+    },
+    {
+      "epoch": 0.20597102728141775,
+      "grad_norm": 1.388517141342163,
+      "learning_rate": 0.00018420719481986026,
+      "loss": 0.5837,
+      "step": 2801
+    },
+    {
+      "epoch": 0.20604456210015443,
+      "grad_norm": 1.557358980178833,
+      "learning_rate": 0.00018419434681136892,
+      "loss": 0.5733,
+      "step": 2802
+    },
+    {
+      "epoch": 0.2061180969188911,
+      "grad_norm": 1.28078293800354,
+      "learning_rate": 0.00018418149402728547,
+      "loss": 0.5965,
+      "step": 2803
+    },
+    {
+      "epoch": 0.20619163173762778,
+      "grad_norm": 2.0768957138061523,
+      "learning_rate": 0.00018416863646833877,
+      "loss": 0.606,
+      "step": 2804
+    },
+    {
+      "epoch": 0.20626516655636443,
+      "grad_norm": 1.8228501081466675,
+      "learning_rate": 0.0001841557741352582,
+      "loss": 0.5812,
+      "step": 2805
+    },
+    {
+      "epoch": 0.2063387013751011,
+      "grad_norm": 1.4607256650924683,
+      "learning_rate": 0.0001841429070287733,
+      "loss": 0.5873,
+      "step": 2806
+    },
+    {
+      "epoch": 0.20641223619383778,
+      "grad_norm": 1.4164206981658936,
+      "learning_rate": 0.00018413003514961395,
+      "loss": 0.5848,
+      "step": 2807
+    },
+    {
+      "epoch": 0.20648577101257445,
+      "grad_norm": 1.198466181755066,
+      "learning_rate": 0.00018411715849851013,
+      "loss": 0.5915,
+      "step": 2808
+    },
+    {
+      "epoch": 0.20655930583131113,
+      "grad_norm": 1.1063514947891235,
+      "learning_rate": 0.0001841042770761924,
+      "loss": 0.5871,
+      "step": 2809
+    },
+    {
+      "epoch": 0.2066328406500478,
+      "grad_norm": 1.5778658390045166,
+      "learning_rate": 0.0001840913908833913,
+      "loss": 0.6093,
+      "step": 2810
+    },
+    {
+      "epoch": 0.20670637546878448,
+      "grad_norm": 1.0965237617492676,
+      "learning_rate": 0.00018407849992083773,
+      "loss": 0.571,
+      "step": 2811
+    },
+    {
+      "epoch": 0.20677991028752113,
+      "grad_norm": 1.945220947265625,
+      "learning_rate": 0.00018406560418926293,
+      "loss": 0.5746,
+      "step": 2812
+    },
+    {
+      "epoch": 0.2068534451062578,
+      "grad_norm": 1.5386419296264648,
+      "learning_rate": 0.00018405270368939834,
+      "loss": 0.6352,
+      "step": 2813
+    },
+    {
+      "epoch": 0.20692697992499448,
+      "grad_norm": 1.4466360807418823,
+      "learning_rate": 0.00018403979842197571,
+      "loss": 0.5588,
+      "step": 2814
+    },
+    {
+      "epoch": 0.20700051474373116,
+      "grad_norm": 1.3098206520080566,
+      "learning_rate": 0.00018402688838772702,
+      "loss": 0.5646,
+      "step": 2815
+    },
+    {
+      "epoch": 0.20707404956246783,
+      "grad_norm": 1.3013105392456055,
+      "learning_rate": 0.00018401397358738453,
+      "loss": 0.5747,
+      "step": 2816
+    },
+    {
+      "epoch": 0.2071475843812045,
+      "grad_norm": 1.1836161613464355,
+      "learning_rate": 0.00018400105402168078,
+      "loss": 0.5883,
+      "step": 2817
+    },
+    {
+      "epoch": 0.20722111919994118,
+      "grad_norm": 1.8129907846450806,
+      "learning_rate": 0.00018398812969134862,
+      "loss": 0.591,
+      "step": 2818
+    },
+    {
+      "epoch": 0.20729465401867783,
+      "grad_norm": 1.784497857093811,
+      "learning_rate": 0.0001839752005971211,
+      "loss": 0.5924,
+      "step": 2819
+    },
+    {
+      "epoch": 0.2073681888374145,
+      "grad_norm": 1.353366494178772,
+      "learning_rate": 0.00018396226673973156,
+      "loss": 0.572,
+      "step": 2820
+    },
+    {
+      "epoch": 0.20744172365615118,
+      "grad_norm": 1.4704248905181885,
+      "learning_rate": 0.00018394932811991368,
+      "loss": 0.5619,
+      "step": 2821
+    },
+    {
+      "epoch": 0.20751525847488786,
+      "grad_norm": 1.1631700992584229,
+      "learning_rate": 0.0001839363847384013,
+      "loss": 0.5617,
+      "step": 2822
+    },
+    {
+      "epoch": 0.20758879329362454,
+      "grad_norm": 1.3120712041854858,
+      "learning_rate": 0.00018392343659592858,
+      "loss": 0.5556,
+      "step": 2823
+    },
+    {
+      "epoch": 0.2076623281123612,
+      "grad_norm": 1.3754310607910156,
+      "learning_rate": 0.00018391048369323,
+      "loss": 0.5942,
+      "step": 2824
+    },
+    {
+      "epoch": 0.2077358629310979,
+      "grad_norm": 1.2404136657714844,
+      "learning_rate": 0.00018389752603104018,
+      "loss": 0.5791,
+      "step": 2825
+    },
+    {
+      "epoch": 0.20780939774983453,
+      "grad_norm": 1.3346428871154785,
+      "learning_rate": 0.00018388456361009419,
+      "loss": 0.5772,
+      "step": 2826
+    },
+    {
+      "epoch": 0.2078829325685712,
+      "grad_norm": 1.0855741500854492,
+      "learning_rate": 0.00018387159643112723,
+      "loss": 0.5892,
+      "step": 2827
+    },
+    {
+      "epoch": 0.2079564673873079,
+      "grad_norm": 1.3909128904342651,
+      "learning_rate": 0.00018385862449487477,
+      "loss": 0.5963,
+      "step": 2828
+    },
+    {
+      "epoch": 0.20803000220604456,
+      "grad_norm": 1.1573638916015625,
+      "learning_rate": 0.00018384564780207262,
+      "loss": 0.5619,
+      "step": 2829
+    },
+    {
+      "epoch": 0.20810353702478124,
+      "grad_norm": 1.6951537132263184,
+      "learning_rate": 0.00018383266635345688,
+      "loss": 0.5727,
+      "step": 2830
+    },
+    {
+      "epoch": 0.2081770718435179,
+      "grad_norm": 1.6182575225830078,
+      "learning_rate": 0.00018381968014976382,
+      "loss": 0.5638,
+      "step": 2831
+    },
+    {
+      "epoch": 0.2082506066622546,
+      "grad_norm": 1.3190699815750122,
+      "learning_rate": 0.00018380668919173006,
+      "loss": 0.5514,
+      "step": 2832
+    },
+    {
+      "epoch": 0.20832414148099124,
+      "grad_norm": 1.1180505752563477,
+      "learning_rate": 0.0001837936934800924,
+      "loss": 0.5563,
+      "step": 2833
+    },
+    {
+      "epoch": 0.2083976762997279,
+      "grad_norm": 1.4441022872924805,
+      "learning_rate": 0.00018378069301558804,
+      "loss": 0.5496,
+      "step": 2834
+    },
+    {
+      "epoch": 0.2084712111184646,
+      "grad_norm": 1.1558384895324707,
+      "learning_rate": 0.00018376768779895434,
+      "loss": 0.567,
+      "step": 2835
+    },
+    {
+      "epoch": 0.20854474593720126,
+      "grad_norm": 1.7985193729400635,
+      "learning_rate": 0.000183754677830929,
+      "loss": 0.5633,
+      "step": 2836
+    },
+    {
+      "epoch": 0.20861828075593794,
+      "grad_norm": 1.400950312614441,
+      "learning_rate": 0.00018374166311224997,
+      "loss": 0.6025,
+      "step": 2837
+    },
+    {
+      "epoch": 0.20869181557467462,
+      "grad_norm": 1.4239730834960938,
+      "learning_rate": 0.0001837286436436554,
+      "loss": 0.5642,
+      "step": 2838
+    },
+    {
+      "epoch": 0.2087653503934113,
+      "grad_norm": 1.086327075958252,
+      "learning_rate": 0.00018371561942588383,
+      "loss": 0.5572,
+      "step": 2839
+    },
+    {
+      "epoch": 0.20883888521214794,
+      "grad_norm": 1.623172640800476,
+      "learning_rate": 0.00018370259045967394,
+      "loss": 0.5722,
+      "step": 2840
+    },
+    {
+      "epoch": 0.20891242003088462,
+      "grad_norm": 1.2297663688659668,
+      "learning_rate": 0.0001836895567457648,
+      "loss": 0.5687,
+      "step": 2841
+    },
+    {
+      "epoch": 0.2089859548496213,
+      "grad_norm": 1.7174121141433716,
+      "learning_rate": 0.00018367651828489568,
+      "loss": 0.5915,
+      "step": 2842
+    },
+    {
+      "epoch": 0.20905948966835797,
+      "grad_norm": 1.2633963823318481,
+      "learning_rate": 0.00018366347507780612,
+      "loss": 0.5461,
+      "step": 2843
+    },
+    {
+      "epoch": 0.20913302448709464,
+      "grad_norm": 1.5371973514556885,
+      "learning_rate": 0.000183650427125236,
+      "loss": 0.6063,
+      "step": 2844
+    },
+    {
+      "epoch": 0.20920655930583132,
+      "grad_norm": 1.3906015157699585,
+      "learning_rate": 0.00018363737442792533,
+      "loss": 0.5335,
+      "step": 2845
+    },
+    {
+      "epoch": 0.209280094124568,
+      "grad_norm": 1.545803189277649,
+      "learning_rate": 0.00018362431698661452,
+      "loss": 0.5866,
+      "step": 2846
+    },
+    {
+      "epoch": 0.20935362894330464,
+      "grad_norm": 1.5055469274520874,
+      "learning_rate": 0.0001836112548020442,
+      "loss": 0.5537,
+      "step": 2847
+    },
+    {
+      "epoch": 0.20942716376204132,
+      "grad_norm": 1.2694060802459717,
+      "learning_rate": 0.0001835981878749553,
+      "loss": 0.55,
+      "step": 2848
+    },
+    {
+      "epoch": 0.209500698580778,
+      "grad_norm": 1.241575002670288,
+      "learning_rate": 0.00018358511620608894,
+      "loss": 0.5654,
+      "step": 2849
+    },
+    {
+      "epoch": 0.20957423339951467,
+      "grad_norm": 1.4436458349227905,
+      "learning_rate": 0.00018357203979618657,
+      "loss": 0.5802,
+      "step": 2850
+    },
+    {
+      "epoch": 0.20964776821825135,
+      "grad_norm": 1.096796989440918,
+      "learning_rate": 0.0001835589586459899,
+      "loss": 0.5886,
+      "step": 2851
+    },
+    {
+      "epoch": 0.20972130303698802,
+      "grad_norm": 1.7189993858337402,
+      "learning_rate": 0.00018354587275624097,
+      "loss": 0.5379,
+      "step": 2852
+    },
+    {
+      "epoch": 0.2097948378557247,
+      "grad_norm": 1.3353079557418823,
+      "learning_rate": 0.0001835327821276819,
+      "loss": 0.573,
+      "step": 2853
+    },
+    {
+      "epoch": 0.20986837267446135,
+      "grad_norm": 1.4592704772949219,
+      "learning_rate": 0.00018351968676105532,
+      "loss": 0.5635,
+      "step": 2854
+    },
+    {
+      "epoch": 0.20994190749319802,
+      "grad_norm": 1.0787458419799805,
+      "learning_rate": 0.00018350658665710392,
+      "loss": 0.5718,
+      "step": 2855
+    },
+    {
+      "epoch": 0.2100154423119347,
+      "grad_norm": 1.2916628122329712,
+      "learning_rate": 0.0001834934818165708,
+      "loss": 0.5624,
+      "step": 2856
+    },
+    {
+      "epoch": 0.21008897713067137,
+      "grad_norm": 0.9856218695640564,
+      "learning_rate": 0.0001834803722401993,
+      "loss": 0.563,
+      "step": 2857
+    },
+    {
+      "epoch": 0.21016251194940805,
+      "grad_norm": 1.7601271867752075,
+      "learning_rate": 0.000183467257928733,
+      "loss": 0.5563,
+      "step": 2858
+    },
+    {
+      "epoch": 0.21023604676814472,
+      "grad_norm": 1.4991081953048706,
+      "learning_rate": 0.00018345413888291572,
+      "loss": 0.5471,
+      "step": 2859
+    },
+    {
+      "epoch": 0.2103095815868814,
+      "grad_norm": 1.3091179132461548,
+      "learning_rate": 0.0001834410151034916,
+      "loss": 0.6059,
+      "step": 2860
+    },
+    {
+      "epoch": 0.21038311640561805,
+      "grad_norm": 1.0682049989700317,
+      "learning_rate": 0.00018342788659120508,
+      "loss": 0.5639,
+      "step": 2861
+    },
+    {
+      "epoch": 0.21045665122435472,
+      "grad_norm": 1.3862104415893555,
+      "learning_rate": 0.00018341475334680076,
+      "loss": 0.5727,
+      "step": 2862
+    },
+    {
+      "epoch": 0.2105301860430914,
+      "grad_norm": 1.110353708267212,
+      "learning_rate": 0.00018340161537102362,
+      "loss": 0.5645,
+      "step": 2863
+    },
+    {
+      "epoch": 0.21060372086182808,
+      "grad_norm": 1.5367006063461304,
+      "learning_rate": 0.0001833884726646188,
+      "loss": 0.5844,
+      "step": 2864
+    },
+    {
+      "epoch": 0.21067725568056475,
+      "grad_norm": 1.2548023462295532,
+      "learning_rate": 0.00018337532522833184,
+      "loss": 0.5585,
+      "step": 2865
+    },
+    {
+      "epoch": 0.21075079049930143,
+      "grad_norm": 1.3144752979278564,
+      "learning_rate": 0.00018336217306290844,
+      "loss": 0.5469,
+      "step": 2866
+    },
+    {
+      "epoch": 0.2108243253180381,
+      "grad_norm": 1.0409570932388306,
+      "learning_rate": 0.0001833490161690946,
+      "loss": 0.5483,
+      "step": 2867
+    },
+    {
+      "epoch": 0.21089786013677475,
+      "grad_norm": 1.5468132495880127,
+      "learning_rate": 0.0001833358545476366,
+      "loss": 0.545,
+      "step": 2868
+    },
+    {
+      "epoch": 0.21097139495551143,
+      "grad_norm": 1.0659055709838867,
+      "learning_rate": 0.000183322688199281,
+      "loss": 0.5486,
+      "step": 2869
+    },
+    {
+      "epoch": 0.2110449297742481,
+      "grad_norm": 1.5632880926132202,
+      "learning_rate": 0.00018330951712477458,
+      "loss": 0.6024,
+      "step": 2870
+    },
+    {
+      "epoch": 0.21111846459298478,
+      "grad_norm": 1.0727630853652954,
+      "learning_rate": 0.00018329634132486442,
+      "loss": 0.5597,
+      "step": 2871
+    },
+    {
+      "epoch": 0.21119199941172145,
+      "grad_norm": 1.7708021402359009,
+      "learning_rate": 0.00018328316080029787,
+      "loss": 0.5656,
+      "step": 2872
+    },
+    {
+      "epoch": 0.21126553423045813,
+      "grad_norm": 1.2210580110549927,
+      "learning_rate": 0.00018326997555182258,
+      "loss": 0.5551,
+      "step": 2873
+    },
+    {
+      "epoch": 0.2113390690491948,
+      "grad_norm": 1.7090801000595093,
+      "learning_rate": 0.00018325678558018637,
+      "loss": 0.5426,
+      "step": 2874
+    },
+    {
+      "epoch": 0.21141260386793145,
+      "grad_norm": 1.2940007448196411,
+      "learning_rate": 0.00018324359088613742,
+      "loss": 0.5578,
+      "step": 2875
+    },
+    {
+      "epoch": 0.21148613868666813,
+      "grad_norm": 1.4750893115997314,
+      "learning_rate": 0.00018323039147042416,
+      "loss": 0.5848,
+      "step": 2876
+    },
+    {
+      "epoch": 0.2115596735054048,
+      "grad_norm": 1.381226897239685,
+      "learning_rate": 0.00018321718733379524,
+      "loss": 0.566,
+      "step": 2877
+    },
+    {
+      "epoch": 0.21163320832414148,
+      "grad_norm": 1.355833649635315,
+      "learning_rate": 0.00018320397847699965,
+      "loss": 0.5556,
+      "step": 2878
+    },
+    {
+      "epoch": 0.21170674314287816,
+      "grad_norm": 1.1924527883529663,
+      "learning_rate": 0.00018319076490078658,
+      "loss": 0.581,
+      "step": 2879
+    },
+    {
+      "epoch": 0.21178027796161483,
+      "grad_norm": 1.3908332586288452,
+      "learning_rate": 0.00018317754660590553,
+      "loss": 0.5379,
+      "step": 2880
+    },
+    {
+      "epoch": 0.2118538127803515,
+      "grad_norm": 1.1826207637786865,
+      "learning_rate": 0.00018316432359310626,
+      "loss": 0.5799,
+      "step": 2881
+    },
+    {
+      "epoch": 0.21192734759908816,
+      "grad_norm": 1.523348331451416,
+      "learning_rate": 0.0001831510958631388,
+      "loss": 0.5621,
+      "step": 2882
+    },
+    {
+      "epoch": 0.21200088241782483,
+      "grad_norm": 1.3590301275253296,
+      "learning_rate": 0.0001831378634167534,
+      "loss": 0.5437,
+      "step": 2883
+    },
+    {
+      "epoch": 0.2120744172365615,
+      "grad_norm": 1.1726497411727905,
+      "learning_rate": 0.00018312462625470066,
+      "loss": 0.5438,
+      "step": 2884
+    },
+    {
+      "epoch": 0.21214795205529818,
+      "grad_norm": 1.2389098405838013,
+      "learning_rate": 0.0001831113843777314,
+      "loss": 0.5561,
+      "step": 2885
+    },
+    {
+      "epoch": 0.21222148687403486,
+      "grad_norm": 1.1923811435699463,
+      "learning_rate": 0.0001830981377865967,
+      "loss": 0.5565,
+      "step": 2886
+    },
+    {
+      "epoch": 0.21229502169277154,
+      "grad_norm": 1.1164624691009521,
+      "learning_rate": 0.00018308488648204792,
+      "loss": 0.597,
+      "step": 2887
+    },
+    {
+      "epoch": 0.2123685565115082,
+      "grad_norm": 1.640863060951233,
+      "learning_rate": 0.00018307163046483673,
+      "loss": 0.5698,
+      "step": 2888
+    },
+    {
+      "epoch": 0.21244209133024486,
+      "grad_norm": 1.6564183235168457,
+      "learning_rate": 0.00018305836973571497,
+      "loss": 0.5531,
+      "step": 2889
+    },
+    {
+      "epoch": 0.21251562614898153,
+      "grad_norm": 1.3389441967010498,
+      "learning_rate": 0.00018304510429543482,
+      "loss": 0.5669,
+      "step": 2890
+    },
+    {
+      "epoch": 0.2125891609677182,
+      "grad_norm": 1.4294819831848145,
+      "learning_rate": 0.00018303183414474875,
+      "loss": 0.575,
+      "step": 2891
+    },
+    {
+      "epoch": 0.2126626957864549,
+      "grad_norm": 1.3099455833435059,
+      "learning_rate": 0.00018301855928440935,
+      "loss": 0.5838,
+      "step": 2892
+    },
+    {
+      "epoch": 0.21273623060519156,
+      "grad_norm": 1.1975899934768677,
+      "learning_rate": 0.0001830052797151697,
+      "loss": 0.5672,
+      "step": 2893
+    },
+    {
+      "epoch": 0.21280976542392824,
+      "grad_norm": 1.5240440368652344,
+      "learning_rate": 0.00018299199543778298,
+      "loss": 0.5741,
+      "step": 2894
+    },
+    {
+      "epoch": 0.2128833002426649,
+      "grad_norm": 1.4865225553512573,
+      "learning_rate": 0.0001829787064530027,
+      "loss": 0.5273,
+      "step": 2895
+    },
+    {
+      "epoch": 0.21295683506140156,
+      "grad_norm": 1.3831321001052856,
+      "learning_rate": 0.0001829654127615826,
+      "loss": 0.5466,
+      "step": 2896
+    },
+    {
+      "epoch": 0.21303036988013824,
+      "grad_norm": 1.4533921480178833,
+      "learning_rate": 0.00018295211436427676,
+      "loss": 0.561,
+      "step": 2897
+    },
+    {
+      "epoch": 0.2131039046988749,
+      "grad_norm": 1.262459635734558,
+      "learning_rate": 0.0001829388112618394,
+      "loss": 0.5328,
+      "step": 2898
+    },
+    {
+      "epoch": 0.2131774395176116,
+      "grad_norm": 1.1636648178100586,
+      "learning_rate": 0.0001829255034550252,
+      "loss": 0.5577,
+      "step": 2899
+    },
+    {
+      "epoch": 0.21325097433634826,
+      "grad_norm": 1.3677254915237427,
+      "learning_rate": 0.0001829121909445889,
+      "loss": 0.5311,
+      "step": 2900
+    },
+    {
+      "epoch": 0.21332450915508494,
+      "grad_norm": 1.2248530387878418,
+      "learning_rate": 0.00018289887373128561,
+      "loss": 0.5421,
+      "step": 2901
+    },
+    {
+      "epoch": 0.21339804397382162,
+      "grad_norm": 1.571791648864746,
+      "learning_rate": 0.00018288555181587076,
+      "loss": 0.5637,
+      "step": 2902
+    },
+    {
+      "epoch": 0.21347157879255826,
+      "grad_norm": 1.4395170211791992,
+      "learning_rate": 0.00018287222519909987,
+      "loss": 0.5448,
+      "step": 2903
+    },
+    {
+      "epoch": 0.21354511361129494,
+      "grad_norm": 1.4251290559768677,
+      "learning_rate": 0.00018285889388172896,
+      "loss": 0.5335,
+      "step": 2904
+    },
+    {
+      "epoch": 0.21361864843003162,
+      "grad_norm": 1.3303669691085815,
+      "learning_rate": 0.00018284555786451416,
+      "loss": 0.5319,
+      "step": 2905
+    },
+    {
+      "epoch": 0.2136921832487683,
+      "grad_norm": 1.2890651226043701,
+      "learning_rate": 0.00018283221714821183,
+      "loss": 0.5655,
+      "step": 2906
+    },
+    {
+      "epoch": 0.21376571806750497,
+      "grad_norm": 1.1711236238479614,
+      "learning_rate": 0.00018281887173357878,
+      "loss": 0.5351,
+      "step": 2907
+    },
+    {
+      "epoch": 0.21383925288624164,
+      "grad_norm": 1.2209774255752563,
+      "learning_rate": 0.0001828055216213719,
+      "loss": 0.5341,
+      "step": 2908
+    },
+    {
+      "epoch": 0.21391278770497832,
+      "grad_norm": 0.9704241156578064,
+      "learning_rate": 0.00018279216681234845,
+      "loss": 0.5487,
+      "step": 2909
+    },
+    {
+      "epoch": 0.21398632252371497,
+      "grad_norm": 1.5567630529403687,
+      "learning_rate": 0.0001827788073072659,
+      "loss": 0.5574,
+      "step": 2910
+    },
+    {
+      "epoch": 0.21405985734245164,
+      "grad_norm": 1.018020749092102,
+      "learning_rate": 0.00018276544310688207,
+      "loss": 0.5296,
+      "step": 2911
+    },
+    {
+      "epoch": 0.21413339216118832,
+      "grad_norm": 1.5725622177124023,
+      "learning_rate": 0.00018275207421195494,
+      "loss": 0.5559,
+      "step": 2912
+    },
+    {
+      "epoch": 0.214206926979925,
+      "grad_norm": 1.147858738899231,
+      "learning_rate": 0.00018273870062324285,
+      "loss": 0.5177,
+      "step": 2913
+    },
+    {
+      "epoch": 0.21428046179866167,
+      "grad_norm": 1.4399924278259277,
+      "learning_rate": 0.00018272532234150436,
+      "loss": 0.5331,
+      "step": 2914
+    },
+    {
+      "epoch": 0.21435399661739835,
+      "grad_norm": 1.4297422170639038,
+      "learning_rate": 0.00018271193936749824,
+      "loss": 0.5292,
+      "step": 2915
+    },
+    {
+      "epoch": 0.21442753143613502,
+      "grad_norm": 1.5445528030395508,
+      "learning_rate": 0.00018269855170198365,
+      "loss": 0.5531,
+      "step": 2916
+    },
+    {
+      "epoch": 0.21450106625487167,
+      "grad_norm": 1.2804207801818848,
+      "learning_rate": 0.00018268515934571992,
+      "loss": 0.5809,
+      "step": 2917
+    },
+    {
+      "epoch": 0.21457460107360835,
+      "grad_norm": 1.566281795501709,
+      "learning_rate": 0.0001826717622994667,
+      "loss": 0.5383,
+      "step": 2918
+    },
+    {
+      "epoch": 0.21464813589234502,
+      "grad_norm": 1.2450507879257202,
+      "learning_rate": 0.00018265836056398387,
+      "loss": 0.5473,
+      "step": 2919
+    },
+    {
+      "epoch": 0.2147216707110817,
+      "grad_norm": 1.4447340965270996,
+      "learning_rate": 0.00018264495414003159,
+      "loss": 0.5341,
+      "step": 2920
+    },
+    {
+      "epoch": 0.21479520552981837,
+      "grad_norm": 1.3136258125305176,
+      "learning_rate": 0.0001826315430283703,
+      "loss": 0.56,
+      "step": 2921
+    },
+    {
+      "epoch": 0.21486874034855505,
+      "grad_norm": 1.2370400428771973,
+      "learning_rate": 0.0001826181272297607,
+      "loss": 0.5884,
+      "step": 2922
+    },
+    {
+      "epoch": 0.21494227516729172,
+      "grad_norm": 1.126532793045044,
+      "learning_rate": 0.00018260470674496374,
+      "loss": 0.5967,
+      "step": 2923
+    },
+    {
+      "epoch": 0.21501580998602837,
+      "grad_norm": 1.6091411113739014,
+      "learning_rate": 0.00018259128157474062,
+      "loss": 0.5381,
+      "step": 2924
+    },
+    {
+      "epoch": 0.21508934480476505,
+      "grad_norm": 1.3106461763381958,
+      "learning_rate": 0.00018257785171985283,
+      "loss": 0.5445,
+      "step": 2925
+    },
+    {
+      "epoch": 0.21516287962350172,
+      "grad_norm": 1.425999402999878,
+      "learning_rate": 0.00018256441718106215,
+      "loss": 0.5793,
+      "step": 2926
+    },
+    {
+      "epoch": 0.2152364144422384,
+      "grad_norm": 1.414546251296997,
+      "learning_rate": 0.0001825509779591306,
+      "loss": 0.5406,
+      "step": 2927
+    },
+    {
+      "epoch": 0.21530994926097508,
+      "grad_norm": 1.0518776178359985,
+      "learning_rate": 0.00018253753405482047,
+      "loss": 0.527,
+      "step": 2928
+    },
+    {
+      "epoch": 0.21538348407971175,
+      "grad_norm": 0.9168993830680847,
+      "learning_rate": 0.00018252408546889432,
+      "loss": 0.5309,
+      "step": 2929
+    },
+    {
+      "epoch": 0.21545701889844843,
+      "grad_norm": 1.2502708435058594,
+      "learning_rate": 0.00018251063220211492,
+      "loss": 0.5312,
+      "step": 2930
+    },
+    {
+      "epoch": 0.21553055371718507,
+      "grad_norm": 0.8570818901062012,
+      "learning_rate": 0.0001824971742552454,
+      "loss": 0.5698,
+      "step": 2931
+    },
+    {
+      "epoch": 0.21560408853592175,
+      "grad_norm": 1.3225476741790771,
+      "learning_rate": 0.00018248371162904906,
+      "loss": 0.5437,
+      "step": 2932
+    },
+    {
+      "epoch": 0.21567762335465843,
+      "grad_norm": 0.7682308554649353,
+      "learning_rate": 0.0001824702443242896,
+      "loss": 0.545,
+      "step": 2933
+    },
+    {
+      "epoch": 0.2157511581733951,
+      "grad_norm": 1.2084381580352783,
+      "learning_rate": 0.0001824567723417308,
+      "loss": 0.5556,
+      "step": 2934
+    },
+    {
+      "epoch": 0.21582469299213178,
+      "grad_norm": 0.9033636450767517,
+      "learning_rate": 0.00018244329568213688,
+      "loss": 0.5484,
+      "step": 2935
+    },
+    {
+      "epoch": 0.21589822781086845,
+      "grad_norm": 1.238548755645752,
+      "learning_rate": 0.0001824298143462722,
+      "loss": 0.5365,
+      "step": 2936
+    },
+    {
+      "epoch": 0.21597176262960513,
+      "grad_norm": 0.8866881132125854,
+      "learning_rate": 0.0001824163283349015,
+      "loss": 0.542,
+      "step": 2937
+    },
+    {
+      "epoch": 0.21604529744834178,
+      "grad_norm": 0.9582094550132751,
+      "learning_rate": 0.00018240283764878965,
+      "loss": 0.5218,
+      "step": 2938
+    },
+    {
+      "epoch": 0.21611883226707845,
+      "grad_norm": 0.8529465198516846,
+      "learning_rate": 0.0001823893422887019,
+      "loss": 0.5322,
+      "step": 2939
+    },
+    {
+      "epoch": 0.21619236708581513,
+      "grad_norm": 0.8557735681533813,
+      "learning_rate": 0.0001823758422554037,
+      "loss": 0.5009,
+      "step": 2940
+    },
+    {
+      "epoch": 0.2162659019045518,
+      "grad_norm": 0.7988099455833435,
+      "learning_rate": 0.00018236233754966076,
+      "loss": 0.5834,
+      "step": 2941
+    },
+    {
+      "epoch": 0.21633943672328848,
+      "grad_norm": 1.0773651599884033,
+      "learning_rate": 0.00018234882817223916,
+      "loss": 0.534,
+      "step": 2942
+    },
+    {
+      "epoch": 0.21641297154202516,
+      "grad_norm": 0.6712372899055481,
+      "learning_rate": 0.0001823353141239051,
+      "loss": 0.5255,
+      "step": 2943
+    },
+    {
+      "epoch": 0.21648650636076183,
+      "grad_norm": 0.8130234479904175,
+      "learning_rate": 0.00018232179540542514,
+      "loss": 0.5191,
+      "step": 2944
+    },
+    {
+      "epoch": 0.21656004117949848,
+      "grad_norm": 0.9147355556488037,
+      "learning_rate": 0.00018230827201756603,
+      "loss": 0.5147,
+      "step": 2945
+    },
+    {
+      "epoch": 0.21663357599823516,
+      "grad_norm": 0.7353776097297668,
+      "learning_rate": 0.0001822947439610949,
+      "loss": 0.5292,
+      "step": 2946
+    },
+    {
+      "epoch": 0.21670711081697183,
+      "grad_norm": 0.7399029731750488,
+      "learning_rate": 0.00018228121123677906,
+      "loss": 0.5203,
+      "step": 2947
+    },
+    {
+      "epoch": 0.2167806456357085,
+      "grad_norm": 0.7483555674552917,
+      "learning_rate": 0.00018226767384538607,
+      "loss": 0.549,
+      "step": 2948
+    },
+    {
+      "epoch": 0.21685418045444518,
+      "grad_norm": 0.5925317406654358,
+      "learning_rate": 0.00018225413178768381,
+      "loss": 0.5383,
+      "step": 2949
+    },
+    {
+      "epoch": 0.21692771527318186,
+      "grad_norm": 0.7870122194290161,
+      "learning_rate": 0.00018224058506444035,
+      "loss": 0.5122,
+      "step": 2950
+    },
+    {
+      "epoch": 0.21700125009191853,
+      "grad_norm": 0.6600607633590698,
+      "learning_rate": 0.00018222703367642415,
+      "loss": 0.533,
+      "step": 2951
+    },
+    {
+      "epoch": 0.21707478491065518,
+      "grad_norm": 0.6445098519325256,
+      "learning_rate": 0.0001822134776244038,
+      "loss": 0.4951,
+      "step": 2952
+    },
+    {
+      "epoch": 0.21714831972939186,
+      "grad_norm": 0.6419938802719116,
+      "learning_rate": 0.00018219991690914827,
+      "loss": 0.5008,
+      "step": 2953
+    },
+    {
+      "epoch": 0.21722185454812853,
+      "grad_norm": 0.6079204678535461,
+      "learning_rate": 0.00018218635153142664,
+      "loss": 0.5336,
+      "step": 2954
+    },
+    {
+      "epoch": 0.2172953893668652,
+      "grad_norm": 0.5952870845794678,
+      "learning_rate": 0.0001821727814920085,
+      "loss": 0.5434,
+      "step": 2955
+    },
+    {
+      "epoch": 0.21736892418560189,
+      "grad_norm": 0.5915732979774475,
+      "learning_rate": 0.0001821592067916634,
+      "loss": 0.5408,
+      "step": 2956
+    },
+    {
+      "epoch": 0.21744245900433856,
+      "grad_norm": 0.6558079719543457,
+      "learning_rate": 0.00018214562743116143,
+      "loss": 0.5496,
+      "step": 2957
+    },
+    {
+      "epoch": 0.21751599382307524,
+      "grad_norm": 0.5513221025466919,
+      "learning_rate": 0.00018213204341127276,
+      "loss": 0.522,
+      "step": 2958
+    },
+    {
+      "epoch": 0.21758952864181189,
+      "grad_norm": 0.7003249526023865,
+      "learning_rate": 0.0001821184547327679,
+      "loss": 0.5478,
+      "step": 2959
+    },
+    {
+      "epoch": 0.21766306346054856,
+      "grad_norm": 0.5708374381065369,
+      "learning_rate": 0.00018210486139641762,
+      "loss": 0.5105,
+      "step": 2960
+    },
+    {
+      "epoch": 0.21773659827928524,
+      "grad_norm": 0.5709456205368042,
+      "learning_rate": 0.000182091263402993,
+      "loss": 0.5219,
+      "step": 2961
+    },
+    {
+      "epoch": 0.2178101330980219,
+      "grad_norm": 0.5473268032073975,
+      "learning_rate": 0.00018207766075326525,
+      "loss": 0.5169,
+      "step": 2962
+    },
+    {
+      "epoch": 0.2178836679167586,
+      "grad_norm": 0.6497728824615479,
+      "learning_rate": 0.00018206405344800596,
+      "loss": 0.5476,
+      "step": 2963
+    },
+    {
+      "epoch": 0.21795720273549526,
+      "grad_norm": 0.5642459392547607,
+      "learning_rate": 0.000182050441487987,
+      "loss": 0.5525,
+      "step": 2964
+    },
+    {
+      "epoch": 0.21803073755423194,
+      "grad_norm": 0.6211819052696228,
+      "learning_rate": 0.00018203682487398032,
+      "loss": 0.5421,
+      "step": 2965
+    },
+    {
+      "epoch": 0.2181042723729686,
+      "grad_norm": 0.8039196133613586,
+      "learning_rate": 0.00018202320360675842,
+      "loss": 0.5566,
+      "step": 2966
+    },
+    {
+      "epoch": 0.21817780719170526,
+      "grad_norm": 0.547728419303894,
+      "learning_rate": 0.00018200957768709384,
+      "loss": 0.544,
+      "step": 2967
+    },
+    {
+      "epoch": 0.21825134201044194,
+      "grad_norm": 1.152121663093567,
+      "learning_rate": 0.00018199594711575947,
+      "loss": 0.5464,
+      "step": 2968
+    },
+    {
+      "epoch": 0.21832487682917862,
+      "grad_norm": 1.065087080001831,
+      "learning_rate": 0.00018198231189352844,
+      "loss": 0.5434,
+      "step": 2969
+    },
+    {
+      "epoch": 0.2183984116479153,
+      "grad_norm": 0.6706288456916809,
+      "learning_rate": 0.00018196867202117413,
+      "loss": 0.5188,
+      "step": 2970
+    },
+    {
+      "epoch": 0.21847194646665197,
+      "grad_norm": 0.9068410396575928,
+      "learning_rate": 0.00018195502749947026,
+      "loss": 0.5675,
+      "step": 2971
+    },
+    {
+      "epoch": 0.21854548128538864,
+      "grad_norm": 1.509711503982544,
+      "learning_rate": 0.00018194137832919073,
+      "loss": 0.5126,
+      "step": 2972
+    },
+    {
+      "epoch": 0.2186190161041253,
+      "grad_norm": 0.7947085499763489,
+      "learning_rate": 0.00018192772451110977,
+      "loss": 0.5522,
+      "step": 2973
+    },
+    {
+      "epoch": 0.21869255092286197,
+      "grad_norm": 2.1711583137512207,
+      "learning_rate": 0.00018191406604600182,
+      "loss": 0.549,
+      "step": 2974
+    },
+    {
+      "epoch": 0.21876608574159864,
+      "grad_norm": 1.6515501737594604,
+      "learning_rate": 0.00018190040293464157,
+      "loss": 0.5238,
+      "step": 2975
+    },
+    {
+      "epoch": 0.21883962056033532,
+      "grad_norm": 2.079498291015625,
+      "learning_rate": 0.00018188673517780405,
+      "loss": 0.5468,
+      "step": 2976
+    },
+    {
+      "epoch": 0.218913155379072,
+      "grad_norm": 2.4215359687805176,
+      "learning_rate": 0.00018187306277626445,
+      "loss": 0.5687,
+      "step": 2977
+    },
+    {
+      "epoch": 0.21898669019780867,
+      "grad_norm": 0.6824780702590942,
+      "learning_rate": 0.00018185938573079838,
+      "loss": 0.529,
+      "step": 2978
+    },
+    {
+      "epoch": 0.21906022501654535,
+      "grad_norm": 2.2429933547973633,
+      "learning_rate": 0.00018184570404218155,
+      "loss": 0.5466,
+      "step": 2979
+    },
+    {
+      "epoch": 0.219133759835282,
+      "grad_norm": 1.2417845726013184,
+      "learning_rate": 0.00018183201771118998,
+      "loss": 0.568,
+      "step": 2980
+    },
+    {
+      "epoch": 0.21920729465401867,
+      "grad_norm": 2.026502847671509,
+      "learning_rate": 0.00018181832673860003,
+      "loss": 0.5493,
+      "step": 2981
+    },
+    {
+      "epoch": 0.21928082947275535,
+      "grad_norm": 2.0270910263061523,
+      "learning_rate": 0.00018180463112518825,
+      "loss": 0.5452,
+      "step": 2982
+    },
+    {
+      "epoch": 0.21935436429149202,
+      "grad_norm": 0.9847967028617859,
+      "learning_rate": 0.00018179093087173146,
+      "loss": 0.5451,
+      "step": 2983
+    },
+    {
+      "epoch": 0.2194278991102287,
+      "grad_norm": 1.9717069864273071,
+      "learning_rate": 0.00018177722597900674,
+      "loss": 0.5251,
+      "step": 2984
+    },
+    {
+      "epoch": 0.21950143392896537,
+      "grad_norm": 1.5626556873321533,
+      "learning_rate": 0.00018176351644779145,
+      "loss": 0.5249,
+      "step": 2985
+    },
+    {
+      "epoch": 0.21957496874770205,
+      "grad_norm": 1.2404801845550537,
+      "learning_rate": 0.00018174980227886325,
+      "loss": 0.4994,
+      "step": 2986
+    },
+    {
+      "epoch": 0.2196485035664387,
+      "grad_norm": 1.4244028329849243,
+      "learning_rate": 0.00018173608347299997,
+      "loss": 0.5327,
+      "step": 2987
+    },
+    {
+      "epoch": 0.21972203838517537,
+      "grad_norm": 1.0698027610778809,
+      "learning_rate": 0.00018172236003097977,
+      "loss": 0.5669,
+      "step": 2988
+    },
+    {
+      "epoch": 0.21979557320391205,
+      "grad_norm": 1.3782232999801636,
+      "learning_rate": 0.00018170863195358106,
+      "loss": 0.5259,
+      "step": 2989
+    },
+    {
+      "epoch": 0.21986910802264872,
+      "grad_norm": 1.0606563091278076,
+      "learning_rate": 0.00018169489924158252,
+      "loss": 0.5511,
+      "step": 2990
+    },
+    {
+      "epoch": 0.2199426428413854,
+      "grad_norm": 1.502944827079773,
+      "learning_rate": 0.0001816811618957631,
+      "loss": 0.5215,
+      "step": 2991
+    },
+    {
+      "epoch": 0.22001617766012208,
+      "grad_norm": 1.4065078496932983,
+      "learning_rate": 0.00018166741991690194,
+      "loss": 0.5413,
+      "step": 2992
+    },
+    {
+      "epoch": 0.22008971247885875,
+      "grad_norm": 1.8072099685668945,
+      "learning_rate": 0.00018165367330577858,
+      "loss": 0.5187,
+      "step": 2993
+    },
+    {
+      "epoch": 0.2201632472975954,
+      "grad_norm": 1.496889352798462,
+      "learning_rate": 0.00018163992206317265,
+      "loss": 0.5293,
+      "step": 2994
+    },
+    {
+      "epoch": 0.22023678211633207,
+      "grad_norm": 1.2236113548278809,
+      "learning_rate": 0.00018162616618986422,
+      "loss": 0.5186,
+      "step": 2995
+    },
+    {
+      "epoch": 0.22031031693506875,
+      "grad_norm": 1.0579713582992554,
+      "learning_rate": 0.00018161240568663348,
+      "loss": 0.5347,
+      "step": 2996
+    },
+    {
+      "epoch": 0.22038385175380543,
+      "grad_norm": 1.6502026319503784,
+      "learning_rate": 0.00018159864055426097,
+      "loss": 0.509,
+      "step": 2997
+    },
+    {
+      "epoch": 0.2204573865725421,
+      "grad_norm": 1.2475897073745728,
+      "learning_rate": 0.00018158487079352744,
+      "loss": 0.522,
+      "step": 2998
+    },
+    {
+      "epoch": 0.22053092139127878,
+      "grad_norm": 1.6342942714691162,
+      "learning_rate": 0.00018157109640521397,
+      "loss": 0.5369,
+      "step": 2999
+    },
+    {
+      "epoch": 0.22060445621001545,
+      "grad_norm": 1.6629966497421265,
+      "learning_rate": 0.00018155731739010177,
+      "loss": 0.5096,
+      "step": 3000
+    },
+    {
+      "epoch": 0.2206779910287521,
+      "grad_norm": 0.9001145958900452,
+      "learning_rate": 0.0001815435337489725,
+      "loss": 0.5366,
+      "step": 3001
+    },
+    {
+      "epoch": 0.22075152584748878,
+      "grad_norm": 1.352590799331665,
+      "learning_rate": 0.00018152974548260795,
+      "loss": 0.536,
+      "step": 3002
+    },
+    {
+      "epoch": 0.22082506066622545,
+      "grad_norm": 1.0985625982284546,
+      "learning_rate": 0.00018151595259179018,
+      "loss": 0.5243,
+      "step": 3003
+    },
+    {
+      "epoch": 0.22089859548496213,
+      "grad_norm": 0.9768959283828735,
+      "learning_rate": 0.00018150215507730156,
+      "loss": 0.5302,
+      "step": 3004
+    },
+    {
+      "epoch": 0.2209721303036988,
+      "grad_norm": 1.3413007259368896,
+      "learning_rate": 0.00018148835293992468,
+      "loss": 0.5253,
+      "step": 3005
+    },
+    {
+      "epoch": 0.22104566512243548,
+      "grad_norm": 1.123515009880066,
+      "learning_rate": 0.0001814745461804424,
+      "loss": 0.5408,
+      "step": 3006
+    },
+    {
+      "epoch": 0.22111919994117216,
+      "grad_norm": 1.508075475692749,
+      "learning_rate": 0.0001814607347996379,
+      "loss": 0.5121,
+      "step": 3007
+    },
+    {
+      "epoch": 0.2211927347599088,
+      "grad_norm": 1.2609809637069702,
+      "learning_rate": 0.00018144691879829456,
+      "loss": 0.5424,
+      "step": 3008
+    },
+    {
+      "epoch": 0.22126626957864548,
+      "grad_norm": 1.2567862272262573,
+      "learning_rate": 0.00018143309817719602,
+      "loss": 0.502,
+      "step": 3009
+    },
+    {
+      "epoch": 0.22133980439738216,
+      "grad_norm": 0.9325829148292542,
+      "learning_rate": 0.0001814192729371262,
+      "loss": 0.5516,
+      "step": 3010
+    },
+    {
+      "epoch": 0.22141333921611883,
+      "grad_norm": 1.5447005033493042,
+      "learning_rate": 0.00018140544307886932,
+      "loss": 0.5302,
+      "step": 3011
+    },
+    {
+      "epoch": 0.2214868740348555,
+      "grad_norm": 0.9622365236282349,
+      "learning_rate": 0.00018139160860320978,
+      "loss": 0.5368,
+      "step": 3012
+    },
+    {
+      "epoch": 0.22156040885359218,
+      "grad_norm": 1.5810542106628418,
+      "learning_rate": 0.0001813777695109323,
+      "loss": 0.529,
+      "step": 3013
+    },
+    {
+      "epoch": 0.22163394367232886,
+      "grad_norm": 1.3386882543563843,
+      "learning_rate": 0.00018136392580282186,
+      "loss": 0.5208,
+      "step": 3014
+    },
+    {
+      "epoch": 0.2217074784910655,
+      "grad_norm": 1.5296090841293335,
+      "learning_rate": 0.0001813500774796637,
+      "loss": 0.5171,
+      "step": 3015
+    },
+    {
+      "epoch": 0.22178101330980218,
+      "grad_norm": 1.4581636190414429,
+      "learning_rate": 0.0001813362245422433,
+      "loss": 0.5408,
+      "step": 3016
+    },
+    {
+      "epoch": 0.22185454812853886,
+      "grad_norm": 1.3624422550201416,
+      "learning_rate": 0.00018132236699134637,
+      "loss": 0.5278,
+      "step": 3017
+    },
+    {
+      "epoch": 0.22192808294727553,
+      "grad_norm": 1.2575162649154663,
+      "learning_rate": 0.00018130850482775897,
+      "loss": 0.5138,
+      "step": 3018
+    },
+    {
+      "epoch": 0.2220016177660122,
+      "grad_norm": 1.285211205482483,
+      "learning_rate": 0.00018129463805226738,
+      "loss": 0.5286,
+      "step": 3019
+    },
+    {
+      "epoch": 0.22207515258474889,
+      "grad_norm": 0.7743325233459473,
+      "learning_rate": 0.00018128076666565812,
+      "loss": 0.5045,
+      "step": 3020
+    },
+    {
+      "epoch": 0.22214868740348556,
+      "grad_norm": 1.2699816226959229,
+      "learning_rate": 0.000181266890668718,
+      "loss": 0.5573,
+      "step": 3021
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 0.6297335624694824,
+      "learning_rate": 0.00018125301006223404,
+      "loss": 0.5114,
+      "step": 3022
+    },
+    {
+      "epoch": 0.22229575704095889,
+      "grad_norm": 1.2781275510787964,
+      "learning_rate": 0.00018123912484699367,
+      "loss": 0.545,
+      "step": 3023
+    },
+    {
+      "epoch": 0.22236929185969556,
+      "grad_norm": 0.6280782222747803,
+      "learning_rate": 0.00018122523502378437,
+      "loss": 0.5162,
+      "step": 3024
+    },
+    {
+      "epoch": 0.22244282667843224,
+      "grad_norm": 1.307356595993042,
+      "learning_rate": 0.000181211340593394,
+      "loss": 0.5415,
+      "step": 3025
+    },
+    {
+      "epoch": 0.2225163614971689,
+      "grad_norm": 0.560202956199646,
+      "learning_rate": 0.0001811974415566107,
+      "loss": 0.5243,
+      "step": 3026
+    },
+    {
+      "epoch": 0.2225898963159056,
+      "grad_norm": 1.3329452276229858,
+      "learning_rate": 0.0001811835379142228,
+      "loss": 0.5136,
+      "step": 3027
+    },
+    {
+      "epoch": 0.22266343113464226,
+      "grad_norm": 0.6758899688720703,
+      "learning_rate": 0.000181169629667019,
+      "loss": 0.5204,
+      "step": 3028
+    },
+    {
+      "epoch": 0.2227369659533789,
+      "grad_norm": 1.2574031352996826,
+      "learning_rate": 0.0001811557168157881,
+      "loss": 0.544,
+      "step": 3029
+    },
+    {
+      "epoch": 0.2228105007721156,
+      "grad_norm": 0.7752076387405396,
+      "learning_rate": 0.00018114179936131933,
+      "loss": 0.5182,
+      "step": 3030
+    },
+    {
+      "epoch": 0.22288403559085226,
+      "grad_norm": 0.9776637554168701,
+      "learning_rate": 0.00018112787730440206,
+      "loss": 0.5057,
+      "step": 3031
+    },
+    {
+      "epoch": 0.22295757040958894,
+      "grad_norm": 0.7321799993515015,
+      "learning_rate": 0.00018111395064582597,
+      "loss": 0.5085,
+      "step": 3032
+    },
+    {
+      "epoch": 0.22303110522832562,
+      "grad_norm": 0.8070792555809021,
+      "learning_rate": 0.000181100019386381,
+      "loss": 0.5615,
+      "step": 3033
+    },
+    {
+      "epoch": 0.2231046400470623,
+      "grad_norm": 0.7019020915031433,
+      "learning_rate": 0.00018108608352685734,
+      "loss": 0.5324,
+      "step": 3034
+    },
+    {
+      "epoch": 0.22317817486579897,
+      "grad_norm": 0.6784930229187012,
+      "learning_rate": 0.00018107214306804542,
+      "loss": 0.5381,
+      "step": 3035
+    },
+    {
+      "epoch": 0.22325170968453562,
+      "grad_norm": 0.6365440487861633,
+      "learning_rate": 0.00018105819801073604,
+      "loss": 0.5126,
+      "step": 3036
+    },
+    {
+      "epoch": 0.2233252445032723,
+      "grad_norm": 0.7887497544288635,
+      "learning_rate": 0.00018104424835572007,
+      "loss": 0.5285,
+      "step": 3037
+    },
+    {
+      "epoch": 0.22339877932200897,
+      "grad_norm": 0.7934899926185608,
+      "learning_rate": 0.00018103029410378881,
+      "loss": 0.5413,
+      "step": 3038
+    },
+    {
+      "epoch": 0.22347231414074564,
+      "grad_norm": 0.6918529868125916,
+      "learning_rate": 0.00018101633525573378,
+      "loss": 0.5474,
+      "step": 3039
+    },
+    {
+      "epoch": 0.22354584895948232,
+      "grad_norm": 0.8107700347900391,
+      "learning_rate": 0.0001810023718123467,
+      "loss": 0.5503,
+      "step": 3040
+    },
+    {
+      "epoch": 0.223619383778219,
+      "grad_norm": 0.8838568329811096,
+      "learning_rate": 0.00018098840377441962,
+      "loss": 0.5398,
+      "step": 3041
+    },
+    {
+      "epoch": 0.22369291859695567,
+      "grad_norm": 0.8887888193130493,
+      "learning_rate": 0.0001809744311427448,
+      "loss": 0.5106,
+      "step": 3042
+    },
+    {
+      "epoch": 0.22376645341569232,
+      "grad_norm": 0.5996225476264954,
+      "learning_rate": 0.0001809604539181148,
+      "loss": 0.5357,
+      "step": 3043
+    },
+    {
+      "epoch": 0.223839988234429,
+      "grad_norm": 0.9603798389434814,
+      "learning_rate": 0.00018094647210132242,
+      "loss": 0.5444,
+      "step": 3044
+    },
+    {
+      "epoch": 0.22391352305316567,
+      "grad_norm": 0.9682263731956482,
+      "learning_rate": 0.00018093248569316068,
+      "loss": 0.5308,
+      "step": 3045
+    },
+    {
+      "epoch": 0.22398705787190235,
+      "grad_norm": 0.5701063871383667,
+      "learning_rate": 0.00018091849469442295,
+      "loss": 0.5031,
+      "step": 3046
+    },
+    {
+      "epoch": 0.22406059269063902,
+      "grad_norm": 1.323050856590271,
+      "learning_rate": 0.0001809044991059028,
+      "loss": 0.5409,
+      "step": 3047
+    },
+    {
+      "epoch": 0.2241341275093757,
+      "grad_norm": 0.738703191280365,
+      "learning_rate": 0.0001808904989283941,
+      "loss": 0.5628,
+      "step": 3048
+    },
+    {
+      "epoch": 0.22420766232811237,
+      "grad_norm": 0.6796239614486694,
+      "learning_rate": 0.00018087649416269095,
+      "loss": 0.5113,
+      "step": 3049
+    },
+    {
+      "epoch": 0.22428119714684902,
+      "grad_norm": 0.8646931052207947,
+      "learning_rate": 0.00018086248480958766,
+      "loss": 0.506,
+      "step": 3050
+    },
+    {
+      "epoch": 0.2243547319655857,
+      "grad_norm": 0.9002664685249329,
+      "learning_rate": 0.0001808484708698789,
+      "loss": 0.5301,
+      "step": 3051
+    },
+    {
+      "epoch": 0.22442826678432237,
+      "grad_norm": 0.6328312158584595,
+      "learning_rate": 0.00018083445234435956,
+      "loss": 0.5249,
+      "step": 3052
+    },
+    {
+      "epoch": 0.22450180160305905,
+      "grad_norm": 0.6152592301368713,
+      "learning_rate": 0.00018082042923382477,
+      "loss": 0.5188,
+      "step": 3053
+    },
+    {
+      "epoch": 0.22457533642179572,
+      "grad_norm": 0.7026908993721008,
+      "learning_rate": 0.00018080640153906994,
+      "loss": 0.5276,
+      "step": 3054
+    },
+    {
+      "epoch": 0.2246488712405324,
+      "grad_norm": 0.6074684262275696,
+      "learning_rate": 0.00018079236926089073,
+      "loss": 0.5359,
+      "step": 3055
+    },
+    {
+      "epoch": 0.22472240605926908,
+      "grad_norm": 0.6889917254447937,
+      "learning_rate": 0.00018077833240008308,
+      "loss": 0.5258,
+      "step": 3056
+    },
+    {
+      "epoch": 0.22479594087800572,
+      "grad_norm": 0.6339432001113892,
+      "learning_rate": 0.00018076429095744318,
+      "loss": 0.5277,
+      "step": 3057
+    },
+    {
+      "epoch": 0.2248694756967424,
+      "grad_norm": 0.8525620698928833,
+      "learning_rate": 0.00018075024493376747,
+      "loss": 0.5053,
+      "step": 3058
+    },
+    {
+      "epoch": 0.22494301051547907,
+      "grad_norm": 0.9036215543746948,
+      "learning_rate": 0.00018073619432985264,
+      "loss": 0.4966,
+      "step": 3059
+    },
+    {
+      "epoch": 0.22501654533421575,
+      "grad_norm": 0.6496038436889648,
+      "learning_rate": 0.00018072213914649562,
+      "loss": 0.5006,
+      "step": 3060
+    },
+    {
+      "epoch": 0.22509008015295243,
+      "grad_norm": 1.2909085750579834,
+      "learning_rate": 0.00018070807938449373,
+      "loss": 0.5556,
+      "step": 3061
+    },
+    {
+      "epoch": 0.2251636149716891,
+      "grad_norm": 0.8376094102859497,
+      "learning_rate": 0.0001806940150446444,
+      "loss": 0.5063,
+      "step": 3062
+    },
+    {
+      "epoch": 0.22523714979042578,
+      "grad_norm": 0.6260842084884644,
+      "learning_rate": 0.00018067994612774535,
+      "loss": 0.5437,
+      "step": 3063
+    },
+    {
+      "epoch": 0.22531068460916243,
+      "grad_norm": 0.8623825907707214,
+      "learning_rate": 0.0001806658726345946,
+      "loss": 0.5062,
+      "step": 3064
+    },
+    {
+      "epoch": 0.2253842194278991,
+      "grad_norm": 0.6364173293113708,
+      "learning_rate": 0.00018065179456599047,
+      "loss": 0.5124,
+      "step": 3065
+    },
+    {
+      "epoch": 0.22545775424663578,
+      "grad_norm": 0.6941555142402649,
+      "learning_rate": 0.0001806377119227314,
+      "loss": 0.5174,
+      "step": 3066
+    },
+    {
+      "epoch": 0.22553128906537245,
+      "grad_norm": 0.6140319108963013,
+      "learning_rate": 0.0001806236247056162,
+      "loss": 0.5205,
+      "step": 3067
+    },
+    {
+      "epoch": 0.22560482388410913,
+      "grad_norm": 0.5681672692298889,
+      "learning_rate": 0.00018060953291544396,
+      "loss": 0.555,
+      "step": 3068
+    },
+    {
+      "epoch": 0.2256783587028458,
+      "grad_norm": 0.7637145519256592,
+      "learning_rate": 0.00018059543655301395,
+      "loss": 0.4938,
+      "step": 3069
+    },
+    {
+      "epoch": 0.22575189352158248,
+      "grad_norm": 0.9820735454559326,
+      "learning_rate": 0.00018058133561912567,
+      "loss": 0.5452,
+      "step": 3070
+    },
+    {
+      "epoch": 0.22582542834031913,
+      "grad_norm": 0.7238828539848328,
+      "learning_rate": 0.00018056723011457902,
+      "loss": 0.5429,
+      "step": 3071
+    },
+    {
+      "epoch": 0.2258989631590558,
+      "grad_norm": 0.6457898020744324,
+      "learning_rate": 0.00018055312004017405,
+      "loss": 0.5123,
+      "step": 3072
+    },
+    {
+      "epoch": 0.22597249797779248,
+      "grad_norm": 0.8162586688995361,
+      "learning_rate": 0.0001805390053967111,
+      "loss": 0.4899,
+      "step": 3073
+    },
+    {
+      "epoch": 0.22604603279652916,
+      "grad_norm": 0.7217814922332764,
+      "learning_rate": 0.00018052488618499078,
+      "loss": 0.5127,
+      "step": 3074
+    },
+    {
+      "epoch": 0.22611956761526583,
+      "grad_norm": 0.6414519548416138,
+      "learning_rate": 0.00018051076240581393,
+      "loss": 0.5562,
+      "step": 3075
+    },
+    {
+      "epoch": 0.2261931024340025,
+      "grad_norm": 0.8061267733573914,
+      "learning_rate": 0.00018049663405998167,
+      "loss": 0.5216,
+      "step": 3076
+    },
+    {
+      "epoch": 0.22626663725273918,
+      "grad_norm": 0.7553830146789551,
+      "learning_rate": 0.00018048250114829535,
+      "loss": 0.5269,
+      "step": 3077
+    },
+    {
+      "epoch": 0.22634017207147583,
+      "grad_norm": 0.7285786867141724,
+      "learning_rate": 0.00018046836367155666,
+      "loss": 0.5464,
+      "step": 3078
+    },
+    {
+      "epoch": 0.2264137068902125,
+      "grad_norm": 0.6305744051933289,
+      "learning_rate": 0.00018045422163056745,
+      "loss": 0.5191,
+      "step": 3079
+    },
+    {
+      "epoch": 0.22648724170894918,
+      "grad_norm": 0.7322538495063782,
+      "learning_rate": 0.00018044007502612987,
+      "loss": 0.4827,
+      "step": 3080
+    },
+    {
+      "epoch": 0.22656077652768586,
+      "grad_norm": 0.7398054003715515,
+      "learning_rate": 0.00018042592385904635,
+      "loss": 0.5353,
+      "step": 3081
+    },
+    {
+      "epoch": 0.22663431134642253,
+      "grad_norm": 0.7575973272323608,
+      "learning_rate": 0.00018041176813011957,
+      "loss": 0.5205,
+      "step": 3082
+    },
+    {
+      "epoch": 0.2267078461651592,
+      "grad_norm": 1.1059502363204956,
+      "learning_rate": 0.00018039760784015243,
+      "loss": 0.4916,
+      "step": 3083
+    },
+    {
+      "epoch": 0.22678138098389589,
+      "grad_norm": 0.9677029252052307,
+      "learning_rate": 0.0001803834429899481,
+      "loss": 0.5407,
+      "step": 3084
+    },
+    {
+      "epoch": 0.22685491580263253,
+      "grad_norm": 0.9165573120117188,
+      "learning_rate": 0.0001803692735803101,
+      "loss": 0.5288,
+      "step": 3085
+    },
+    {
+      "epoch": 0.2269284506213692,
+      "grad_norm": 0.8137401342391968,
+      "learning_rate": 0.00018035509961204205,
+      "loss": 0.5246,
+      "step": 3086
+    },
+    {
+      "epoch": 0.22700198544010589,
+      "grad_norm": 0.7558435797691345,
+      "learning_rate": 0.00018034092108594796,
+      "loss": 0.5126,
+      "step": 3087
+    },
+    {
+      "epoch": 0.22707552025884256,
+      "grad_norm": 1.1888021230697632,
+      "learning_rate": 0.00018032673800283204,
+      "loss": 0.5398,
+      "step": 3088
+    },
+    {
+      "epoch": 0.22714905507757924,
+      "grad_norm": 0.596146285533905,
+      "learning_rate": 0.00018031255036349877,
+      "loss": 0.5304,
+      "step": 3089
+    },
+    {
+      "epoch": 0.2272225898963159,
+      "grad_norm": 1.399083137512207,
+      "learning_rate": 0.00018029835816875286,
+      "loss": 0.5106,
+      "step": 3090
+    },
+    {
+      "epoch": 0.2272961247150526,
+      "grad_norm": 0.8315428495407104,
+      "learning_rate": 0.00018028416141939934,
+      "loss": 0.5432,
+      "step": 3091
+    },
+    {
+      "epoch": 0.22736965953378924,
+      "grad_norm": 0.9779343605041504,
+      "learning_rate": 0.0001802699601162435,
+      "loss": 0.522,
+      "step": 3092
+    },
+    {
+      "epoch": 0.2274431943525259,
+      "grad_norm": 1.5273317098617554,
+      "learning_rate": 0.0001802557542600908,
+      "loss": 0.5119,
+      "step": 3093
+    },
+    {
+      "epoch": 0.2275167291712626,
+      "grad_norm": 0.9510782361030579,
+      "learning_rate": 0.000180241543851747,
+      "loss": 0.5497,
+      "step": 3094
+    },
+    {
+      "epoch": 0.22759026398999926,
+      "grad_norm": 2.0662176609039307,
+      "learning_rate": 0.00018022732889201812,
+      "loss": 0.54,
+      "step": 3095
+    },
+    {
+      "epoch": 0.22766379880873594,
+      "grad_norm": 1.7783054113388062,
+      "learning_rate": 0.0001802131093817105,
+      "loss": 0.5548,
+      "step": 3096
+    },
+    {
+      "epoch": 0.22773733362747262,
+      "grad_norm": 1.6743342876434326,
+      "learning_rate": 0.0001801988853216307,
+      "loss": 0.5253,
+      "step": 3097
+    },
+    {
+      "epoch": 0.2278108684462093,
+      "grad_norm": 1.7447561025619507,
+      "learning_rate": 0.00018018465671258547,
+      "loss": 0.5186,
+      "step": 3098
+    },
+    {
+      "epoch": 0.22788440326494594,
+      "grad_norm": 1.4999849796295166,
+      "learning_rate": 0.00018017042355538188,
+      "loss": 0.5315,
+      "step": 3099
+    },
+    {
+      "epoch": 0.22795793808368262,
+      "grad_norm": 1.2950221300125122,
+      "learning_rate": 0.00018015618585082724,
+      "loss": 0.508,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2280314729024193,
+      "grad_norm": 1.5859287977218628,
+      "learning_rate": 0.00018014194359972918,
+      "loss": 0.523,
+      "step": 3101
+    },
+    {
+      "epoch": 0.22810500772115597,
+      "grad_norm": 1.458201289176941,
+      "learning_rate": 0.00018012769680289547,
+      "loss": 0.5279,
+      "step": 3102
+    },
+    {
+      "epoch": 0.22817854253989264,
+      "grad_norm": 1.5418975353240967,
+      "learning_rate": 0.00018011344546113424,
+      "loss": 0.5143,
+      "step": 3103
+    },
+    {
+      "epoch": 0.22825207735862932,
+      "grad_norm": 1.3917889595031738,
+      "learning_rate": 0.00018009918957525388,
+      "loss": 0.517,
+      "step": 3104
+    },
+    {
+      "epoch": 0.228325612177366,
+      "grad_norm": 1.8034965991973877,
+      "learning_rate": 0.0001800849291460629,
+      "loss": 0.53,
+      "step": 3105
+    },
+    {
+      "epoch": 0.22839914699610264,
+      "grad_norm": 1.5921541452407837,
+      "learning_rate": 0.00018007066417437026,
+      "loss": 0.502,
+      "step": 3106
+    },
+    {
+      "epoch": 0.22847268181483932,
+      "grad_norm": 1.530457854270935,
+      "learning_rate": 0.00018005639466098502,
+      "loss": 0.4728,
+      "step": 3107
+    },
+    {
+      "epoch": 0.228546216633576,
+      "grad_norm": 1.613051414489746,
+      "learning_rate": 0.0001800421206067166,
+      "loss": 0.5192,
+      "step": 3108
+    },
+    {
+      "epoch": 0.22861975145231267,
+      "grad_norm": 1.1361582279205322,
+      "learning_rate": 0.00018002784201237462,
+      "loss": 0.5397,
+      "step": 3109
+    },
+    {
+      "epoch": 0.22869328627104935,
+      "grad_norm": 1.1316853761672974,
+      "learning_rate": 0.00018001355887876897,
+      "loss": 0.5289,
+      "step": 3110
+    },
+    {
+      "epoch": 0.22876682108978602,
+      "grad_norm": 1.4322855472564697,
+      "learning_rate": 0.00017999927120670986,
+      "loss": 0.5369,
+      "step": 3111
+    },
+    {
+      "epoch": 0.2288403559085227,
+      "grad_norm": 0.9391867518424988,
+      "learning_rate": 0.00017998497899700765,
+      "loss": 0.5216,
+      "step": 3112
+    },
+    {
+      "epoch": 0.22891389072725934,
+      "grad_norm": 1.713124394416809,
+      "learning_rate": 0.00017997068225047302,
+      "loss": 0.5554,
+      "step": 3113
+    },
+    {
+      "epoch": 0.22898742554599602,
+      "grad_norm": 1.3684942722320557,
+      "learning_rate": 0.00017995638096791688,
+      "loss": 0.5116,
+      "step": 3114
+    },
+    {
+      "epoch": 0.2290609603647327,
+      "grad_norm": 1.4238970279693604,
+      "learning_rate": 0.00017994207515015042,
+      "loss": 0.5276,
+      "step": 3115
+    },
+    {
+      "epoch": 0.22913449518346937,
+      "grad_norm": 1.1278380155563354,
+      "learning_rate": 0.00017992776479798512,
+      "loss": 0.5413,
+      "step": 3116
+    },
+    {
+      "epoch": 0.22920803000220605,
+      "grad_norm": 2.0153989791870117,
+      "learning_rate": 0.00017991344991223263,
+      "loss": 0.5464,
+      "step": 3117
+    },
+    {
+      "epoch": 0.22928156482094272,
+      "grad_norm": 1.5502228736877441,
+      "learning_rate": 0.00017989913049370495,
+      "loss": 0.5136,
+      "step": 3118
+    },
+    {
+      "epoch": 0.2293550996396794,
+      "grad_norm": 1.7264782190322876,
+      "learning_rate": 0.00017988480654321424,
+      "loss": 0.5329,
+      "step": 3119
+    },
+    {
+      "epoch": 0.22942863445841605,
+      "grad_norm": 1.5331337451934814,
+      "learning_rate": 0.000179870478061573,
+      "loss": 0.5319,
+      "step": 3120
+    },
+    {
+      "epoch": 0.22950216927715272,
+      "grad_norm": 1.339672327041626,
+      "learning_rate": 0.00017985614504959397,
+      "loss": 0.5212,
+      "step": 3121
+    },
+    {
+      "epoch": 0.2295757040958894,
+      "grad_norm": 1.4618507623672485,
+      "learning_rate": 0.0001798418075080901,
+      "loss": 0.5282,
+      "step": 3122
+    },
+    {
+      "epoch": 0.22964923891462607,
+      "grad_norm": 1.4696424007415771,
+      "learning_rate": 0.00017982746543787465,
+      "loss": 0.5435,
+      "step": 3123
+    },
+    {
+      "epoch": 0.22972277373336275,
+      "grad_norm": 1.34312105178833,
+      "learning_rate": 0.00017981311883976113,
+      "loss": 0.5462,
+      "step": 3124
+    },
+    {
+      "epoch": 0.22979630855209943,
+      "grad_norm": 1.375502109527588,
+      "learning_rate": 0.00017979876771456326,
+      "loss": 0.5168,
+      "step": 3125
+    },
+    {
+      "epoch": 0.2298698433708361,
+      "grad_norm": 1.1883599758148193,
+      "learning_rate": 0.0001797844120630951,
+      "loss": 0.5446,
+      "step": 3126
+    },
+    {
+      "epoch": 0.22994337818957275,
+      "grad_norm": 1.243552803993225,
+      "learning_rate": 0.00017977005188617086,
+      "loss": 0.5278,
+      "step": 3127
+    },
+    {
+      "epoch": 0.23001691300830943,
+      "grad_norm": 1.0758516788482666,
+      "learning_rate": 0.0001797556871846051,
+      "loss": 0.525,
+      "step": 3128
+    },
+    {
+      "epoch": 0.2300904478270461,
+      "grad_norm": 1.6998021602630615,
+      "learning_rate": 0.0001797413179592126,
+      "loss": 0.5191,
+      "step": 3129
+    },
+    {
+      "epoch": 0.23016398264578278,
+      "grad_norm": 1.233639121055603,
+      "learning_rate": 0.00017972694421080837,
+      "loss": 0.5806,
+      "step": 3130
+    },
+    {
+      "epoch": 0.23023751746451945,
+      "grad_norm": 1.8989745378494263,
+      "learning_rate": 0.00017971256594020773,
+      "loss": 0.5467,
+      "step": 3131
+    },
+    {
+      "epoch": 0.23031105228325613,
+      "grad_norm": 2.2017319202423096,
+      "learning_rate": 0.00017969818314822624,
+      "loss": 0.555,
+      "step": 3132
+    },
+    {
+      "epoch": 0.2303845871019928,
+      "grad_norm": 1.1337777376174927,
+      "learning_rate": 0.00017968379583567973,
+      "loss": 0.5477,
+      "step": 3133
+    },
+    {
+      "epoch": 0.23045812192072945,
+      "grad_norm": 1.392576813697815,
+      "learning_rate": 0.00017966940400338416,
+      "loss": 0.5249,
+      "step": 3134
+    },
+    {
+      "epoch": 0.23053165673946613,
+      "grad_norm": 0.9758958220481873,
+      "learning_rate": 0.00017965500765215593,
+      "loss": 0.537,
+      "step": 3135
+    },
+    {
+      "epoch": 0.2306051915582028,
+      "grad_norm": 1.2852485179901123,
+      "learning_rate": 0.00017964060678281164,
+      "loss": 0.5612,
+      "step": 3136
+    },
+    {
+      "epoch": 0.23067872637693948,
+      "grad_norm": 0.8679046630859375,
+      "learning_rate": 0.00017962620139616803,
+      "loss": 0.5171,
+      "step": 3137
+    },
+    {
+      "epoch": 0.23075226119567616,
+      "grad_norm": 1.4607651233673096,
+      "learning_rate": 0.0001796117914930423,
+      "loss": 0.5271,
+      "step": 3138
+    },
+    {
+      "epoch": 0.23082579601441283,
+      "grad_norm": 0.9169206023216248,
+      "learning_rate": 0.00017959737707425171,
+      "loss": 0.5841,
+      "step": 3139
+    },
+    {
+      "epoch": 0.2308993308331495,
+      "grad_norm": 1.5769826173782349,
+      "learning_rate": 0.00017958295814061388,
+      "loss": 0.5277,
+      "step": 3140
+    },
+    {
+      "epoch": 0.23097286565188616,
+      "grad_norm": 1.3459501266479492,
+      "learning_rate": 0.00017956853469294667,
+      "loss": 0.558,
+      "step": 3141
+    },
+    {
+      "epoch": 0.23104640047062283,
+      "grad_norm": 1.2536426782608032,
+      "learning_rate": 0.00017955410673206826,
+      "loss": 0.5352,
+      "step": 3142
+    },
+    {
+      "epoch": 0.2311199352893595,
+      "grad_norm": 1.2397085428237915,
+      "learning_rate": 0.0001795396742587969,
+      "loss": 0.5629,
+      "step": 3143
+    },
+    {
+      "epoch": 0.23119347010809618,
+      "grad_norm": 1.005817174911499,
+      "learning_rate": 0.0001795252372739513,
+      "loss": 0.5433,
+      "step": 3144
+    },
+    {
+      "epoch": 0.23126700492683286,
+      "grad_norm": 1.288962483406067,
+      "learning_rate": 0.0001795107957783503,
+      "loss": 0.5377,
+      "step": 3145
+    },
+    {
+      "epoch": 0.23134053974556953,
+      "grad_norm": 0.7910787463188171,
+      "learning_rate": 0.00017949634977281305,
+      "loss": 0.5466,
+      "step": 3146
+    },
+    {
+      "epoch": 0.2314140745643062,
+      "grad_norm": 1.2703769207000732,
+      "learning_rate": 0.00017948189925815896,
+      "loss": 0.5532,
+      "step": 3147
+    },
+    {
+      "epoch": 0.23148760938304286,
+      "grad_norm": 0.7892640233039856,
+      "learning_rate": 0.00017946744423520767,
+      "loss": 0.5054,
+      "step": 3148
+    },
+    {
+      "epoch": 0.23156114420177953,
+      "grad_norm": 0.9602465033531189,
+      "learning_rate": 0.00017945298470477906,
+      "loss": 0.5162,
+      "step": 3149
+    },
+    {
+      "epoch": 0.2316346790205162,
+      "grad_norm": 0.8257586359977722,
+      "learning_rate": 0.0001794385206676933,
+      "loss": 0.5385,
+      "step": 3150
+    },
+    {
+      "epoch": 0.23170821383925289,
+      "grad_norm": 0.9210731983184814,
+      "learning_rate": 0.00017942405212477086,
+      "loss": 0.5238,
+      "step": 3151
+    },
+    {
+      "epoch": 0.23178174865798956,
+      "grad_norm": 0.7127185463905334,
+      "learning_rate": 0.00017940957907683233,
+      "loss": 0.5431,
+      "step": 3152
+    },
+    {
+      "epoch": 0.23185528347672624,
+      "grad_norm": 0.8464113473892212,
+      "learning_rate": 0.00017939510152469867,
+      "loss": 0.5198,
+      "step": 3153
+    },
+    {
+      "epoch": 0.2319288182954629,
+      "grad_norm": 0.6597235798835754,
+      "learning_rate": 0.00017938061946919107,
+      "loss": 0.5287,
+      "step": 3154
+    },
+    {
+      "epoch": 0.23200235311419956,
+      "grad_norm": 0.638741672039032,
+      "learning_rate": 0.00017936613291113094,
+      "loss": 0.5751,
+      "step": 3155
+    },
+    {
+      "epoch": 0.23207588793293624,
+      "grad_norm": 0.6284526586532593,
+      "learning_rate": 0.00017935164185134002,
+      "loss": 0.528,
+      "step": 3156
+    },
+    {
+      "epoch": 0.2321494227516729,
+      "grad_norm": 0.779295027256012,
+      "learning_rate": 0.00017933714629064023,
+      "loss": 0.5231,
+      "step": 3157
+    },
+    {
+      "epoch": 0.2322229575704096,
+      "grad_norm": 0.7517540454864502,
+      "learning_rate": 0.00017932264622985374,
+      "loss": 0.5266,
+      "step": 3158
+    },
+    {
+      "epoch": 0.23229649238914626,
+      "grad_norm": 0.7404706478118896,
+      "learning_rate": 0.00017930814166980307,
+      "loss": 0.547,
+      "step": 3159
+    },
+    {
+      "epoch": 0.23237002720788294,
+      "grad_norm": 1.2140958309173584,
+      "learning_rate": 0.0001792936326113109,
+      "loss": 0.5381,
+      "step": 3160
+    },
+    {
+      "epoch": 0.23244356202661962,
+      "grad_norm": 0.6454242467880249,
+      "learning_rate": 0.0001792791190552002,
+      "loss": 0.515,
+      "step": 3161
+    },
+    {
+      "epoch": 0.23251709684535626,
+      "grad_norm": 1.5269042253494263,
+      "learning_rate": 0.00017926460100229423,
+      "loss": 0.5501,
+      "step": 3162
+    },
+    {
+      "epoch": 0.23259063166409294,
+      "grad_norm": 0.7543985843658447,
+      "learning_rate": 0.00017925007845341642,
+      "loss": 0.5254,
+      "step": 3163
+    },
+    {
+      "epoch": 0.23266416648282962,
+      "grad_norm": 1.787487268447876,
+      "learning_rate": 0.00017923555140939053,
+      "loss": 0.5058,
+      "step": 3164
+    },
+    {
+      "epoch": 0.2327377013015663,
+      "grad_norm": 1.1462316513061523,
+      "learning_rate": 0.00017922101987104055,
+      "loss": 0.5016,
+      "step": 3165
+    },
+    {
+      "epoch": 0.23281123612030297,
+      "grad_norm": 1.947117567062378,
+      "learning_rate": 0.0001792064838391907,
+      "loss": 0.5358,
+      "step": 3166
+    },
+    {
+      "epoch": 0.23288477093903964,
+      "grad_norm": 1.727980375289917,
+      "learning_rate": 0.0001791919433146655,
+      "loss": 0.5401,
+      "step": 3167
+    },
+    {
+      "epoch": 0.23295830575777632,
+      "grad_norm": 1.0562809705734253,
+      "learning_rate": 0.0001791773982982897,
+      "loss": 0.5318,
+      "step": 3168
+    },
+    {
+      "epoch": 0.23303184057651297,
+      "grad_norm": 0.8148366808891296,
+      "learning_rate": 0.00017916284879088829,
+      "loss": 0.5174,
+      "step": 3169
+    },
+    {
+      "epoch": 0.23310537539524964,
+      "grad_norm": 1.253170132637024,
+      "learning_rate": 0.00017914829479328658,
+      "loss": 0.5313,
+      "step": 3170
+    },
+    {
+      "epoch": 0.23317891021398632,
+      "grad_norm": 0.6402531266212463,
+      "learning_rate": 0.00017913373630631008,
+      "loss": 0.5246,
+      "step": 3171
+    },
+    {
+      "epoch": 0.233252445032723,
+      "grad_norm": 1.1424249410629272,
+      "learning_rate": 0.00017911917333078453,
+      "loss": 0.5143,
+      "step": 3172
+    },
+    {
+      "epoch": 0.23332597985145967,
+      "grad_norm": 1.0288498401641846,
+      "learning_rate": 0.00017910460586753594,
+      "loss": 0.519,
+      "step": 3173
+    },
+    {
+      "epoch": 0.23339951467019635,
+      "grad_norm": 0.7351285219192505,
+      "learning_rate": 0.00017909003391739065,
+      "loss": 0.5682,
+      "step": 3174
+    },
+    {
+      "epoch": 0.23347304948893302,
+      "grad_norm": 0.9924508333206177,
+      "learning_rate": 0.00017907545748117515,
+      "loss": 0.5267,
+      "step": 3175
+    },
+    {
+      "epoch": 0.23354658430766967,
+      "grad_norm": 0.7266468405723572,
+      "learning_rate": 0.00017906087655971626,
+      "loss": 0.5575,
+      "step": 3176
+    },
+    {
+      "epoch": 0.23362011912640634,
+      "grad_norm": 0.6440916061401367,
+      "learning_rate": 0.00017904629115384103,
+      "loss": 0.5318,
+      "step": 3177
+    },
+    {
+      "epoch": 0.23369365394514302,
+      "grad_norm": 0.7686768770217896,
+      "learning_rate": 0.00017903170126437674,
+      "loss": 0.5085,
+      "step": 3178
+    },
+    {
+      "epoch": 0.2337671887638797,
+      "grad_norm": 0.7251720428466797,
+      "learning_rate": 0.00017901710689215093,
+      "loss": 0.543,
+      "step": 3179
+    },
+    {
+      "epoch": 0.23384072358261637,
+      "grad_norm": 0.6307001709938049,
+      "learning_rate": 0.00017900250803799144,
+      "loss": 0.5697,
+      "step": 3180
+    },
+    {
+      "epoch": 0.23391425840135305,
+      "grad_norm": 0.6474977731704712,
+      "learning_rate": 0.0001789879047027263,
+      "loss": 0.5317,
+      "step": 3181
+    },
+    {
+      "epoch": 0.23398779322008972,
+      "grad_norm": 0.67755126953125,
+      "learning_rate": 0.00017897329688718386,
+      "loss": 0.5254,
+      "step": 3182
+    },
+    {
+      "epoch": 0.23406132803882637,
+      "grad_norm": 0.7944599986076355,
+      "learning_rate": 0.00017895868459219273,
+      "loss": 0.493,
+      "step": 3183
+    },
+    {
+      "epoch": 0.23413486285756305,
+      "grad_norm": 0.9276211261749268,
+      "learning_rate": 0.0001789440678185816,
+      "loss": 0.5592,
+      "step": 3184
+    },
+    {
+      "epoch": 0.23420839767629972,
+      "grad_norm": 0.7298025488853455,
+      "learning_rate": 0.00017892944656717967,
+      "loss": 0.5237,
+      "step": 3185
+    },
+    {
+      "epoch": 0.2342819324950364,
+      "grad_norm": 0.6264386773109436,
+      "learning_rate": 0.00017891482083881623,
+      "loss": 0.5606,
+      "step": 3186
+    },
+    {
+      "epoch": 0.23435546731377307,
+      "grad_norm": 0.8703876733779907,
+      "learning_rate": 0.00017890019063432087,
+      "loss": 0.5246,
+      "step": 3187
+    },
+    {
+      "epoch": 0.23442900213250975,
+      "grad_norm": 0.9654495120048523,
+      "learning_rate": 0.0001788855559545234,
+      "loss": 0.5382,
+      "step": 3188
+    },
+    {
+      "epoch": 0.23450253695124643,
+      "grad_norm": 0.5907428860664368,
+      "learning_rate": 0.000178870916800254,
+      "loss": 0.5279,
+      "step": 3189
+    },
+    {
+      "epoch": 0.23457607176998307,
+      "grad_norm": 0.9454459547996521,
+      "learning_rate": 0.00017885627317234292,
+      "loss": 0.5315,
+      "step": 3190
+    },
+    {
+      "epoch": 0.23464960658871975,
+      "grad_norm": 1.262338638305664,
+      "learning_rate": 0.00017884162507162082,
+      "loss": 0.5236,
+      "step": 3191
+    },
+    {
+      "epoch": 0.23472314140745643,
+      "grad_norm": 0.8067899942398071,
+      "learning_rate": 0.00017882697249891854,
+      "loss": 0.5468,
+      "step": 3192
+    },
+    {
+      "epoch": 0.2347966762261931,
+      "grad_norm": 0.5544705986976624,
+      "learning_rate": 0.00017881231545506721,
+      "loss": 0.5127,
+      "step": 3193
+    },
+    {
+      "epoch": 0.23487021104492978,
+      "grad_norm": 0.7082903981208801,
+      "learning_rate": 0.00017879765394089814,
+      "loss": 0.5227,
+      "step": 3194
+    },
+    {
+      "epoch": 0.23494374586366645,
+      "grad_norm": 0.8847938776016235,
+      "learning_rate": 0.000178782987957243,
+      "loss": 0.5255,
+      "step": 3195
+    },
+    {
+      "epoch": 0.23501728068240313,
+      "grad_norm": 0.795775830745697,
+      "learning_rate": 0.00017876831750493363,
+      "loss": 0.5407,
+      "step": 3196
+    },
+    {
+      "epoch": 0.23509081550113978,
+      "grad_norm": 0.6575978994369507,
+      "learning_rate": 0.00017875364258480216,
+      "loss": 0.5598,
+      "step": 3197
+    },
+    {
+      "epoch": 0.23516435031987645,
+      "grad_norm": 0.7581764459609985,
+      "learning_rate": 0.000178738963197681,
+      "loss": 0.5105,
+      "step": 3198
+    },
+    {
+      "epoch": 0.23523788513861313,
+      "grad_norm": 1.2716944217681885,
+      "learning_rate": 0.00017872427934440273,
+      "loss": 0.5876,
+      "step": 3199
+    },
+    {
+      "epoch": 0.2353114199573498,
+      "grad_norm": 0.6582130193710327,
+      "learning_rate": 0.00017870959102580025,
+      "loss": 0.5182,
+      "step": 3200
+    },
+    {
+      "epoch": 0.23538495477608648,
+      "grad_norm": 0.7281584143638611,
+      "learning_rate": 0.00017869489824270672,
+      "loss": 0.5308,
+      "step": 3201
+    },
+    {
+      "epoch": 0.23545848959482316,
+      "grad_norm": 0.6283013820648193,
+      "learning_rate": 0.0001786802009959555,
+      "loss": 0.5196,
+      "step": 3202
+    },
+    {
+      "epoch": 0.23553202441355983,
+      "grad_norm": 0.7866979241371155,
+      "learning_rate": 0.00017866549928638022,
+      "loss": 0.5008,
+      "step": 3203
+    },
+    {
+      "epoch": 0.23560555923229648,
+      "grad_norm": 1.0993731021881104,
+      "learning_rate": 0.00017865079311481482,
+      "loss": 0.5598,
+      "step": 3204
+    },
+    {
+      "epoch": 0.23567909405103316,
+      "grad_norm": 0.9130511283874512,
+      "learning_rate": 0.00017863608248209345,
+      "loss": 0.5373,
+      "step": 3205
+    },
+    {
+      "epoch": 0.23575262886976983,
+      "grad_norm": 0.7703906297683716,
+      "learning_rate": 0.00017862136738905048,
+      "loss": 0.5407,
+      "step": 3206
+    },
+    {
+      "epoch": 0.2358261636885065,
+      "grad_norm": 0.6448803544044495,
+      "learning_rate": 0.00017860664783652056,
+      "loss": 0.5075,
+      "step": 3207
+    },
+    {
+      "epoch": 0.23589969850724318,
+      "grad_norm": 1.0177886486053467,
+      "learning_rate": 0.00017859192382533865,
+      "loss": 0.5202,
+      "step": 3208
+    },
+    {
+      "epoch": 0.23597323332597986,
+      "grad_norm": 1.4698995351791382,
+      "learning_rate": 0.00017857719535633985,
+      "loss": 0.5513,
+      "step": 3209
+    },
+    {
+      "epoch": 0.23604676814471653,
+      "grad_norm": 0.6019389033317566,
+      "learning_rate": 0.00017856246243035963,
+      "loss": 0.5377,
+      "step": 3210
+    },
+    {
+      "epoch": 0.23612030296345318,
+      "grad_norm": 1.0798295736312866,
+      "learning_rate": 0.0001785477250482336,
+      "loss": 0.528,
+      "step": 3211
+    },
+    {
+      "epoch": 0.23619383778218986,
+      "grad_norm": 0.9453423023223877,
+      "learning_rate": 0.00017853298321079776,
+      "loss": 0.4807,
+      "step": 3212
+    },
+    {
+      "epoch": 0.23626737260092653,
+      "grad_norm": 0.6428825855255127,
+      "learning_rate": 0.0001785182369188882,
+      "loss": 0.5201,
+      "step": 3213
+    },
+    {
+      "epoch": 0.2363409074196632,
+      "grad_norm": 0.6674232482910156,
+      "learning_rate": 0.00017850348617334135,
+      "loss": 0.5124,
+      "step": 3214
+    },
+    {
+      "epoch": 0.23641444223839989,
+      "grad_norm": 0.9731060862541199,
+      "learning_rate": 0.00017848873097499397,
+      "loss": 0.5505,
+      "step": 3215
+    },
+    {
+      "epoch": 0.23648797705713656,
+      "grad_norm": 1.1321821212768555,
+      "learning_rate": 0.00017847397132468294,
+      "loss": 0.5046,
+      "step": 3216
+    },
+    {
+      "epoch": 0.23656151187587324,
+      "grad_norm": 0.8465587496757507,
+      "learning_rate": 0.00017845920722324538,
+      "loss": 0.4912,
+      "step": 3217
+    },
+    {
+      "epoch": 0.23663504669460989,
+      "grad_norm": 0.5495790839195251,
+      "learning_rate": 0.00017844443867151883,
+      "loss": 0.498,
+      "step": 3218
+    },
+    {
+      "epoch": 0.23670858151334656,
+      "grad_norm": 1.0680586099624634,
+      "learning_rate": 0.00017842966567034092,
+      "loss": 0.5404,
+      "step": 3219
+    },
+    {
+      "epoch": 0.23678211633208324,
+      "grad_norm": 1.314777135848999,
+      "learning_rate": 0.0001784148882205496,
+      "loss": 0.5161,
+      "step": 3220
+    },
+    {
+      "epoch": 0.2368556511508199,
+      "grad_norm": 0.6852834820747375,
+      "learning_rate": 0.00017840010632298308,
+      "loss": 0.538,
+      "step": 3221
+    },
+    {
+      "epoch": 0.2369291859695566,
+      "grad_norm": 1.4184808731079102,
+      "learning_rate": 0.00017838531997847978,
+      "loss": 0.5152,
+      "step": 3222
+    },
+    {
+      "epoch": 0.23700272078829326,
+      "grad_norm": 0.756058394908905,
+      "learning_rate": 0.0001783705291878784,
+      "loss": 0.5422,
+      "step": 3223
+    },
+    {
+      "epoch": 0.23707625560702994,
+      "grad_norm": 0.6005216836929321,
+      "learning_rate": 0.00017835573395201792,
+      "loss": 0.5119,
+      "step": 3224
+    },
+    {
+      "epoch": 0.2371497904257666,
+      "grad_norm": 0.9470319747924805,
+      "learning_rate": 0.0001783409342717375,
+      "loss": 0.5117,
+      "step": 3225
+    },
+    {
+      "epoch": 0.23722332524450326,
+      "grad_norm": 0.9777472615242004,
+      "learning_rate": 0.0001783261301478766,
+      "loss": 0.5369,
+      "step": 3226
+    },
+    {
+      "epoch": 0.23729686006323994,
+      "grad_norm": 0.8512688875198364,
+      "learning_rate": 0.00017831132158127496,
+      "loss": 0.5098,
+      "step": 3227
+    },
+    {
+      "epoch": 0.23737039488197662,
+      "grad_norm": 0.6358827352523804,
+      "learning_rate": 0.0001782965085727725,
+      "loss": 0.5288,
+      "step": 3228
+    },
+    {
+      "epoch": 0.2374439297007133,
+      "grad_norm": 0.5504454374313354,
+      "learning_rate": 0.00017828169112320947,
+      "loss": 0.5072,
+      "step": 3229
+    },
+    {
+      "epoch": 0.23751746451944997,
+      "grad_norm": 0.7223479747772217,
+      "learning_rate": 0.0001782668692334263,
+      "loss": 0.5541,
+      "step": 3230
+    },
+    {
+      "epoch": 0.23759099933818664,
+      "grad_norm": 0.6727440357208252,
+      "learning_rate": 0.00017825204290426367,
+      "loss": 0.5013,
+      "step": 3231
+    },
+    {
+      "epoch": 0.2376645341569233,
+      "grad_norm": 0.5376930236816406,
+      "learning_rate": 0.0001782372121365626,
+      "loss": 0.5185,
+      "step": 3232
+    },
+    {
+      "epoch": 0.23773806897565997,
+      "grad_norm": 0.8265194892883301,
+      "learning_rate": 0.0001782223769311643,
+      "loss": 0.5482,
+      "step": 3233
+    },
+    {
+      "epoch": 0.23781160379439664,
+      "grad_norm": 0.9248981475830078,
+      "learning_rate": 0.00017820753728891025,
+      "loss": 0.4992,
+      "step": 3234
+    },
+    {
+      "epoch": 0.23788513861313332,
+      "grad_norm": 0.8250258564949036,
+      "learning_rate": 0.00017819269321064213,
+      "loss": 0.5159,
+      "step": 3235
+    },
+    {
+      "epoch": 0.23795867343187,
+      "grad_norm": 0.6082636713981628,
+      "learning_rate": 0.0001781778446972019,
+      "loss": 0.5439,
+      "step": 3236
+    },
+    {
+      "epoch": 0.23803220825060667,
+      "grad_norm": 0.6778237819671631,
+      "learning_rate": 0.00017816299174943184,
+      "loss": 0.5116,
+      "step": 3237
+    },
+    {
+      "epoch": 0.23810574306934335,
+      "grad_norm": 1.171964406967163,
+      "learning_rate": 0.00017814813436817443,
+      "loss": 0.5112,
+      "step": 3238
+    },
+    {
+      "epoch": 0.23817927788808,
+      "grad_norm": 1.1319867372512817,
+      "learning_rate": 0.00017813327255427233,
+      "loss": 0.5364,
+      "step": 3239
+    },
+    {
+      "epoch": 0.23825281270681667,
+      "grad_norm": 0.6984413266181946,
+      "learning_rate": 0.00017811840630856858,
+      "loss": 0.5429,
+      "step": 3240
+    },
+    {
+      "epoch": 0.23832634752555334,
+      "grad_norm": 0.7275168299674988,
+      "learning_rate": 0.00017810353563190635,
+      "loss": 0.512,
+      "step": 3241
+    },
+    {
+      "epoch": 0.23839988234429002,
+      "grad_norm": 0.8972579836845398,
+      "learning_rate": 0.00017808866052512916,
+      "loss": 0.4942,
+      "step": 3242
+    },
+    {
+      "epoch": 0.2384734171630267,
+      "grad_norm": 0.8024085164070129,
+      "learning_rate": 0.00017807378098908075,
+      "loss": 0.4996,
+      "step": 3243
+    },
+    {
+      "epoch": 0.23854695198176337,
+      "grad_norm": 0.5620863437652588,
+      "learning_rate": 0.0001780588970246051,
+      "loss": 0.5124,
+      "step": 3244
+    },
+    {
+      "epoch": 0.23862048680050005,
+      "grad_norm": 0.592943012714386,
+      "learning_rate": 0.0001780440086325464,
+      "loss": 0.5287,
+      "step": 3245
+    },
+    {
+      "epoch": 0.2386940216192367,
+      "grad_norm": 0.6995099782943726,
+      "learning_rate": 0.00017802911581374917,
+      "loss": 0.4848,
+      "step": 3246
+    },
+    {
+      "epoch": 0.23876755643797337,
+      "grad_norm": 0.6860783696174622,
+      "learning_rate": 0.00017801421856905816,
+      "loss": 0.4852,
+      "step": 3247
+    },
+    {
+      "epoch": 0.23884109125671005,
+      "grad_norm": 0.6919470429420471,
+      "learning_rate": 0.0001779993168993183,
+      "loss": 0.5042,
+      "step": 3248
+    },
+    {
+      "epoch": 0.23891462607544672,
+      "grad_norm": 0.6326781511306763,
+      "learning_rate": 0.00017798441080537494,
+      "loss": 0.516,
+      "step": 3249
+    },
+    {
+      "epoch": 0.2389881608941834,
+      "grad_norm": 0.6665195822715759,
+      "learning_rate": 0.00017796950028807346,
+      "loss": 0.5061,
+      "step": 3250
+    },
+    {
+      "epoch": 0.23906169571292007,
+      "grad_norm": 0.6875540614128113,
+      "learning_rate": 0.00017795458534825963,
+      "loss": 0.5034,
+      "step": 3251
+    },
+    {
+      "epoch": 0.23913523053165675,
+      "grad_norm": 0.7194721102714539,
+      "learning_rate": 0.00017793966598677946,
+      "loss": 0.5249,
+      "step": 3252
+    },
+    {
+      "epoch": 0.2392087653503934,
+      "grad_norm": 1.2353428602218628,
+      "learning_rate": 0.00017792474220447918,
+      "loss": 0.5266,
+      "step": 3253
+    },
+    {
+      "epoch": 0.23928230016913007,
+      "grad_norm": 0.7887352108955383,
+      "learning_rate": 0.0001779098140022053,
+      "loss": 0.5288,
+      "step": 3254
+    },
+    {
+      "epoch": 0.23935583498786675,
+      "grad_norm": 0.5252473950386047,
+      "learning_rate": 0.00017789488138080455,
+      "loss": 0.5068,
+      "step": 3255
+    },
+    {
+      "epoch": 0.23942936980660343,
+      "grad_norm": 0.7392098903656006,
+      "learning_rate": 0.00017787994434112392,
+      "loss": 0.5208,
+      "step": 3256
+    },
+    {
+      "epoch": 0.2395029046253401,
+      "grad_norm": 0.9486863613128662,
+      "learning_rate": 0.00017786500288401064,
+      "loss": 0.5082,
+      "step": 3257
+    },
+    {
+      "epoch": 0.23957643944407678,
+      "grad_norm": 1.1246739625930786,
+      "learning_rate": 0.00017785005701031223,
+      "loss": 0.5334,
+      "step": 3258
+    },
+    {
+      "epoch": 0.23964997426281345,
+      "grad_norm": 0.8276765942573547,
+      "learning_rate": 0.00017783510672087643,
+      "loss": 0.499,
+      "step": 3259
+    },
+    {
+      "epoch": 0.2397235090815501,
+      "grad_norm": 0.6085115075111389,
+      "learning_rate": 0.00017782015201655123,
+      "loss": 0.5154,
+      "step": 3260
+    },
+    {
+      "epoch": 0.23979704390028678,
+      "grad_norm": 0.5479869246482849,
+      "learning_rate": 0.00017780519289818488,
+      "loss": 0.5237,
+      "step": 3261
+    },
+    {
+      "epoch": 0.23987057871902345,
+      "grad_norm": 0.8235727548599243,
+      "learning_rate": 0.00017779022936662586,
+      "loss": 0.4886,
+      "step": 3262
+    },
+    {
+      "epoch": 0.23994411353776013,
+      "grad_norm": 0.8937106132507324,
+      "learning_rate": 0.00017777526142272296,
+      "loss": 0.5471,
+      "step": 3263
+    },
+    {
+      "epoch": 0.2400176483564968,
+      "grad_norm": 0.6820952892303467,
+      "learning_rate": 0.00017776028906732515,
+      "loss": 0.5242,
+      "step": 3264
+    },
+    {
+      "epoch": 0.24009118317523348,
+      "grad_norm": 0.6086716651916504,
+      "learning_rate": 0.0001777453123012817,
+      "loss": 0.487,
+      "step": 3265
+    },
+    {
+      "epoch": 0.24016471799397016,
+      "grad_norm": 0.5869222283363342,
+      "learning_rate": 0.00017773033112544203,
+      "loss": 0.5183,
+      "step": 3266
+    },
+    {
+      "epoch": 0.2402382528127068,
+      "grad_norm": 0.8675061464309692,
+      "learning_rate": 0.00017771534554065599,
+      "loss": 0.4944,
+      "step": 3267
+    },
+    {
+      "epoch": 0.24031178763144348,
+      "grad_norm": 1.1876158714294434,
+      "learning_rate": 0.00017770035554777352,
+      "loss": 0.5066,
+      "step": 3268
+    },
+    {
+      "epoch": 0.24038532245018016,
+      "grad_norm": 1.1457223892211914,
+      "learning_rate": 0.0001776853611476449,
+      "loss": 0.5392,
+      "step": 3269
+    },
+    {
+      "epoch": 0.24045885726891683,
+      "grad_norm": 0.970066249370575,
+      "learning_rate": 0.00017767036234112062,
+      "loss": 0.4901,
+      "step": 3270
+    },
+    {
+      "epoch": 0.2405323920876535,
+      "grad_norm": 0.7567079067230225,
+      "learning_rate": 0.00017765535912905137,
+      "loss": 0.5386,
+      "step": 3271
+    },
+    {
+      "epoch": 0.24060592690639018,
+      "grad_norm": 0.6031343340873718,
+      "learning_rate": 0.00017764035151228823,
+      "loss": 0.5296,
+      "step": 3272
+    },
+    {
+      "epoch": 0.24067946172512686,
+      "grad_norm": 0.656265139579773,
+      "learning_rate": 0.00017762533949168243,
+      "loss": 0.5053,
+      "step": 3273
+    },
+    {
+      "epoch": 0.2407529965438635,
+      "grad_norm": 0.7647505402565002,
+      "learning_rate": 0.00017761032306808542,
+      "loss": 0.5378,
+      "step": 3274
+    },
+    {
+      "epoch": 0.24082653136260018,
+      "grad_norm": 0.5961447954177856,
+      "learning_rate": 0.00017759530224234902,
+      "loss": 0.5307,
+      "step": 3275
+    },
+    {
+      "epoch": 0.24090006618133686,
+      "grad_norm": 0.6864814162254333,
+      "learning_rate": 0.00017758027701532517,
+      "loss": 0.5005,
+      "step": 3276
+    },
+    {
+      "epoch": 0.24097360100007353,
+      "grad_norm": 0.6611002683639526,
+      "learning_rate": 0.0001775652473878661,
+      "loss": 0.5057,
+      "step": 3277
+    },
+    {
+      "epoch": 0.2410471358188102,
+      "grad_norm": 0.6315354108810425,
+      "learning_rate": 0.00017755021336082435,
+      "loss": 0.5081,
+      "step": 3278
+    },
+    {
+      "epoch": 0.24112067063754689,
+      "grad_norm": 0.8532983660697937,
+      "learning_rate": 0.0001775351749350527,
+      "loss": 0.5411,
+      "step": 3279
+    },
+    {
+      "epoch": 0.24119420545628356,
+      "grad_norm": 0.8003588914871216,
+      "learning_rate": 0.00017752013211140409,
+      "loss": 0.5126,
+      "step": 3280
+    },
+    {
+      "epoch": 0.2412677402750202,
+      "grad_norm": 0.9724708795547485,
+      "learning_rate": 0.00017750508489073173,
+      "loss": 0.5451,
+      "step": 3281
+    },
+    {
+      "epoch": 0.24134127509375689,
+      "grad_norm": 1.5560559034347534,
+      "learning_rate": 0.00017749003327388921,
+      "loss": 0.5278,
+      "step": 3282
+    },
+    {
+      "epoch": 0.24141480991249356,
+      "grad_norm": 0.700489342212677,
+      "learning_rate": 0.0001774749772617302,
+      "loss": 0.5144,
+      "step": 3283
+    },
+    {
+      "epoch": 0.24148834473123024,
+      "grad_norm": 1.4402235746383667,
+      "learning_rate": 0.00017745991685510872,
+      "loss": 0.5117,
+      "step": 3284
+    },
+    {
+      "epoch": 0.2415618795499669,
+      "grad_norm": 1.2321498394012451,
+      "learning_rate": 0.000177444852054879,
+      "loss": 0.5285,
+      "step": 3285
+    },
+    {
+      "epoch": 0.2416354143687036,
+      "grad_norm": 1.0200116634368896,
+      "learning_rate": 0.00017742978286189556,
+      "loss": 0.4833,
+      "step": 3286
+    },
+    {
+      "epoch": 0.24170894918744026,
+      "grad_norm": 1.0136723518371582,
+      "learning_rate": 0.00017741470927701313,
+      "loss": 0.5045,
+      "step": 3287
+    },
+    {
+      "epoch": 0.2417824840061769,
+      "grad_norm": 1.076463222503662,
+      "learning_rate": 0.00017739963130108664,
+      "loss": 0.5365,
+      "step": 3288
+    },
+    {
+      "epoch": 0.2418560188249136,
+      "grad_norm": 1.5140540599822998,
+      "learning_rate": 0.00017738454893497145,
+      "loss": 0.5365,
+      "step": 3289
+    },
+    {
+      "epoch": 0.24192955364365026,
+      "grad_norm": 0.7378954291343689,
+      "learning_rate": 0.00017736946217952295,
+      "loss": 0.5368,
+      "step": 3290
+    },
+    {
+      "epoch": 0.24200308846238694,
+      "grad_norm": 2.1352195739746094,
+      "learning_rate": 0.0001773543710355969,
+      "loss": 0.5419,
+      "step": 3291
+    },
+    {
+      "epoch": 0.24207662328112362,
+      "grad_norm": 1.3270277976989746,
+      "learning_rate": 0.00017733927550404928,
+      "loss": 0.5108,
+      "step": 3292
+    },
+    {
+      "epoch": 0.2421501580998603,
+      "grad_norm": 2.7129104137420654,
+      "learning_rate": 0.00017732417558573632,
+      "loss": 0.5323,
+      "step": 3293
+    },
+    {
+      "epoch": 0.24222369291859697,
+      "grad_norm": 2.5540785789489746,
+      "learning_rate": 0.00017730907128151455,
+      "loss": 0.4976,
+      "step": 3294
+    },
+    {
+      "epoch": 0.24229722773733361,
+      "grad_norm": 1.083964228630066,
+      "learning_rate": 0.00017729396259224064,
+      "loss": 0.5187,
+      "step": 3295
+    },
+    {
+      "epoch": 0.2423707625560703,
+      "grad_norm": 1.639491081237793,
+      "learning_rate": 0.00017727884951877163,
+      "loss": 0.5321,
+      "step": 3296
+    },
+    {
+      "epoch": 0.24244429737480697,
+      "grad_norm": 1.2214126586914062,
+      "learning_rate": 0.0001772637320619647,
+      "loss": 0.5522,
+      "step": 3297
+    },
+    {
+      "epoch": 0.24251783219354364,
+      "grad_norm": 1.230615496635437,
+      "learning_rate": 0.00017724861022267737,
+      "loss": 0.4952,
+      "step": 3298
+    },
+    {
+      "epoch": 0.24259136701228032,
+      "grad_norm": 1.2257493734359741,
+      "learning_rate": 0.00017723348400176732,
+      "loss": 0.55,
+      "step": 3299
+    },
+    {
+      "epoch": 0.242664901831017,
+      "grad_norm": 1.0142035484313965,
+      "learning_rate": 0.00017721835340009254,
+      "loss": 0.5616,
+      "step": 3300
+    },
+    {
+      "epoch": 0.24273843664975367,
+      "grad_norm": 1.6587673425674438,
+      "learning_rate": 0.0001772032184185113,
+      "loss": 0.5287,
+      "step": 3301
+    },
+    {
+      "epoch": 0.24281197146849032,
+      "grad_norm": 1.2059881687164307,
+      "learning_rate": 0.000177188079057882,
+      "loss": 0.5353,
+      "step": 3302
+    },
+    {
+      "epoch": 0.242885506287227,
+      "grad_norm": 1.8394315242767334,
+      "learning_rate": 0.00017717293531906343,
+      "loss": 0.517,
+      "step": 3303
+    },
+    {
+      "epoch": 0.24295904110596367,
+      "grad_norm": 1.5391452312469482,
+      "learning_rate": 0.0001771577872029145,
+      "loss": 0.5288,
+      "step": 3304
+    },
+    {
+      "epoch": 0.24303257592470034,
+      "grad_norm": 1.469218134880066,
+      "learning_rate": 0.00017714263471029447,
+      "loss": 0.5516,
+      "step": 3305
+    },
+    {
+      "epoch": 0.24310611074343702,
+      "grad_norm": 1.092346429824829,
+      "learning_rate": 0.00017712747784206275,
+      "loss": 0.5174,
+      "step": 3306
+    },
+    {
+      "epoch": 0.2431796455621737,
+      "grad_norm": 1.8403077125549316,
+      "learning_rate": 0.00017711231659907914,
+      "loss": 0.5476,
+      "step": 3307
+    },
+    {
+      "epoch": 0.24325318038091037,
+      "grad_norm": 1.40420401096344,
+      "learning_rate": 0.00017709715098220352,
+      "loss": 0.5131,
+      "step": 3308
+    },
+    {
+      "epoch": 0.24332671519964702,
+      "grad_norm": 1.7101171016693115,
+      "learning_rate": 0.00017708198099229615,
+      "loss": 0.4963,
+      "step": 3309
+    },
+    {
+      "epoch": 0.2434002500183837,
+      "grad_norm": 1.733238935470581,
+      "learning_rate": 0.00017706680663021746,
+      "loss": 0.5193,
+      "step": 3310
+    },
+    {
+      "epoch": 0.24347378483712037,
+      "grad_norm": 1.361214518547058,
+      "learning_rate": 0.00017705162789682816,
+      "loss": 0.5087,
+      "step": 3311
+    },
+    {
+      "epoch": 0.24354731965585705,
+      "grad_norm": 1.446799874305725,
+      "learning_rate": 0.00017703644479298921,
+      "loss": 0.5448,
+      "step": 3312
+    },
+    {
+      "epoch": 0.24362085447459372,
+      "grad_norm": 1.9828060865402222,
+      "learning_rate": 0.00017702125731956184,
+      "loss": 0.5344,
+      "step": 3313
+    },
+    {
+      "epoch": 0.2436943892933304,
+      "grad_norm": 1.5851126909255981,
+      "learning_rate": 0.00017700606547740746,
+      "loss": 0.5471,
+      "step": 3314
+    },
+    {
+      "epoch": 0.24376792411206707,
+      "grad_norm": 1.7168811559677124,
+      "learning_rate": 0.00017699086926738775,
+      "loss": 0.5534,
+      "step": 3315
+    },
+    {
+      "epoch": 0.24384145893080372,
+      "grad_norm": 1.4371575117111206,
+      "learning_rate": 0.00017697566869036472,
+      "loss": 0.5546,
+      "step": 3316
+    },
+    {
+      "epoch": 0.2439149937495404,
+      "grad_norm": 1.5104557275772095,
+      "learning_rate": 0.0001769604637472005,
+      "loss": 0.503,
+      "step": 3317
+    },
+    {
+      "epoch": 0.24398852856827707,
+      "grad_norm": 1.361249327659607,
+      "learning_rate": 0.00017694525443875757,
+      "loss": 0.5256,
+      "step": 3318
+    },
+    {
+      "epoch": 0.24406206338701375,
+      "grad_norm": 1.499018907546997,
+      "learning_rate": 0.0001769300407658986,
+      "loss": 0.5149,
+      "step": 3319
+    },
+    {
+      "epoch": 0.24413559820575043,
+      "grad_norm": 1.292407512664795,
+      "learning_rate": 0.00017691482272948653,
+      "loss": 0.5359,
+      "step": 3320
+    },
+    {
+      "epoch": 0.2442091330244871,
+      "grad_norm": 1.7259883880615234,
+      "learning_rate": 0.00017689960033038458,
+      "loss": 0.5165,
+      "step": 3321
+    },
+    {
+      "epoch": 0.24428266784322378,
+      "grad_norm": 1.5149351358413696,
+      "learning_rate": 0.00017688437356945609,
+      "loss": 0.5081,
+      "step": 3322
+    },
+    {
+      "epoch": 0.24435620266196043,
+      "grad_norm": 1.4708054065704346,
+      "learning_rate": 0.0001768691424475648,
+      "loss": 0.5262,
+      "step": 3323
+    },
+    {
+      "epoch": 0.2444297374806971,
+      "grad_norm": 1.2986764907836914,
+      "learning_rate": 0.00017685390696557465,
+      "loss": 0.5395,
+      "step": 3324
+    },
+    {
+      "epoch": 0.24450327229943378,
+      "grad_norm": 1.7840728759765625,
+      "learning_rate": 0.00017683866712434978,
+      "loss": 0.5406,
+      "step": 3325
+    },
+    {
+      "epoch": 0.24457680711817045,
+      "grad_norm": 1.6200637817382812,
+      "learning_rate": 0.0001768234229247546,
+      "loss": 0.4944,
+      "step": 3326
+    },
+    {
+      "epoch": 0.24465034193690713,
+      "grad_norm": 1.661208987236023,
+      "learning_rate": 0.0001768081743676538,
+      "loss": 0.5071,
+      "step": 3327
+    },
+    {
+      "epoch": 0.2447238767556438,
+      "grad_norm": 1.5023136138916016,
+      "learning_rate": 0.0001767929214539123,
+      "loss": 0.5023,
+      "step": 3328
+    },
+    {
+      "epoch": 0.24479741157438048,
+      "grad_norm": 1.4780282974243164,
+      "learning_rate": 0.00017677766418439529,
+      "loss": 0.4938,
+      "step": 3329
+    },
+    {
+      "epoch": 0.24487094639311713,
+      "grad_norm": 1.019964575767517,
+      "learning_rate": 0.00017676240255996812,
+      "loss": 0.5129,
+      "step": 3330
+    },
+    {
+      "epoch": 0.2449444812118538,
+      "grad_norm": 1.6500020027160645,
+      "learning_rate": 0.00017674713658149643,
+      "loss": 0.5664,
+      "step": 3331
+    },
+    {
+      "epoch": 0.24501801603059048,
+      "grad_norm": 1.319800853729248,
+      "learning_rate": 0.0001767318662498462,
+      "loss": 0.5145,
+      "step": 3332
+    },
+    {
+      "epoch": 0.24509155084932716,
+      "grad_norm": 1.8916194438934326,
+      "learning_rate": 0.00017671659156588348,
+      "loss": 0.4894,
+      "step": 3333
+    },
+    {
+      "epoch": 0.24516508566806383,
+      "grad_norm": 1.7976435422897339,
+      "learning_rate": 0.00017670131253047478,
+      "loss": 0.5109,
+      "step": 3334
+    },
+    {
+      "epoch": 0.2452386204868005,
+      "grad_norm": 1.302098274230957,
+      "learning_rate": 0.0001766860291444867,
+      "loss": 0.5384,
+      "step": 3335
+    },
+    {
+      "epoch": 0.24531215530553718,
+      "grad_norm": 1.4755133390426636,
+      "learning_rate": 0.0001766707414087861,
+      "loss": 0.4855,
+      "step": 3336
+    },
+    {
+      "epoch": 0.24538569012427383,
+      "grad_norm": 1.431869387626648,
+      "learning_rate": 0.0001766554493242401,
+      "loss": 0.5008,
+      "step": 3337
+    },
+    {
+      "epoch": 0.2454592249430105,
+      "grad_norm": 1.3506207466125488,
+      "learning_rate": 0.00017664015289171615,
+      "loss": 0.5637,
+      "step": 3338
+    },
+    {
+      "epoch": 0.24553275976174718,
+      "grad_norm": 1.5770291090011597,
+      "learning_rate": 0.00017662485211208183,
+      "loss": 0.4984,
+      "step": 3339
+    },
+    {
+      "epoch": 0.24560629458048386,
+      "grad_norm": 1.3264353275299072,
+      "learning_rate": 0.00017660954698620502,
+      "loss": 0.529,
+      "step": 3340
+    },
+    {
+      "epoch": 0.24567982939922053,
+      "grad_norm": 1.3953007459640503,
+      "learning_rate": 0.0001765942375149539,
+      "loss": 0.5005,
+      "step": 3341
+    },
+    {
+      "epoch": 0.2457533642179572,
+      "grad_norm": 1.2087165117263794,
+      "learning_rate": 0.00017657892369919676,
+      "loss": 0.5417,
+      "step": 3342
+    },
+    {
+      "epoch": 0.24582689903669389,
+      "grad_norm": 1.6700578927993774,
+      "learning_rate": 0.00017656360553980224,
+      "loss": 0.5269,
+      "step": 3343
+    },
+    {
+      "epoch": 0.24590043385543053,
+      "grad_norm": 1.4768197536468506,
+      "learning_rate": 0.00017654828303763926,
+      "loss": 0.5032,
+      "step": 3344
+    },
+    {
+      "epoch": 0.2459739686741672,
+      "grad_norm": 1.2841062545776367,
+      "learning_rate": 0.00017653295619357687,
+      "loss": 0.5108,
+      "step": 3345
+    },
+    {
+      "epoch": 0.24604750349290389,
+      "grad_norm": 1.1352163553237915,
+      "learning_rate": 0.0001765176250084844,
+      "loss": 0.5025,
+      "step": 3346
+    },
+    {
+      "epoch": 0.24612103831164056,
+      "grad_norm": 1.6485034227371216,
+      "learning_rate": 0.0001765022894832315,
+      "loss": 0.5194,
+      "step": 3347
+    },
+    {
+      "epoch": 0.24619457313037724,
+      "grad_norm": 1.1153843402862549,
+      "learning_rate": 0.00017648694961868802,
+      "loss": 0.5003,
+      "step": 3348
+    },
+    {
+      "epoch": 0.2462681079491139,
+      "grad_norm": 1.8381693363189697,
+      "learning_rate": 0.00017647160541572402,
+      "loss": 0.5284,
+      "step": 3349
+    },
+    {
+      "epoch": 0.2463416427678506,
+      "grad_norm": 1.9335131645202637,
+      "learning_rate": 0.00017645625687520987,
+      "loss": 0.5375,
+      "step": 3350
+    },
+    {
+      "epoch": 0.24641517758658724,
+      "grad_norm": 1.3404608964920044,
+      "learning_rate": 0.00017644090399801613,
+      "loss": 0.5233,
+      "step": 3351
+    },
+    {
+      "epoch": 0.2464887124053239,
+      "grad_norm": 1.1653528213500977,
+      "learning_rate": 0.00017642554678501366,
+      "loss": 0.5389,
+      "step": 3352
+    },
+    {
+      "epoch": 0.2465622472240606,
+      "grad_norm": 1.1620469093322754,
+      "learning_rate": 0.0001764101852370735,
+      "loss": 0.5468,
+      "step": 3353
+    },
+    {
+      "epoch": 0.24663578204279726,
+      "grad_norm": 0.9703438878059387,
+      "learning_rate": 0.00017639481935506702,
+      "loss": 0.5333,
+      "step": 3354
+    },
+    {
+      "epoch": 0.24670931686153394,
+      "grad_norm": 1.5395838022232056,
+      "learning_rate": 0.00017637944913986576,
+      "loss": 0.5345,
+      "step": 3355
+    },
+    {
+      "epoch": 0.24678285168027062,
+      "grad_norm": 0.9253973364830017,
+      "learning_rate": 0.00017636407459234153,
+      "loss": 0.5142,
+      "step": 3356
+    },
+    {
+      "epoch": 0.2468563864990073,
+      "grad_norm": 1.7917006015777588,
+      "learning_rate": 0.00017634869571336642,
+      "loss": 0.5515,
+      "step": 3357
+    },
+    {
+      "epoch": 0.24692992131774394,
+      "grad_norm": 1.4626343250274658,
+      "learning_rate": 0.00017633331250381268,
+      "loss": 0.5417,
+      "step": 3358
+    },
+    {
+      "epoch": 0.24700345613648061,
+      "grad_norm": 1.8260400295257568,
+      "learning_rate": 0.00017631792496455292,
+      "loss": 0.4968,
+      "step": 3359
+    },
+    {
+      "epoch": 0.2470769909552173,
+      "grad_norm": 1.657170057296753,
+      "learning_rate": 0.0001763025330964599,
+      "loss": 0.5212,
+      "step": 3360
+    },
+    {
+      "epoch": 0.24715052577395397,
+      "grad_norm": 1.0119084119796753,
+      "learning_rate": 0.0001762871369004067,
+      "loss": 0.5384,
+      "step": 3361
+    },
+    {
+      "epoch": 0.24722406059269064,
+      "grad_norm": 1.145167350769043,
+      "learning_rate": 0.00017627173637726657,
+      "loss": 0.5071,
+      "step": 3362
+    },
+    {
+      "epoch": 0.24729759541142732,
+      "grad_norm": 1.5250725746154785,
+      "learning_rate": 0.00017625633152791305,
+      "loss": 0.4878,
+      "step": 3363
+    },
+    {
+      "epoch": 0.247371130230164,
+      "grad_norm": 0.8985742330551147,
+      "learning_rate": 0.00017624092235321993,
+      "loss": 0.5195,
+      "step": 3364
+    },
+    {
+      "epoch": 0.24744466504890064,
+      "grad_norm": 1.5195847749710083,
+      "learning_rate": 0.00017622550885406125,
+      "loss": 0.5269,
+      "step": 3365
+    },
+    {
+      "epoch": 0.24751819986763732,
+      "grad_norm": 1.142179250717163,
+      "learning_rate": 0.00017621009103131125,
+      "loss": 0.5126,
+      "step": 3366
+    },
+    {
+      "epoch": 0.247591734686374,
+      "grad_norm": 1.5041115283966064,
+      "learning_rate": 0.00017619466888584446,
+      "loss": 0.4941,
+      "step": 3367
+    },
+    {
+      "epoch": 0.24766526950511067,
+      "grad_norm": 1.3160271644592285,
+      "learning_rate": 0.00017617924241853565,
+      "loss": 0.5424,
+      "step": 3368
+    },
+    {
+      "epoch": 0.24773880432384734,
+      "grad_norm": 1.9514425992965698,
+      "learning_rate": 0.0001761638116302598,
+      "loss": 0.5155,
+      "step": 3369
+    },
+    {
+      "epoch": 0.24781233914258402,
+      "grad_norm": 1.597891092300415,
+      "learning_rate": 0.0001761483765218922,
+      "loss": 0.5377,
+      "step": 3370
+    },
+    {
+      "epoch": 0.2478858739613207,
+      "grad_norm": 1.1418787240982056,
+      "learning_rate": 0.0001761329370943083,
+      "loss": 0.5088,
+      "step": 3371
+    },
+    {
+      "epoch": 0.24795940878005734,
+      "grad_norm": 1.3442343473434448,
+      "learning_rate": 0.00017611749334838387,
+      "loss": 0.5071,
+      "step": 3372
+    },
+    {
+      "epoch": 0.24803294359879402,
+      "grad_norm": 0.9597196578979492,
+      "learning_rate": 0.00017610204528499489,
+      "loss": 0.485,
+      "step": 3373
+    },
+    {
+      "epoch": 0.2481064784175307,
+      "grad_norm": 1.3506253957748413,
+      "learning_rate": 0.00017608659290501758,
+      "loss": 0.4897,
+      "step": 3374
+    },
+    {
+      "epoch": 0.24818001323626737,
+      "grad_norm": 0.7476764917373657,
+      "learning_rate": 0.00017607113620932846,
+      "loss": 0.5163,
+      "step": 3375
+    },
+    {
+      "epoch": 0.24825354805500405,
+      "grad_norm": 1.524475336074829,
+      "learning_rate": 0.00017605567519880414,
+      "loss": 0.4833,
+      "step": 3376
+    },
+    {
+      "epoch": 0.24832708287374072,
+      "grad_norm": 0.6100408434867859,
+      "learning_rate": 0.00017604020987432173,
+      "loss": 0.4902,
+      "step": 3377
+    },
+    {
+      "epoch": 0.2484006176924774,
+      "grad_norm": 1.6818281412124634,
+      "learning_rate": 0.0001760247402367583,
+      "loss": 0.5395,
+      "step": 3378
+    },
+    {
+      "epoch": 0.24847415251121405,
+      "grad_norm": 0.8439916968345642,
+      "learning_rate": 0.00017600926628699143,
+      "loss": 0.5088,
+      "step": 3379
+    },
+    {
+      "epoch": 0.24854768732995072,
+      "grad_norm": 1.9433478116989136,
+      "learning_rate": 0.00017599378802589873,
+      "loss": 0.4982,
+      "step": 3380
+    },
+    {
+      "epoch": 0.2486212221486874,
+      "grad_norm": 1.5820481777191162,
+      "learning_rate": 0.00017597830545435818,
+      "loss": 0.5126,
+      "step": 3381
+    },
+    {
+      "epoch": 0.24869475696742407,
+      "grad_norm": 1.4140586853027344,
+      "learning_rate": 0.00017596281857324797,
+      "loss": 0.4941,
+      "step": 3382
+    },
+    {
+      "epoch": 0.24876829178616075,
+      "grad_norm": 1.401703119277954,
+      "learning_rate": 0.00017594732738344653,
+      "loss": 0.5327,
+      "step": 3383
+    },
+    {
+      "epoch": 0.24884182660489743,
+      "grad_norm": 1.0257630348205566,
+      "learning_rate": 0.0001759318318858325,
+      "loss": 0.5201,
+      "step": 3384
+    },
+    {
+      "epoch": 0.2489153614236341,
+      "grad_norm": 1.2602800130844116,
+      "learning_rate": 0.00017591633208128483,
+      "loss": 0.4873,
+      "step": 3385
+    },
+    {
+      "epoch": 0.24898889624237075,
+      "grad_norm": 1.0674105882644653,
+      "learning_rate": 0.00017590082797068273,
+      "loss": 0.5172,
+      "step": 3386
+    },
+    {
+      "epoch": 0.24906243106110743,
+      "grad_norm": 1.1928852796554565,
+      "learning_rate": 0.00017588531955490554,
+      "loss": 0.5182,
+      "step": 3387
+    },
+    {
+      "epoch": 0.2491359658798441,
+      "grad_norm": 0.8238458037376404,
+      "learning_rate": 0.00017586980683483293,
+      "loss": 0.5316,
+      "step": 3388
+    },
+    {
+      "epoch": 0.24920950069858078,
+      "grad_norm": 1.1800771951675415,
+      "learning_rate": 0.00017585428981134484,
+      "loss": 0.5397,
+      "step": 3389
+    },
+    {
+      "epoch": 0.24928303551731745,
+      "grad_norm": 0.7805677056312561,
+      "learning_rate": 0.00017583876848532135,
+      "loss": 0.5284,
+      "step": 3390
+    },
+    {
+      "epoch": 0.24935657033605413,
+      "grad_norm": 0.9345836639404297,
+      "learning_rate": 0.00017582324285764288,
+      "loss": 0.5251,
+      "step": 3391
+    },
+    {
+      "epoch": 0.2494301051547908,
+      "grad_norm": 0.932178258895874,
+      "learning_rate": 0.00017580771292919004,
+      "loss": 0.5002,
+      "step": 3392
+    },
+    {
+      "epoch": 0.24950363997352745,
+      "grad_norm": 0.9584727883338928,
+      "learning_rate": 0.00017579217870084375,
+      "loss": 0.5164,
+      "step": 3393
+    },
+    {
+      "epoch": 0.24957717479226413,
+      "grad_norm": 0.795303463935852,
+      "learning_rate": 0.0001757766401734851,
+      "loss": 0.5113,
+      "step": 3394
+    },
+    {
+      "epoch": 0.2496507096110008,
+      "grad_norm": 1.0711016654968262,
+      "learning_rate": 0.00017576109734799541,
+      "loss": 0.5175,
+      "step": 3395
+    },
+    {
+      "epoch": 0.24972424442973748,
+      "grad_norm": 0.9599931836128235,
+      "learning_rate": 0.00017574555022525638,
+      "loss": 0.5184,
+      "step": 3396
+    },
+    {
+      "epoch": 0.24979777924847416,
+      "grad_norm": 0.7070573568344116,
+      "learning_rate": 0.00017572999880614978,
+      "loss": 0.4972,
+      "step": 3397
+    },
+    {
+      "epoch": 0.24987131406721083,
+      "grad_norm": 1.1555147171020508,
+      "learning_rate": 0.0001757144430915577,
+      "loss": 0.5036,
+      "step": 3398
+    },
+    {
+      "epoch": 0.2499448488859475,
+      "grad_norm": 0.8258602023124695,
+      "learning_rate": 0.00017569888308236254,
+      "loss": 0.5205,
+      "step": 3399
+    },
+    {
+      "epoch": 0.25001838370468416,
+      "grad_norm": 0.7488269805908203,
+      "learning_rate": 0.00017568331877944684,
+      "loss": 0.5201,
+      "step": 3400
+    },
+    {
+      "epoch": 0.25009191852342083,
+      "grad_norm": 0.7239822149276733,
+      "learning_rate": 0.0001756677501836934,
+      "loss": 0.5262,
+      "step": 3401
+    },
+    {
+      "epoch": 0.2501654533421575,
+      "grad_norm": 0.6919550895690918,
+      "learning_rate": 0.00017565217729598534,
+      "loss": 0.5212,
+      "step": 3402
+    },
+    {
+      "epoch": 0.2502389881608942,
+      "grad_norm": 0.7035146355628967,
+      "learning_rate": 0.00017563660011720594,
+      "loss": 0.4903,
+      "step": 3403
+    },
+    {
+      "epoch": 0.25031252297963086,
+      "grad_norm": 0.7395893931388855,
+      "learning_rate": 0.00017562101864823873,
+      "loss": 0.4943,
+      "step": 3404
+    },
+    {
+      "epoch": 0.25038605779836753,
+      "grad_norm": 0.7047514915466309,
+      "learning_rate": 0.0001756054328899676,
+      "loss": 0.4954,
+      "step": 3405
+    },
+    {
+      "epoch": 0.2504595926171042,
+      "grad_norm": 0.6527854800224304,
+      "learning_rate": 0.00017558984284327643,
+      "loss": 0.5221,
+      "step": 3406
+    },
+    {
+      "epoch": 0.2505331274358409,
+      "grad_norm": 0.7951874732971191,
+      "learning_rate": 0.0001755742485090497,
+      "loss": 0.5193,
+      "step": 3407
+    },
+    {
+      "epoch": 0.25060666225457756,
+      "grad_norm": 1.1609755754470825,
+      "learning_rate": 0.0001755586498881718,
+      "loss": 0.5012,
+      "step": 3408
+    },
+    {
+      "epoch": 0.25068019707331424,
+      "grad_norm": 0.7335014343261719,
+      "learning_rate": 0.0001755430469815275,
+      "loss": 0.5327,
+      "step": 3409
+    },
+    {
+      "epoch": 0.2507537318920509,
+      "grad_norm": 0.8500760197639465,
+      "learning_rate": 0.0001755274397900019,
+      "loss": 0.524,
+      "step": 3410
+    },
+    {
+      "epoch": 0.25082726671078753,
+      "grad_norm": 1.3866677284240723,
+      "learning_rate": 0.00017551182831448018,
+      "loss": 0.5185,
+      "step": 3411
+    },
+    {
+      "epoch": 0.2509008015295242,
+      "grad_norm": 0.645565390586853,
+      "learning_rate": 0.00017549621255584787,
+      "loss": 0.523,
+      "step": 3412
+    },
+    {
+      "epoch": 0.2509743363482609,
+      "grad_norm": 1.1158292293548584,
+      "learning_rate": 0.0001754805925149907,
+      "loss": 0.5335,
+      "step": 3413
+    },
+    {
+      "epoch": 0.25104787116699756,
+      "grad_norm": 1.1583325862884521,
+      "learning_rate": 0.00017546496819279472,
+      "loss": 0.5287,
+      "step": 3414
+    },
+    {
+      "epoch": 0.25112140598573424,
+      "grad_norm": 0.6475653052330017,
+      "learning_rate": 0.00017544933959014606,
+      "loss": 0.5371,
+      "step": 3415
+    },
+    {
+      "epoch": 0.2511949408044709,
+      "grad_norm": 1.1882866621017456,
+      "learning_rate": 0.00017543370670793126,
+      "loss": 0.5287,
+      "step": 3416
+    },
+    {
+      "epoch": 0.2512684756232076,
+      "grad_norm": 1.4753557443618774,
+      "learning_rate": 0.00017541806954703696,
+      "loss": 0.5378,
+      "step": 3417
+    },
+    {
+      "epoch": 0.25134201044194426,
+      "grad_norm": 0.6707854866981506,
+      "learning_rate": 0.0001754024281083502,
+      "loss": 0.519,
+      "step": 3418
+    },
+    {
+      "epoch": 0.25141554526068094,
+      "grad_norm": 1.2548067569732666,
+      "learning_rate": 0.00017538678239275812,
+      "loss": 0.5293,
+      "step": 3419
+    },
+    {
+      "epoch": 0.2514890800794176,
+      "grad_norm": 1.058680772781372,
+      "learning_rate": 0.0001753711324011482,
+      "loss": 0.538,
+      "step": 3420
+    },
+    {
+      "epoch": 0.2515626148981543,
+      "grad_norm": 0.7423090934753418,
+      "learning_rate": 0.0001753554781344081,
+      "loss": 0.5181,
+      "step": 3421
+    },
+    {
+      "epoch": 0.25163614971689097,
+      "grad_norm": 0.6148707270622253,
+      "learning_rate": 0.00017533981959342576,
+      "loss": 0.4941,
+      "step": 3422
+    },
+    {
+      "epoch": 0.25170968453562764,
+      "grad_norm": 0.5918654203414917,
+      "learning_rate": 0.00017532415677908936,
+      "loss": 0.5141,
+      "step": 3423
+    },
+    {
+      "epoch": 0.2517832193543643,
+      "grad_norm": 0.8030121922492981,
+      "learning_rate": 0.00017530848969228726,
+      "loss": 0.5044,
+      "step": 3424
+    },
+    {
+      "epoch": 0.25185675417310094,
+      "grad_norm": 1.0869699716567993,
+      "learning_rate": 0.00017529281833390817,
+      "loss": 0.5081,
+      "step": 3425
+    },
+    {
+      "epoch": 0.2519302889918376,
+      "grad_norm": 0.6675809621810913,
+      "learning_rate": 0.00017527714270484097,
+      "loss": 0.5355,
+      "step": 3426
+    },
+    {
+      "epoch": 0.2520038238105743,
+      "grad_norm": 0.9632925987243652,
+      "learning_rate": 0.00017526146280597477,
+      "loss": 0.5362,
+      "step": 3427
+    },
+    {
+      "epoch": 0.25207735862931097,
+      "grad_norm": 1.6322532892227173,
+      "learning_rate": 0.000175245778638199,
+      "loss": 0.532,
+      "step": 3428
+    },
+    {
+      "epoch": 0.25215089344804764,
+      "grad_norm": 0.8517298102378845,
+      "learning_rate": 0.00017523009020240323,
+      "loss": 0.5298,
+      "step": 3429
+    },
+    {
+      "epoch": 0.2522244282667843,
+      "grad_norm": 2.400578498840332,
+      "learning_rate": 0.00017521439749947738,
+      "loss": 0.521,
+      "step": 3430
+    },
+    {
+      "epoch": 0.252297963085521,
+      "grad_norm": 1.4775300025939941,
+      "learning_rate": 0.0001751987005303115,
+      "loss": 0.4905,
+      "step": 3431
+    },
+    {
+      "epoch": 0.25237149790425767,
+      "grad_norm": 2.430269241333008,
+      "learning_rate": 0.00017518299929579598,
+      "loss": 0.5217,
+      "step": 3432
+    },
+    {
+      "epoch": 0.25244503272299434,
+      "grad_norm": 2.501922130584717,
+      "learning_rate": 0.00017516729379682137,
+      "loss": 0.5316,
+      "step": 3433
+    },
+    {
+      "epoch": 0.252518567541731,
+      "grad_norm": 1.1419591903686523,
+      "learning_rate": 0.00017515158403427856,
+      "loss": 0.5226,
+      "step": 3434
+    },
+    {
+      "epoch": 0.2525921023604677,
+      "grad_norm": 1.2829303741455078,
+      "learning_rate": 0.00017513587000905857,
+      "loss": 0.4984,
+      "step": 3435
+    },
+    {
+      "epoch": 0.25266563717920437,
+      "grad_norm": 1.4749765396118164,
+      "learning_rate": 0.00017512015172205274,
+      "loss": 0.5378,
+      "step": 3436
+    },
+    {
+      "epoch": 0.25273917199794105,
+      "grad_norm": 0.9431778788566589,
+      "learning_rate": 0.00017510442917415263,
+      "loss": 0.5344,
+      "step": 3437
+    },
+    {
+      "epoch": 0.2528127068166777,
+      "grad_norm": 1.9229687452316284,
+      "learning_rate": 0.00017508870236625007,
+      "loss": 0.5226,
+      "step": 3438
+    },
+    {
+      "epoch": 0.25288624163541434,
+      "grad_norm": 1.1740612983703613,
+      "learning_rate": 0.00017507297129923702,
+      "loss": 0.5643,
+      "step": 3439
+    },
+    {
+      "epoch": 0.252959776454151,
+      "grad_norm": 2.323031425476074,
+      "learning_rate": 0.00017505723597400585,
+      "loss": 0.5434,
+      "step": 3440
+    },
+    {
+      "epoch": 0.2530333112728877,
+      "grad_norm": 2.251256227493286,
+      "learning_rate": 0.00017504149639144902,
+      "loss": 0.5211,
+      "step": 3441
+    },
+    {
+      "epoch": 0.25310684609162437,
+      "grad_norm": 1.0761462450027466,
+      "learning_rate": 0.00017502575255245933,
+      "loss": 0.5396,
+      "step": 3442
+    },
+    {
+      "epoch": 0.25318038091036105,
+      "grad_norm": 1.1328190565109253,
+      "learning_rate": 0.0001750100044579298,
+      "loss": 0.4961,
+      "step": 3443
+    },
+    {
+      "epoch": 0.2532539157290977,
+      "grad_norm": 1.250245213508606,
+      "learning_rate": 0.00017499425210875362,
+      "loss": 0.5136,
+      "step": 3444
+    },
+    {
+      "epoch": 0.2533274505478344,
+      "grad_norm": 0.9548202753067017,
+      "learning_rate": 0.00017497849550582434,
+      "loss": 0.5181,
+      "step": 3445
+    },
+    {
+      "epoch": 0.2534009853665711,
+      "grad_norm": 1.763776421546936,
+      "learning_rate": 0.00017496273465003563,
+      "loss": 0.528,
+      "step": 3446
+    },
+    {
+      "epoch": 0.25347452018530775,
+      "grad_norm": 1.2873989343643188,
+      "learning_rate": 0.00017494696954228152,
+      "loss": 0.5171,
+      "step": 3447
+    },
+    {
+      "epoch": 0.2535480550040444,
+      "grad_norm": 1.727494239807129,
+      "learning_rate": 0.0001749312001834562,
+      "loss": 0.5466,
+      "step": 3448
+    },
+    {
+      "epoch": 0.2536215898227811,
+      "grad_norm": 1.667781949043274,
+      "learning_rate": 0.00017491542657445413,
+      "loss": 0.549,
+      "step": 3449
+    },
+    {
+      "epoch": 0.2536951246415178,
+      "grad_norm": 1.3208001852035522,
+      "learning_rate": 0.00017489964871616998,
+      "loss": 0.5235,
+      "step": 3450
+    },
+    {
+      "epoch": 0.25376865946025445,
+      "grad_norm": 1.247423529624939,
+      "learning_rate": 0.00017488386660949876,
+      "loss": 0.5165,
+      "step": 3451
+    },
+    {
+      "epoch": 0.25384219427899113,
+      "grad_norm": 0.9477109909057617,
+      "learning_rate": 0.00017486808025533556,
+      "loss": 0.5097,
+      "step": 3452
+    },
+    {
+      "epoch": 0.25391572909772775,
+      "grad_norm": 1.4159281253814697,
+      "learning_rate": 0.00017485228965457588,
+      "loss": 0.5062,
+      "step": 3453
+    },
+    {
+      "epoch": 0.2539892639164644,
+      "grad_norm": 0.6525847315788269,
+      "learning_rate": 0.0001748364948081153,
+      "loss": 0.5077,
+      "step": 3454
+    },
+    {
+      "epoch": 0.2540627987352011,
+      "grad_norm": 1.5212113857269287,
+      "learning_rate": 0.00017482069571684975,
+      "loss": 0.5817,
+      "step": 3455
+    },
+    {
+      "epoch": 0.2541363335539378,
+      "grad_norm": 0.9377439618110657,
+      "learning_rate": 0.00017480489238167542,
+      "loss": 0.5385,
+      "step": 3456
+    },
+    {
+      "epoch": 0.25420986837267445,
+      "grad_norm": 1.2773314714431763,
+      "learning_rate": 0.0001747890848034886,
+      "loss": 0.5027,
+      "step": 3457
+    },
+    {
+      "epoch": 0.25428340319141113,
+      "grad_norm": 1.024880290031433,
+      "learning_rate": 0.000174773272983186,
+      "loss": 0.5429,
+      "step": 3458
+    },
+    {
+      "epoch": 0.2543569380101478,
+      "grad_norm": 1.4136419296264648,
+      "learning_rate": 0.0001747574569216644,
+      "loss": 0.5235,
+      "step": 3459
+    },
+    {
+      "epoch": 0.2544304728288845,
+      "grad_norm": 1.1413484811782837,
+      "learning_rate": 0.000174741636619821,
+      "loss": 0.5209,
+      "step": 3460
+    },
+    {
+      "epoch": 0.25450400764762116,
+      "grad_norm": 1.4068725109100342,
+      "learning_rate": 0.00017472581207855306,
+      "loss": 0.4692,
+      "step": 3461
+    },
+    {
+      "epoch": 0.25457754246635783,
+      "grad_norm": 1.433855414390564,
+      "learning_rate": 0.0001747099832987582,
+      "loss": 0.5464,
+      "step": 3462
+    },
+    {
+      "epoch": 0.2546510772850945,
+      "grad_norm": 1.278591513633728,
+      "learning_rate": 0.00017469415028133426,
+      "loss": 0.498,
+      "step": 3463
+    },
+    {
+      "epoch": 0.2547246121038312,
+      "grad_norm": 1.2928515672683716,
+      "learning_rate": 0.00017467831302717926,
+      "loss": 0.5242,
+      "step": 3464
+    },
+    {
+      "epoch": 0.25479814692256786,
+      "grad_norm": 0.8344340324401855,
+      "learning_rate": 0.00017466247153719152,
+      "loss": 0.5159,
+      "step": 3465
+    },
+    {
+      "epoch": 0.25487168174130453,
+      "grad_norm": 1.4766181707382202,
+      "learning_rate": 0.00017464662581226964,
+      "loss": 0.5083,
+      "step": 3466
+    },
+    {
+      "epoch": 0.25494521656004115,
+      "grad_norm": 0.5883437395095825,
+      "learning_rate": 0.00017463077585331237,
+      "loss": 0.5304,
+      "step": 3467
+    },
+    {
+      "epoch": 0.25501875137877783,
+      "grad_norm": 1.1766592264175415,
+      "learning_rate": 0.00017461492166121875,
+      "loss": 0.5164,
+      "step": 3468
+    },
+    {
+      "epoch": 0.2550922861975145,
+      "grad_norm": 0.6388810276985168,
+      "learning_rate": 0.00017459906323688798,
+      "loss": 0.4937,
+      "step": 3469
+    },
+    {
+      "epoch": 0.2551658210162512,
+      "grad_norm": 0.8314344882965088,
+      "learning_rate": 0.00017458320058121962,
+      "loss": 0.5017,
+      "step": 3470
+    },
+    {
+      "epoch": 0.25523935583498786,
+      "grad_norm": 1.4428339004516602,
+      "learning_rate": 0.0001745673336951134,
+      "loss": 0.4921,
+      "step": 3471
+    },
+    {
+      "epoch": 0.25531289065372453,
+      "grad_norm": 0.6448944807052612,
+      "learning_rate": 0.0001745514625794694,
+      "loss": 0.5076,
+      "step": 3472
+    },
+    {
+      "epoch": 0.2553864254724612,
+      "grad_norm": 1.9771407842636108,
+      "learning_rate": 0.0001745355872351877,
+      "loss": 0.5229,
+      "step": 3473
+    },
+    {
+      "epoch": 0.2554599602911979,
+      "grad_norm": 0.7750949263572693,
+      "learning_rate": 0.00017451970766316883,
+      "loss": 0.5042,
+      "step": 3474
+    },
+    {
+      "epoch": 0.25553349510993456,
+      "grad_norm": 2.5731561183929443,
+      "learning_rate": 0.00017450382386431353,
+      "loss": 0.5156,
+      "step": 3475
+    },
+    {
+      "epoch": 0.25560702992867124,
+      "grad_norm": 1.3133890628814697,
+      "learning_rate": 0.00017448793583952268,
+      "loss": 0.5408,
+      "step": 3476
+    },
+    {
+      "epoch": 0.2556805647474079,
+      "grad_norm": 2.4682252407073975,
+      "learning_rate": 0.0001744720435896975,
+      "loss": 0.5427,
+      "step": 3477
+    },
+    {
+      "epoch": 0.2557540995661446,
+      "grad_norm": 2.421630859375,
+      "learning_rate": 0.00017445614711573943,
+      "loss": 0.5489,
+      "step": 3478
+    },
+    {
+      "epoch": 0.25582763438488126,
+      "grad_norm": 0.9659860134124756,
+      "learning_rate": 0.00017444024641855013,
+      "loss": 0.5334,
+      "step": 3479
+    },
+    {
+      "epoch": 0.25590116920361794,
+      "grad_norm": 2.021029233932495,
+      "learning_rate": 0.00017442434149903145,
+      "loss": 0.5541,
+      "step": 3480
+    },
+    {
+      "epoch": 0.25597470402235456,
+      "grad_norm": 1.233957290649414,
+      "learning_rate": 0.00017440843235808564,
+      "loss": 0.5411,
+      "step": 3481
+    },
+    {
+      "epoch": 0.25604823884109124,
+      "grad_norm": 1.4962365627288818,
+      "learning_rate": 0.00017439251899661494,
+      "loss": 0.5393,
+      "step": 3482
+    },
+    {
+      "epoch": 0.2561217736598279,
+      "grad_norm": 1.2457542419433594,
+      "learning_rate": 0.00017437660141552213,
+      "loss": 0.535,
+      "step": 3483
+    },
+    {
+      "epoch": 0.2561953084785646,
+      "grad_norm": 1.4313403367996216,
+      "learning_rate": 0.00017436067961570996,
+      "loss": 0.507,
+      "step": 3484
+    },
+    {
+      "epoch": 0.25626884329730126,
+      "grad_norm": 1.5272780656814575,
+      "learning_rate": 0.00017434475359808156,
+      "loss": 0.5242,
+      "step": 3485
+    },
+    {
+      "epoch": 0.25634237811603794,
+      "grad_norm": 0.8657505512237549,
+      "learning_rate": 0.00017432882336354029,
+      "loss": 0.5022,
+      "step": 3486
+    },
+    {
+      "epoch": 0.2564159129347746,
+      "grad_norm": 1.9071382284164429,
+      "learning_rate": 0.00017431288891298972,
+      "loss": 0.5129,
+      "step": 3487
+    },
+    {
+      "epoch": 0.2564894477535113,
+      "grad_norm": 1.485072135925293,
+      "learning_rate": 0.00017429695024733365,
+      "loss": 0.5189,
+      "step": 3488
+    },
+    {
+      "epoch": 0.25656298257224797,
+      "grad_norm": 1.2884128093719482,
+      "learning_rate": 0.00017428100736747616,
+      "loss": 0.5119,
+      "step": 3489
+    },
+    {
+      "epoch": 0.25663651739098464,
+      "grad_norm": 1.505125880241394,
+      "learning_rate": 0.00017426506027432156,
+      "loss": 0.5206,
+      "step": 3490
+    },
+    {
+      "epoch": 0.2567100522097213,
+      "grad_norm": 1.0161021947860718,
+      "learning_rate": 0.00017424910896877434,
+      "loss": 0.5555,
+      "step": 3491
+    },
+    {
+      "epoch": 0.256783587028458,
+      "grad_norm": 1.1017752885818481,
+      "learning_rate": 0.00017423315345173928,
+      "loss": 0.5163,
+      "step": 3492
+    },
+    {
+      "epoch": 0.25685712184719467,
+      "grad_norm": 0.9360860586166382,
+      "learning_rate": 0.00017421719372412146,
+      "loss": 0.5382,
+      "step": 3493
+    },
+    {
+      "epoch": 0.25693065666593135,
+      "grad_norm": 0.9993916153907776,
+      "learning_rate": 0.00017420122978682606,
+      "loss": 0.517,
+      "step": 3494
+    },
+    {
+      "epoch": 0.25700419148466797,
+      "grad_norm": 0.6693781018257141,
+      "learning_rate": 0.00017418526164075863,
+      "loss": 0.5025,
+      "step": 3495
+    },
+    {
+      "epoch": 0.25707772630340464,
+      "grad_norm": 1.0659352540969849,
+      "learning_rate": 0.00017416928928682483,
+      "loss": 0.5249,
+      "step": 3496
+    },
+    {
+      "epoch": 0.2571512611221413,
+      "grad_norm": 0.7521973848342896,
+      "learning_rate": 0.0001741533127259307,
+      "loss": 0.5245,
+      "step": 3497
+    },
+    {
+      "epoch": 0.257224795940878,
+      "grad_norm": 1.171413779258728,
+      "learning_rate": 0.0001741373319589824,
+      "loss": 0.4806,
+      "step": 3498
+    },
+    {
+      "epoch": 0.25729833075961467,
+      "grad_norm": 0.6625381708145142,
+      "learning_rate": 0.0001741213469868864,
+      "loss": 0.5089,
+      "step": 3499
+    },
+    {
+      "epoch": 0.25737186557835134,
+      "grad_norm": 0.9330199360847473,
+      "learning_rate": 0.00017410535781054937,
+      "loss": 0.5468,
+      "step": 3500
+    },
+    {
+      "epoch": 0.257445400397088,
+      "grad_norm": 0.8278741836547852,
+      "learning_rate": 0.0001740893644308782,
+      "loss": 0.4894,
+      "step": 3501
+    },
+    {
+      "epoch": 0.2575189352158247,
+      "grad_norm": 0.8294281959533691,
+      "learning_rate": 0.00017407336684878018,
+      "loss": 0.5054,
+      "step": 3502
+    },
+    {
+      "epoch": 0.25759247003456137,
+      "grad_norm": 0.8835468292236328,
+      "learning_rate": 0.00017405736506516254,
+      "loss": 0.5147,
+      "step": 3503
+    },
+    {
+      "epoch": 0.25766600485329805,
+      "grad_norm": 0.7915424704551697,
+      "learning_rate": 0.000174041359080933,
+      "loss": 0.511,
+      "step": 3504
+    },
+    {
+      "epoch": 0.2577395396720347,
+      "grad_norm": 0.58710777759552,
+      "learning_rate": 0.0001740253488969995,
+      "loss": 0.5136,
+      "step": 3505
+    },
+    {
+      "epoch": 0.2578130744907714,
+      "grad_norm": 0.948517918586731,
+      "learning_rate": 0.00017400933451427002,
+      "loss": 0.5429,
+      "step": 3506
+    },
+    {
+      "epoch": 0.2578866093095081,
+      "grad_norm": 0.9208300113677979,
+      "learning_rate": 0.00017399331593365302,
+      "loss": 0.5387,
+      "step": 3507
+    },
+    {
+      "epoch": 0.25796014412824475,
+      "grad_norm": 0.549121081829071,
+      "learning_rate": 0.00017397729315605704,
+      "loss": 0.5698,
+      "step": 3508
+    },
+    {
+      "epoch": 0.25803367894698137,
+      "grad_norm": 0.79099440574646,
+      "learning_rate": 0.00017396126618239092,
+      "loss": 0.5493,
+      "step": 3509
+    },
+    {
+      "epoch": 0.25810721376571805,
+      "grad_norm": 0.5916030406951904,
+      "learning_rate": 0.00017394523501356373,
+      "loss": 0.5253,
+      "step": 3510
+    },
+    {
+      "epoch": 0.2581807485844547,
+      "grad_norm": 0.9105160236358643,
+      "learning_rate": 0.00017392919965048478,
+      "loss": 0.5084,
+      "step": 3511
+    },
+    {
+      "epoch": 0.2582542834031914,
+      "grad_norm": 1.1867690086364746,
+      "learning_rate": 0.0001739131600940636,
+      "loss": 0.5184,
+      "step": 3512
+    },
+    {
+      "epoch": 0.2583278182219281,
+      "grad_norm": 0.748284101486206,
+      "learning_rate": 0.00017389711634520997,
+      "loss": 0.5232,
+      "step": 3513
+    },
+    {
+      "epoch": 0.25840135304066475,
+      "grad_norm": 0.7308588624000549,
+      "learning_rate": 0.00017388106840483388,
+      "loss": 0.514,
+      "step": 3514
+    },
+    {
+      "epoch": 0.2584748878594014,
+      "grad_norm": 1.220163106918335,
+      "learning_rate": 0.00017386501627384568,
+      "loss": 0.5212,
+      "step": 3515
+    },
+    {
+      "epoch": 0.2585484226781381,
+      "grad_norm": 0.7186478972434998,
+      "learning_rate": 0.00017384895995315578,
+      "loss": 0.4995,
+      "step": 3516
+    },
+    {
+      "epoch": 0.2586219574968748,
+      "grad_norm": 0.7959068417549133,
+      "learning_rate": 0.00017383289944367493,
+      "loss": 0.5217,
+      "step": 3517
+    },
+    {
+      "epoch": 0.25869549231561145,
+      "grad_norm": 1.1476439237594604,
+      "learning_rate": 0.00017381683474631412,
+      "loss": 0.5142,
+      "step": 3518
+    },
+    {
+      "epoch": 0.25876902713434813,
+      "grad_norm": 0.6210100650787354,
+      "learning_rate": 0.00017380076586198453,
+      "loss": 0.5105,
+      "step": 3519
+    },
+    {
+      "epoch": 0.2588425619530848,
+      "grad_norm": 0.6192209720611572,
+      "learning_rate": 0.00017378469279159762,
+      "loss": 0.5383,
+      "step": 3520
+    },
+    {
+      "epoch": 0.2589160967718215,
+      "grad_norm": 0.8168628811836243,
+      "learning_rate": 0.00017376861553606508,
+      "loss": 0.5139,
+      "step": 3521
+    },
+    {
+      "epoch": 0.25898963159055816,
+      "grad_norm": 0.8125194907188416,
+      "learning_rate": 0.0001737525340962988,
+      "loss": 0.506,
+      "step": 3522
+    },
+    {
+      "epoch": 0.2590631664092948,
+      "grad_norm": 0.5424176454544067,
+      "learning_rate": 0.00017373644847321092,
+      "loss": 0.5041,
+      "step": 3523
+    },
+    {
+      "epoch": 0.25913670122803145,
+      "grad_norm": 0.9247700572013855,
+      "learning_rate": 0.0001737203586677139,
+      "loss": 0.5253,
+      "step": 3524
+    },
+    {
+      "epoch": 0.2592102360467681,
+      "grad_norm": 0.8271786570549011,
+      "learning_rate": 0.00017370426468072035,
+      "loss": 0.4957,
+      "step": 3525
+    },
+    {
+      "epoch": 0.2592837708655048,
+      "grad_norm": 0.887243926525116,
+      "learning_rate": 0.0001736881665131431,
+      "loss": 0.5037,
+      "step": 3526
+    },
+    {
+      "epoch": 0.2593573056842415,
+      "grad_norm": 1.1669423580169678,
+      "learning_rate": 0.00017367206416589527,
+      "loss": 0.4971,
+      "step": 3527
+    },
+    {
+      "epoch": 0.25943084050297816,
+      "grad_norm": 0.8345411419868469,
+      "learning_rate": 0.00017365595763989026,
+      "loss": 0.5075,
+      "step": 3528
+    },
+    {
+      "epoch": 0.25950437532171483,
+      "grad_norm": 0.6732757687568665,
+      "learning_rate": 0.00017363984693604154,
+      "loss": 0.5171,
+      "step": 3529
+    },
+    {
+      "epoch": 0.2595779101404515,
+      "grad_norm": 1.4490530490875244,
+      "learning_rate": 0.00017362373205526304,
+      "loss": 0.5274,
+      "step": 3530
+    },
+    {
+      "epoch": 0.2596514449591882,
+      "grad_norm": 0.7624437212944031,
+      "learning_rate": 0.00017360761299846874,
+      "loss": 0.5318,
+      "step": 3531
+    },
+    {
+      "epoch": 0.25972497977792486,
+      "grad_norm": 0.8034105896949768,
+      "learning_rate": 0.0001735914897665729,
+      "loss": 0.4904,
+      "step": 3532
+    },
+    {
+      "epoch": 0.25979851459666153,
+      "grad_norm": 1.574925422668457,
+      "learning_rate": 0.00017357536236049014,
+      "loss": 0.4935,
+      "step": 3533
+    },
+    {
+      "epoch": 0.2598720494153982,
+      "grad_norm": 0.6203644871711731,
+      "learning_rate": 0.00017355923078113522,
+      "loss": 0.4993,
+      "step": 3534
+    },
+    {
+      "epoch": 0.2599455842341349,
+      "grad_norm": 1.5950663089752197,
+      "learning_rate": 0.00017354309502942304,
+      "loss": 0.5521,
+      "step": 3535
+    },
+    {
+      "epoch": 0.26001911905287156,
+      "grad_norm": 0.6820402145385742,
+      "learning_rate": 0.00017352695510626891,
+      "loss": 0.5338,
+      "step": 3536
+    },
+    {
+      "epoch": 0.2600926538716082,
+      "grad_norm": 1.5801585912704468,
+      "learning_rate": 0.00017351081101258831,
+      "loss": 0.5036,
+      "step": 3537
+    },
+    {
+      "epoch": 0.26016618869034486,
+      "grad_norm": 0.7038335204124451,
+      "learning_rate": 0.0001734946627492969,
+      "loss": 0.5233,
+      "step": 3538
+    },
+    {
+      "epoch": 0.26023972350908153,
+      "grad_norm": 1.241039752960205,
+      "learning_rate": 0.00017347851031731067,
+      "loss": 0.5011,
+      "step": 3539
+    },
+    {
+      "epoch": 0.2603132583278182,
+      "grad_norm": 1.2365622520446777,
+      "learning_rate": 0.0001734623537175458,
+      "loss": 0.4982,
+      "step": 3540
+    },
+    {
+      "epoch": 0.2603867931465549,
+      "grad_norm": 0.7523282766342163,
+      "learning_rate": 0.0001734461929509187,
+      "loss": 0.502,
+      "step": 3541
+    },
+    {
+      "epoch": 0.26046032796529156,
+      "grad_norm": 0.6157302856445312,
+      "learning_rate": 0.00017343002801834602,
+      "loss": 0.5006,
+      "step": 3542
+    },
+    {
+      "epoch": 0.26053386278402824,
+      "grad_norm": 0.6891542077064514,
+      "learning_rate": 0.00017341385892074467,
+      "loss": 0.5233,
+      "step": 3543
+    },
+    {
+      "epoch": 0.2606073976027649,
+      "grad_norm": 0.5945039987564087,
+      "learning_rate": 0.00017339768565903174,
+      "loss": 0.5298,
+      "step": 3544
+    },
+    {
+      "epoch": 0.2606809324215016,
+      "grad_norm": 0.6684527397155762,
+      "learning_rate": 0.0001733815082341246,
+      "loss": 0.514,
+      "step": 3545
+    },
+    {
+      "epoch": 0.26075446724023826,
+      "grad_norm": 0.747336208820343,
+      "learning_rate": 0.00017336532664694092,
+      "loss": 0.4893,
+      "step": 3546
+    },
+    {
+      "epoch": 0.26082800205897494,
+      "grad_norm": 0.7327359914779663,
+      "learning_rate": 0.00017334914089839847,
+      "loss": 0.5311,
+      "step": 3547
+    },
+    {
+      "epoch": 0.2609015368777116,
+      "grad_norm": 0.8918739557266235,
+      "learning_rate": 0.00017333295098941534,
+      "loss": 0.4861,
+      "step": 3548
+    },
+    {
+      "epoch": 0.2609750716964483,
+      "grad_norm": 1.3485279083251953,
+      "learning_rate": 0.00017331675692090984,
+      "loss": 0.5431,
+      "step": 3549
+    },
+    {
+      "epoch": 0.26104860651518497,
+      "grad_norm": 0.6665927767753601,
+      "learning_rate": 0.0001733005586938005,
+      "loss": 0.5076,
+      "step": 3550
+    },
+    {
+      "epoch": 0.2611221413339216,
+      "grad_norm": 1.0571483373641968,
+      "learning_rate": 0.00017328435630900612,
+      "loss": 0.5302,
+      "step": 3551
+    },
+    {
+      "epoch": 0.26119567615265826,
+      "grad_norm": 1.3395870923995972,
+      "learning_rate": 0.00017326814976744573,
+      "loss": 0.5096,
+      "step": 3552
+    },
+    {
+      "epoch": 0.26126921097139494,
+      "grad_norm": 0.7181114554405212,
+      "learning_rate": 0.00017325193907003853,
+      "loss": 0.5046,
+      "step": 3553
+    },
+    {
+      "epoch": 0.2613427457901316,
+      "grad_norm": 1.1557368040084839,
+      "learning_rate": 0.00017323572421770405,
+      "loss": 0.5018,
+      "step": 3554
+    },
+    {
+      "epoch": 0.2614162806088683,
+      "grad_norm": 1.404931902885437,
+      "learning_rate": 0.000173219505211362,
+      "loss": 0.5026,
+      "step": 3555
+    },
+    {
+      "epoch": 0.26148981542760497,
+      "grad_norm": 0.7508135437965393,
+      "learning_rate": 0.00017320328205193235,
+      "loss": 0.4952,
+      "step": 3556
+    },
+    {
+      "epoch": 0.26156335024634164,
+      "grad_norm": 0.8352027535438538,
+      "learning_rate": 0.00017318705474033533,
+      "loss": 0.5124,
+      "step": 3557
+    },
+    {
+      "epoch": 0.2616368850650783,
+      "grad_norm": 1.0784698724746704,
+      "learning_rate": 0.00017317082327749126,
+      "loss": 0.5541,
+      "step": 3558
+    },
+    {
+      "epoch": 0.261710419883815,
+      "grad_norm": 1.0144262313842773,
+      "learning_rate": 0.0001731545876643209,
+      "loss": 0.5249,
+      "step": 3559
+    },
+    {
+      "epoch": 0.26178395470255167,
+      "grad_norm": 0.8031282424926758,
+      "learning_rate": 0.00017313834790174513,
+      "loss": 0.4963,
+      "step": 3560
+    },
+    {
+      "epoch": 0.26185748952128834,
+      "grad_norm": 0.6631783843040466,
+      "learning_rate": 0.00017312210399068507,
+      "loss": 0.5341,
+      "step": 3561
+    },
+    {
+      "epoch": 0.261931024340025,
+      "grad_norm": 0.704338014125824,
+      "learning_rate": 0.0001731058559320621,
+      "loss": 0.5168,
+      "step": 3562
+    },
+    {
+      "epoch": 0.2620045591587617,
+      "grad_norm": 0.7107346057891846,
+      "learning_rate": 0.00017308960372679785,
+      "loss": 0.5144,
+      "step": 3563
+    },
+    {
+      "epoch": 0.26207809397749837,
+      "grad_norm": 0.5262987017631531,
+      "learning_rate": 0.00017307334737581414,
+      "loss": 0.5026,
+      "step": 3564
+    },
+    {
+      "epoch": 0.262151628796235,
+      "grad_norm": 0.60494065284729,
+      "learning_rate": 0.00017305708688003301,
+      "loss": 0.5074,
+      "step": 3565
+    },
+    {
+      "epoch": 0.26222516361497167,
+      "grad_norm": 0.7075631022453308,
+      "learning_rate": 0.00017304082224037684,
+      "loss": 0.5028,
+      "step": 3566
+    },
+    {
+      "epoch": 0.26229869843370834,
+      "grad_norm": 0.6588335633277893,
+      "learning_rate": 0.00017302455345776817,
+      "loss": 0.5317,
+      "step": 3567
+    },
+    {
+      "epoch": 0.262372233252445,
+      "grad_norm": 0.8539736270904541,
+      "learning_rate": 0.00017300828053312976,
+      "loss": 0.4903,
+      "step": 3568
+    },
+    {
+      "epoch": 0.2624457680711817,
+      "grad_norm": 1.0006237030029297,
+      "learning_rate": 0.0001729920034673846,
+      "loss": 0.5149,
+      "step": 3569
+    },
+    {
+      "epoch": 0.26251930288991837,
+      "grad_norm": 2.360678195953369,
+      "learning_rate": 0.000172975722261456,
+      "loss": 0.5155,
+      "step": 3570
+    },
+    {
+      "epoch": 0.26259283770865505,
+      "grad_norm": 1.6600347757339478,
+      "learning_rate": 0.0001729594369162674,
+      "loss": 0.5256,
+      "step": 3571
+    },
+    {
+      "epoch": 0.2626663725273917,
+      "grad_norm": 2.3442556858062744,
+      "learning_rate": 0.00017294314743274253,
+      "loss": 0.514,
+      "step": 3572
+    },
+    {
+      "epoch": 0.2627399073461284,
+      "grad_norm": 2.1359944343566895,
+      "learning_rate": 0.0001729268538118054,
+      "loss": 0.5071,
+      "step": 3573
+    },
+    {
+      "epoch": 0.2628134421648651,
+      "grad_norm": 1.6370906829833984,
+      "learning_rate": 0.0001729105560543801,
+      "loss": 0.518,
+      "step": 3574
+    },
+    {
+      "epoch": 0.26288697698360175,
+      "grad_norm": 1.2672994136810303,
+      "learning_rate": 0.00017289425416139117,
+      "loss": 0.5198,
+      "step": 3575
+    },
+    {
+      "epoch": 0.2629605118023384,
+      "grad_norm": 2.3371241092681885,
+      "learning_rate": 0.0001728779481337632,
+      "loss": 0.5315,
+      "step": 3576
+    },
+    {
+      "epoch": 0.2630340466210751,
+      "grad_norm": 1.8448907136917114,
+      "learning_rate": 0.00017286163797242108,
+      "loss": 0.514,
+      "step": 3577
+    },
+    {
+      "epoch": 0.2631075814398118,
+      "grad_norm": 1.9943821430206299,
+      "learning_rate": 0.00017284532367828996,
+      "loss": 0.4932,
+      "step": 3578
+    },
+    {
+      "epoch": 0.2631811162585484,
+      "grad_norm": 1.8963559865951538,
+      "learning_rate": 0.00017282900525229523,
+      "loss": 0.5152,
+      "step": 3579
+    },
+    {
+      "epoch": 0.2632546510772851,
+      "grad_norm": 1.1776231527328491,
+      "learning_rate": 0.00017281268269536243,
+      "loss": 0.517,
+      "step": 3580
+    },
+    {
+      "epoch": 0.26332818589602175,
+      "grad_norm": 1.3109608888626099,
+      "learning_rate": 0.00017279635600841745,
+      "loss": 0.5021,
+      "step": 3581
+    },
+    {
+      "epoch": 0.2634017207147584,
+      "grad_norm": 1.959905982017517,
+      "learning_rate": 0.00017278002519238635,
+      "loss": 0.5274,
+      "step": 3582
+    },
+    {
+      "epoch": 0.2634752555334951,
+      "grad_norm": 1.5371620655059814,
+      "learning_rate": 0.00017276369024819536,
+      "loss": 0.5173,
+      "step": 3583
+    },
+    {
+      "epoch": 0.2635487903522318,
+      "grad_norm": 2.235602378845215,
+      "learning_rate": 0.00017274735117677113,
+      "loss": 0.5394,
+      "step": 3584
+    },
+    {
+      "epoch": 0.26362232517096845,
+      "grad_norm": 2.1092369556427,
+      "learning_rate": 0.00017273100797904033,
+      "loss": 0.4987,
+      "step": 3585
+    },
+    {
+      "epoch": 0.26369585998970513,
+      "grad_norm": 1.4809378385543823,
+      "learning_rate": 0.00017271466065593,
+      "loss": 0.5559,
+      "step": 3586
+    },
+    {
+      "epoch": 0.2637693948084418,
+      "grad_norm": 1.256440281867981,
+      "learning_rate": 0.00017269830920836737,
+      "loss": 0.5603,
+      "step": 3587
+    },
+    {
+      "epoch": 0.2638429296271785,
+      "grad_norm": 1.7470515966415405,
+      "learning_rate": 0.00017268195363727995,
+      "loss": 0.5127,
+      "step": 3588
+    },
+    {
+      "epoch": 0.26391646444591516,
+      "grad_norm": 1.297602891921997,
+      "learning_rate": 0.0001726655939435954,
+      "loss": 0.5167,
+      "step": 3589
+    },
+    {
+      "epoch": 0.26398999926465183,
+      "grad_norm": 1.9560805559158325,
+      "learning_rate": 0.0001726492301282417,
+      "loss": 0.5353,
+      "step": 3590
+    },
+    {
+      "epoch": 0.2640635340833885,
+      "grad_norm": 1.6032581329345703,
+      "learning_rate": 0.00017263286219214695,
+      "loss": 0.5263,
+      "step": 3591
+    },
+    {
+      "epoch": 0.2641370689021252,
+      "grad_norm": 1.6641172170639038,
+      "learning_rate": 0.00017261649013623965,
+      "loss": 0.5112,
+      "step": 3592
+    },
+    {
+      "epoch": 0.2642106037208618,
+      "grad_norm": 1.4870797395706177,
+      "learning_rate": 0.00017260011396144837,
+      "loss": 0.5243,
+      "step": 3593
+    },
+    {
+      "epoch": 0.2642841385395985,
+      "grad_norm": 1.6493858098983765,
+      "learning_rate": 0.000172583733668702,
+      "loss": 0.5498,
+      "step": 3594
+    },
+    {
+      "epoch": 0.26435767335833515,
+      "grad_norm": 1.3164174556732178,
+      "learning_rate": 0.00017256734925892966,
+      "loss": 0.5406,
+      "step": 3595
+    },
+    {
+      "epoch": 0.26443120817707183,
+      "grad_norm": 1.9039572477340698,
+      "learning_rate": 0.0001725509607330607,
+      "loss": 0.5402,
+      "step": 3596
+    },
+    {
+      "epoch": 0.2645047429958085,
+      "grad_norm": 1.3655743598937988,
+      "learning_rate": 0.0001725345680920247,
+      "loss": 0.5229,
+      "step": 3597
+    },
+    {
+      "epoch": 0.2645782778145452,
+      "grad_norm": 1.840683937072754,
+      "learning_rate": 0.0001725181713367514,
+      "loss": 0.5277,
+      "step": 3598
+    },
+    {
+      "epoch": 0.26465181263328186,
+      "grad_norm": 1.1958210468292236,
+      "learning_rate": 0.00017250177046817093,
+      "loss": 0.5184,
+      "step": 3599
+    },
+    {
+      "epoch": 0.26472534745201853,
+      "grad_norm": 1.7613320350646973,
+      "learning_rate": 0.0001724853654872135,
+      "loss": 0.5453,
+      "step": 3600
+    },
+    {
+      "epoch": 0.2647988822707552,
+      "grad_norm": 1.1395760774612427,
+      "learning_rate": 0.00017246895639480965,
+      "loss": 0.4966,
+      "step": 3601
+    },
+    {
+      "epoch": 0.2648724170894919,
+      "grad_norm": 1.8483167886734009,
+      "learning_rate": 0.00017245254319189012,
+      "loss": 0.5045,
+      "step": 3602
+    },
+    {
+      "epoch": 0.26494595190822856,
+      "grad_norm": 1.420143961906433,
+      "learning_rate": 0.00017243612587938584,
+      "loss": 0.4965,
+      "step": 3603
+    },
+    {
+      "epoch": 0.26501948672696524,
+      "grad_norm": 2.0879595279693604,
+      "learning_rate": 0.00017241970445822812,
+      "loss": 0.5106,
+      "step": 3604
+    },
+    {
+      "epoch": 0.2650930215457019,
+      "grad_norm": 1.8127676248550415,
+      "learning_rate": 0.0001724032789293483,
+      "loss": 0.5105,
+      "step": 3605
+    },
+    {
+      "epoch": 0.2651665563644386,
+      "grad_norm": 1.6685919761657715,
+      "learning_rate": 0.00017238684929367807,
+      "loss": 0.5558,
+      "step": 3606
+    },
+    {
+      "epoch": 0.2652400911831752,
+      "grad_norm": 1.3782858848571777,
+      "learning_rate": 0.0001723704155521494,
+      "loss": 0.5055,
+      "step": 3607
+    },
+    {
+      "epoch": 0.2653136260019119,
+      "grad_norm": 1.419631004333496,
+      "learning_rate": 0.00017235397770569435,
+      "loss": 0.5164,
+      "step": 3608
+    },
+    {
+      "epoch": 0.26538716082064856,
+      "grad_norm": 0.9559959173202515,
+      "learning_rate": 0.00017233753575524534,
+      "loss": 0.5208,
+      "step": 3609
+    },
+    {
+      "epoch": 0.26546069563938524,
+      "grad_norm": 1.6700557470321655,
+      "learning_rate": 0.00017232108970173496,
+      "loss": 0.5398,
+      "step": 3610
+    },
+    {
+      "epoch": 0.2655342304581219,
+      "grad_norm": 1.0171443223953247,
+      "learning_rate": 0.00017230463954609606,
+      "loss": 0.5255,
+      "step": 3611
+    },
+    {
+      "epoch": 0.2656077652768586,
+      "grad_norm": 2.163102149963379,
+      "learning_rate": 0.00017228818528926168,
+      "loss": 0.5234,
+      "step": 3612
+    },
+    {
+      "epoch": 0.26568130009559526,
+      "grad_norm": 1.6825686693191528,
+      "learning_rate": 0.00017227172693216514,
+      "loss": 0.5256,
+      "step": 3613
+    },
+    {
+      "epoch": 0.26575483491433194,
+      "grad_norm": 2.0074381828308105,
+      "learning_rate": 0.00017225526447573998,
+      "loss": 0.5267,
+      "step": 3614
+    },
+    {
+      "epoch": 0.2658283697330686,
+      "grad_norm": 1.7167367935180664,
+      "learning_rate": 0.00017223879792092,
+      "loss": 0.537,
+      "step": 3615
+    },
+    {
+      "epoch": 0.2659019045518053,
+      "grad_norm": 1.5490154027938843,
+      "learning_rate": 0.00017222232726863911,
+      "loss": 0.5355,
+      "step": 3616
+    },
+    {
+      "epoch": 0.26597543937054197,
+      "grad_norm": 1.3974647521972656,
+      "learning_rate": 0.00017220585251983165,
+      "loss": 0.517,
+      "step": 3617
+    },
+    {
+      "epoch": 0.26604897418927864,
+      "grad_norm": 1.2915502786636353,
+      "learning_rate": 0.000172189373675432,
+      "loss": 0.5357,
+      "step": 3618
+    },
+    {
+      "epoch": 0.2661225090080153,
+      "grad_norm": 1.0747658014297485,
+      "learning_rate": 0.00017217289073637488,
+      "loss": 0.5309,
+      "step": 3619
+    },
+    {
+      "epoch": 0.266196043826752,
+      "grad_norm": 1.4816964864730835,
+      "learning_rate": 0.00017215640370359523,
+      "loss": 0.4932,
+      "step": 3620
+    },
+    {
+      "epoch": 0.2662695786454886,
+      "grad_norm": 0.855053186416626,
+      "learning_rate": 0.00017213991257802825,
+      "loss": 0.5231,
+      "step": 3621
+    },
+    {
+      "epoch": 0.2663431134642253,
+      "grad_norm": 1.1406821012496948,
+      "learning_rate": 0.00017212341736060928,
+      "loss": 0.5291,
+      "step": 3622
+    },
+    {
+      "epoch": 0.26641664828296197,
+      "grad_norm": 1.1927284002304077,
+      "learning_rate": 0.00017210691805227398,
+      "loss": 0.5384,
+      "step": 3623
+    },
+    {
+      "epoch": 0.26649018310169864,
+      "grad_norm": 0.6163726449012756,
+      "learning_rate": 0.00017209041465395816,
+      "loss": 0.5165,
+      "step": 3624
+    },
+    {
+      "epoch": 0.2665637179204353,
+      "grad_norm": 1.1657469272613525,
+      "learning_rate": 0.00017207390716659794,
+      "loss": 0.505,
+      "step": 3625
+    },
+    {
+      "epoch": 0.266637252739172,
+      "grad_norm": 1.1075866222381592,
+      "learning_rate": 0.00017205739559112966,
+      "loss": 0.524,
+      "step": 3626
+    },
+    {
+      "epoch": 0.26671078755790867,
+      "grad_norm": 0.608222484588623,
+      "learning_rate": 0.00017204087992848984,
+      "loss": 0.5214,
+      "step": 3627
+    },
+    {
+      "epoch": 0.26678432237664534,
+      "grad_norm": 0.721727192401886,
+      "learning_rate": 0.0001720243601796153,
+      "loss": 0.5623,
+      "step": 3628
+    },
+    {
+      "epoch": 0.266857857195382,
+      "grad_norm": 1.1368709802627563,
+      "learning_rate": 0.00017200783634544302,
+      "loss": 0.5236,
+      "step": 3629
+    },
+    {
+      "epoch": 0.2669313920141187,
+      "grad_norm": 0.9082989692687988,
+      "learning_rate": 0.0001719913084269103,
+      "loss": 0.5305,
+      "step": 3630
+    },
+    {
+      "epoch": 0.26700492683285537,
+      "grad_norm": 0.7190804481506348,
+      "learning_rate": 0.00017197477642495458,
+      "loss": 0.5245,
+      "step": 3631
+    },
+    {
+      "epoch": 0.26707846165159205,
+      "grad_norm": 1.5635408163070679,
+      "learning_rate": 0.00017195824034051355,
+      "loss": 0.543,
+      "step": 3632
+    },
+    {
+      "epoch": 0.2671519964703287,
+      "grad_norm": 0.7807452082633972,
+      "learning_rate": 0.00017194170017452524,
+      "loss": 0.5326,
+      "step": 3633
+    },
+    {
+      "epoch": 0.2672255312890654,
+      "grad_norm": 1.5677706003189087,
+      "learning_rate": 0.00017192515592792776,
+      "loss": 0.5467,
+      "step": 3634
+    },
+    {
+      "epoch": 0.267299066107802,
+      "grad_norm": 1.3284684419631958,
+      "learning_rate": 0.0001719086076016595,
+      "loss": 0.5245,
+      "step": 3635
+    },
+    {
+      "epoch": 0.2673726009265387,
+      "grad_norm": 1.112683892250061,
+      "learning_rate": 0.00017189205519665913,
+      "loss": 0.5232,
+      "step": 3636
+    },
+    {
+      "epoch": 0.26744613574527537,
+      "grad_norm": 2.2972960472106934,
+      "learning_rate": 0.00017187549871386555,
+      "loss": 0.5251,
+      "step": 3637
+    },
+    {
+      "epoch": 0.26751967056401205,
+      "grad_norm": 1.85283625125885,
+      "learning_rate": 0.0001718589381542178,
+      "loss": 0.5293,
+      "step": 3638
+    },
+    {
+      "epoch": 0.2675932053827487,
+      "grad_norm": 2.139730453491211,
+      "learning_rate": 0.00017184237351865525,
+      "loss": 0.5006,
+      "step": 3639
+    },
+    {
+      "epoch": 0.2676667402014854,
+      "grad_norm": 2.588585376739502,
+      "learning_rate": 0.00017182580480811746,
+      "loss": 0.5236,
+      "step": 3640
+    },
+    {
+      "epoch": 0.2677402750202221,
+      "grad_norm": 1.2762563228607178,
+      "learning_rate": 0.00017180923202354424,
+      "loss": 0.5052,
+      "step": 3641
+    },
+    {
+      "epoch": 0.26781380983895875,
+      "grad_norm": 1.5151981115341187,
+      "learning_rate": 0.00017179265516587558,
+      "loss": 0.5554,
+      "step": 3642
+    },
+    {
+      "epoch": 0.2678873446576954,
+      "grad_norm": 1.3376989364624023,
+      "learning_rate": 0.00017177607423605174,
+      "loss": 0.5193,
+      "step": 3643
+    },
+    {
+      "epoch": 0.2679608794764321,
+      "grad_norm": 1.1173532009124756,
+      "learning_rate": 0.00017175948923501327,
+      "loss": 0.5308,
+      "step": 3644
+    },
+    {
+      "epoch": 0.2680344142951688,
+      "grad_norm": 2.1713576316833496,
+      "learning_rate": 0.00017174290016370078,
+      "loss": 0.5185,
+      "step": 3645
+    },
+    {
+      "epoch": 0.26810794911390545,
+      "grad_norm": 1.5816570520401,
+      "learning_rate": 0.0001717263070230553,
+      "loss": 0.5452,
+      "step": 3646
+    },
+    {
+      "epoch": 0.26818148393264213,
+      "grad_norm": 1.7513371706008911,
+      "learning_rate": 0.000171709709814018,
+      "loss": 0.4943,
+      "step": 3647
+    },
+    {
+      "epoch": 0.2682550187513788,
+      "grad_norm": 1.5672264099121094,
+      "learning_rate": 0.00017169310853753028,
+      "loss": 0.4985,
+      "step": 3648
+    },
+    {
+      "epoch": 0.2683285535701154,
+      "grad_norm": 1.4721280336380005,
+      "learning_rate": 0.00017167650319453376,
+      "loss": 0.5286,
+      "step": 3649
+    },
+    {
+      "epoch": 0.2684020883888521,
+      "grad_norm": 1.2381142377853394,
+      "learning_rate": 0.00017165989378597037,
+      "loss": 0.5007,
+      "step": 3650
+    },
+    {
+      "epoch": 0.2684756232075888,
+      "grad_norm": 1.6965957880020142,
+      "learning_rate": 0.00017164328031278215,
+      "loss": 0.5174,
+      "step": 3651
+    },
+    {
+      "epoch": 0.26854915802632545,
+      "grad_norm": 1.2849563360214233,
+      "learning_rate": 0.00017162666277591146,
+      "loss": 0.5378,
+      "step": 3652
+    },
+    {
+      "epoch": 0.2686226928450621,
+      "grad_norm": 1.889591097831726,
+      "learning_rate": 0.0001716100411763009,
+      "loss": 0.5307,
+      "step": 3653
+    },
+    {
+      "epoch": 0.2686962276637988,
+      "grad_norm": 1.5855867862701416,
+      "learning_rate": 0.0001715934155148932,
+      "loss": 0.5325,
+      "step": 3654
+    },
+    {
+      "epoch": 0.2687697624825355,
+      "grad_norm": 1.7954556941986084,
+      "learning_rate": 0.00017157678579263143,
+      "loss": 0.5137,
+      "step": 3655
+    },
+    {
+      "epoch": 0.26884329730127215,
+      "grad_norm": 1.4897021055221558,
+      "learning_rate": 0.00017156015201045882,
+      "loss": 0.5429,
+      "step": 3656
+    },
+    {
+      "epoch": 0.26891683212000883,
+      "grad_norm": 1.861596941947937,
+      "learning_rate": 0.00017154351416931888,
+      "loss": 0.53,
+      "step": 3657
+    },
+    {
+      "epoch": 0.2689903669387455,
+      "grad_norm": 1.525008201599121,
+      "learning_rate": 0.00017152687227015533,
+      "loss": 0.5181,
+      "step": 3658
+    },
+    {
+      "epoch": 0.2690639017574822,
+      "grad_norm": 1.72989821434021,
+      "learning_rate": 0.00017151022631391207,
+      "loss": 0.5311,
+      "step": 3659
+    },
+    {
+      "epoch": 0.26913743657621886,
+      "grad_norm": 1.5092711448669434,
+      "learning_rate": 0.00017149357630153332,
+      "loss": 0.5467,
+      "step": 3660
+    },
+    {
+      "epoch": 0.26921097139495553,
+      "grad_norm": 1.534665822982788,
+      "learning_rate": 0.00017147692223396345,
+      "loss": 0.5088,
+      "step": 3661
+    },
+    {
+      "epoch": 0.2692845062136922,
+      "grad_norm": 1.4860234260559082,
+      "learning_rate": 0.00017146026411214715,
+      "loss": 0.528,
+      "step": 3662
+    },
+    {
+      "epoch": 0.26935804103242883,
+      "grad_norm": 1.7936553955078125,
+      "learning_rate": 0.00017144360193702924,
+      "loss": 0.5447,
+      "step": 3663
+    },
+    {
+      "epoch": 0.2694315758511655,
+      "grad_norm": 1.6049896478652954,
+      "learning_rate": 0.0001714269357095548,
+      "loss": 0.5198,
+      "step": 3664
+    },
+    {
+      "epoch": 0.2695051106699022,
+      "grad_norm": 1.6678279638290405,
+      "learning_rate": 0.00017141026543066924,
+      "loss": 0.5196,
+      "step": 3665
+    },
+    {
+      "epoch": 0.26957864548863886,
+      "grad_norm": 1.6991413831710815,
+      "learning_rate": 0.00017139359110131806,
+      "loss": 0.5054,
+      "step": 3666
+    },
+    {
+      "epoch": 0.26965218030737553,
+      "grad_norm": 1.6878455877304077,
+      "learning_rate": 0.000171376912722447,
+      "loss": 0.5316,
+      "step": 3667
+    },
+    {
+      "epoch": 0.2697257151261122,
+      "grad_norm": 1.7929733991622925,
+      "learning_rate": 0.00017136023029500218,
+      "loss": 0.5334,
+      "step": 3668
+    },
+    {
+      "epoch": 0.2697992499448489,
+      "grad_norm": 1.4215071201324463,
+      "learning_rate": 0.00017134354381992975,
+      "loss": 0.5317,
+      "step": 3669
+    },
+    {
+      "epoch": 0.26987278476358556,
+      "grad_norm": 1.3601799011230469,
+      "learning_rate": 0.00017132685329817622,
+      "loss": 0.5183,
+      "step": 3670
+    },
+    {
+      "epoch": 0.26994631958232224,
+      "grad_norm": 1.5326422452926636,
+      "learning_rate": 0.00017131015873068833,
+      "loss": 0.5719,
+      "step": 3671
+    },
+    {
+      "epoch": 0.2700198544010589,
+      "grad_norm": 1.2295050621032715,
+      "learning_rate": 0.00017129346011841295,
+      "loss": 0.5154,
+      "step": 3672
+    },
+    {
+      "epoch": 0.2700933892197956,
+      "grad_norm": 1.6620646715164185,
+      "learning_rate": 0.00017127675746229728,
+      "loss": 0.5529,
+      "step": 3673
+    },
+    {
+      "epoch": 0.27016692403853226,
+      "grad_norm": 1.4500696659088135,
+      "learning_rate": 0.0001712600507632887,
+      "loss": 0.5174,
+      "step": 3674
+    },
+    {
+      "epoch": 0.27024045885726894,
+      "grad_norm": 1.671032428741455,
+      "learning_rate": 0.00017124334002233487,
+      "loss": 0.5254,
+      "step": 3675
+    },
+    {
+      "epoch": 0.2703139936760056,
+      "grad_norm": 1.514920949935913,
+      "learning_rate": 0.00017122662524038357,
+      "loss": 0.5459,
+      "step": 3676
+    },
+    {
+      "epoch": 0.27038752849474224,
+      "grad_norm": 1.4947669506072998,
+      "learning_rate": 0.00017120990641838292,
+      "loss": 0.5067,
+      "step": 3677
+    },
+    {
+      "epoch": 0.2704610633134789,
+      "grad_norm": 1.2357598543167114,
+      "learning_rate": 0.00017119318355728126,
+      "loss": 0.5406,
+      "step": 3678
+    },
+    {
+      "epoch": 0.2705345981322156,
+      "grad_norm": 1.951490879058838,
+      "learning_rate": 0.00017117645665802706,
+      "loss": 0.532,
+      "step": 3679
+    },
+    {
+      "epoch": 0.27060813295095226,
+      "grad_norm": 1.7458256483078003,
+      "learning_rate": 0.00017115972572156912,
+      "loss": 0.5694,
+      "step": 3680
+    },
+    {
+      "epoch": 0.27068166776968894,
+      "grad_norm": 1.5186657905578613,
+      "learning_rate": 0.00017114299074885646,
+      "loss": 0.5345,
+      "step": 3681
+    },
+    {
+      "epoch": 0.2707552025884256,
+      "grad_norm": 1.3323698043823242,
+      "learning_rate": 0.00017112625174083829,
+      "loss": 0.5467,
+      "step": 3682
+    },
+    {
+      "epoch": 0.2708287374071623,
+      "grad_norm": 1.517074465751648,
+      "learning_rate": 0.00017110950869846407,
+      "loss": 0.5186,
+      "step": 3683
+    },
+    {
+      "epoch": 0.27090227222589897,
+      "grad_norm": 1.30636727809906,
+      "learning_rate": 0.00017109276162268343,
+      "loss": 0.5272,
+      "step": 3684
+    },
+    {
+      "epoch": 0.27097580704463564,
+      "grad_norm": 1.5133056640625,
+      "learning_rate": 0.00017107601051444634,
+      "loss": 0.494,
+      "step": 3685
+    },
+    {
+      "epoch": 0.2710493418633723,
+      "grad_norm": 1.4326589107513428,
+      "learning_rate": 0.00017105925537470293,
+      "loss": 0.4963,
+      "step": 3686
+    },
+    {
+      "epoch": 0.271122876682109,
+      "grad_norm": 1.4814363718032837,
+      "learning_rate": 0.00017104249620440357,
+      "loss": 0.4888,
+      "step": 3687
+    },
+    {
+      "epoch": 0.27119641150084567,
+      "grad_norm": 1.33738374710083,
+      "learning_rate": 0.00017102573300449884,
+      "loss": 0.5301,
+      "step": 3688
+    },
+    {
+      "epoch": 0.27126994631958234,
+      "grad_norm": 1.5479962825775146,
+      "learning_rate": 0.00017100896577593958,
+      "loss": 0.4871,
+      "step": 3689
+    },
+    {
+      "epoch": 0.271343481138319,
+      "grad_norm": 1.3791221380233765,
+      "learning_rate": 0.00017099219451967686,
+      "loss": 0.5154,
+      "step": 3690
+    },
+    {
+      "epoch": 0.27141701595705564,
+      "grad_norm": 1.302624225616455,
+      "learning_rate": 0.00017097541923666193,
+      "loss": 0.5265,
+      "step": 3691
+    },
+    {
+      "epoch": 0.2714905507757923,
+      "grad_norm": 1.2849458456039429,
+      "learning_rate": 0.0001709586399278463,
+      "loss": 0.5126,
+      "step": 3692
+    },
+    {
+      "epoch": 0.271564085594529,
+      "grad_norm": 1.4829591512680054,
+      "learning_rate": 0.00017094185659418174,
+      "loss": 0.5344,
+      "step": 3693
+    },
+    {
+      "epoch": 0.27163762041326567,
+      "grad_norm": 1.2711937427520752,
+      "learning_rate": 0.00017092506923662023,
+      "loss": 0.486,
+      "step": 3694
+    },
+    {
+      "epoch": 0.27171115523200234,
+      "grad_norm": 1.6768383979797363,
+      "learning_rate": 0.00017090827785611393,
+      "loss": 0.5415,
+      "step": 3695
+    },
+    {
+      "epoch": 0.271784690050739,
+      "grad_norm": 1.407586693763733,
+      "learning_rate": 0.00017089148245361523,
+      "loss": 0.5162,
+      "step": 3696
+    },
+    {
+      "epoch": 0.2718582248694757,
+      "grad_norm": 1.6269546747207642,
+      "learning_rate": 0.0001708746830300769,
+      "loss": 0.5379,
+      "step": 3697
+    },
+    {
+      "epoch": 0.27193175968821237,
+      "grad_norm": 1.6850175857543945,
+      "learning_rate": 0.00017085787958645172,
+      "loss": 0.5437,
+      "step": 3698
+    },
+    {
+      "epoch": 0.27200529450694905,
+      "grad_norm": 1.2529069185256958,
+      "learning_rate": 0.00017084107212369281,
+      "loss": 0.5246,
+      "step": 3699
+    },
+    {
+      "epoch": 0.2720788293256857,
+      "grad_norm": 1.2996755838394165,
+      "learning_rate": 0.00017082426064275356,
+      "loss": 0.5168,
+      "step": 3700
+    },
+    {
+      "epoch": 0.2721523641444224,
+      "grad_norm": 1.5801390409469604,
+      "learning_rate": 0.0001708074451445875,
+      "loss": 0.5347,
+      "step": 3701
+    },
+    {
+      "epoch": 0.2722258989631591,
+      "grad_norm": 1.0471577644348145,
+      "learning_rate": 0.0001707906256301484,
+      "loss": 0.4963,
+      "step": 3702
+    },
+    {
+      "epoch": 0.27229943378189575,
+      "grad_norm": 2.145655393600464,
+      "learning_rate": 0.00017077380210039035,
+      "loss": 0.5048,
+      "step": 3703
+    },
+    {
+      "epoch": 0.2723729686006324,
+      "grad_norm": 1.7303636074066162,
+      "learning_rate": 0.00017075697455626752,
+      "loss": 0.55,
+      "step": 3704
+    },
+    {
+      "epoch": 0.27244650341936905,
+      "grad_norm": 1.4953199625015259,
+      "learning_rate": 0.00017074014299873444,
+      "loss": 0.5112,
+      "step": 3705
+    },
+    {
+      "epoch": 0.2725200382381057,
+      "grad_norm": 1.336361289024353,
+      "learning_rate": 0.00017072330742874578,
+      "loss": 0.5275,
+      "step": 3706
+    },
+    {
+      "epoch": 0.2725935730568424,
+      "grad_norm": 1.5268298387527466,
+      "learning_rate": 0.00017070646784725648,
+      "loss": 0.4945,
+      "step": 3707
+    },
+    {
+      "epoch": 0.2726671078755791,
+      "grad_norm": 1.0755435228347778,
+      "learning_rate": 0.00017068962425522173,
+      "loss": 0.5113,
+      "step": 3708
+    },
+    {
+      "epoch": 0.27274064269431575,
+      "grad_norm": 1.7274549007415771,
+      "learning_rate": 0.00017067277665359687,
+      "loss": 0.4835,
+      "step": 3709
+    },
+    {
+      "epoch": 0.2728141775130524,
+      "grad_norm": 1.6112632751464844,
+      "learning_rate": 0.00017065592504333753,
+      "loss": 0.48,
+      "step": 3710
+    },
+    {
+      "epoch": 0.2728877123317891,
+      "grad_norm": 1.416008472442627,
+      "learning_rate": 0.00017063906942539956,
+      "loss": 0.4879,
+      "step": 3711
+    },
+    {
+      "epoch": 0.2729612471505258,
+      "grad_norm": 1.4645800590515137,
+      "learning_rate": 0.00017062220980073903,
+      "loss": 0.5012,
+      "step": 3712
+    },
+    {
+      "epoch": 0.27303478196926245,
+      "grad_norm": 1.2543195486068726,
+      "learning_rate": 0.00017060534617031222,
+      "loss": 0.4903,
+      "step": 3713
+    },
+    {
+      "epoch": 0.27310831678799913,
+      "grad_norm": 1.1352492570877075,
+      "learning_rate": 0.00017058847853507566,
+      "loss": 0.5047,
+      "step": 3714
+    },
+    {
+      "epoch": 0.2731818516067358,
+      "grad_norm": 1.7640162706375122,
+      "learning_rate": 0.00017057160689598613,
+      "loss": 0.48,
+      "step": 3715
+    },
+    {
+      "epoch": 0.2732553864254725,
+      "grad_norm": 1.6440311670303345,
+      "learning_rate": 0.00017055473125400056,
+      "loss": 0.4933,
+      "step": 3716
+    },
+    {
+      "epoch": 0.27332892124420916,
+      "grad_norm": 1.5058239698410034,
+      "learning_rate": 0.00017053785161007617,
+      "loss": 0.5493,
+      "step": 3717
+    },
+    {
+      "epoch": 0.27340245606294583,
+      "grad_norm": 1.4277029037475586,
+      "learning_rate": 0.00017052096796517038,
+      "loss": 0.519,
+      "step": 3718
+    },
+    {
+      "epoch": 0.27347599088168245,
+      "grad_norm": 1.553963303565979,
+      "learning_rate": 0.0001705040803202409,
+      "loss": 0.4881,
+      "step": 3719
+    },
+    {
+      "epoch": 0.2735495257004191,
+      "grad_norm": 1.5629254579544067,
+      "learning_rate": 0.0001704871886762456,
+      "loss": 0.4999,
+      "step": 3720
+    },
+    {
+      "epoch": 0.2736230605191558,
+      "grad_norm": 1.31739342212677,
+      "learning_rate": 0.0001704702930341425,
+      "loss": 0.5088,
+      "step": 3721
+    },
+    {
+      "epoch": 0.2736965953378925,
+      "grad_norm": 1.3929870128631592,
+      "learning_rate": 0.00017045339339489003,
+      "loss": 0.4809,
+      "step": 3722
+    },
+    {
+      "epoch": 0.27377013015662915,
+      "grad_norm": 1.1819039583206177,
+      "learning_rate": 0.00017043648975944675,
+      "loss": 0.5028,
+      "step": 3723
+    },
+    {
+      "epoch": 0.27384366497536583,
+      "grad_norm": 1.1807981729507446,
+      "learning_rate": 0.00017041958212877143,
+      "loss": 0.4822,
+      "step": 3724
+    },
+    {
+      "epoch": 0.2739171997941025,
+      "grad_norm": 1.9245295524597168,
+      "learning_rate": 0.0001704026705038231,
+      "loss": 0.4913,
+      "step": 3725
+    },
+    {
+      "epoch": 0.2739907346128392,
+      "grad_norm": 1.3534046411514282,
+      "learning_rate": 0.00017038575488556102,
+      "loss": 0.5279,
+      "step": 3726
+    },
+    {
+      "epoch": 0.27406426943157586,
+      "grad_norm": 1.6327078342437744,
+      "learning_rate": 0.0001703688352749446,
+      "loss": 0.5182,
+      "step": 3727
+    },
+    {
+      "epoch": 0.27413780425031253,
+      "grad_norm": 1.3568240404129028,
+      "learning_rate": 0.00017035191167293358,
+      "loss": 0.4926,
+      "step": 3728
+    },
+    {
+      "epoch": 0.2742113390690492,
+      "grad_norm": 1.6833972930908203,
+      "learning_rate": 0.0001703349840804879,
+      "loss": 0.5317,
+      "step": 3729
+    },
+    {
+      "epoch": 0.2742848738877859,
+      "grad_norm": 1.1968659162521362,
+      "learning_rate": 0.0001703180524985677,
+      "loss": 0.5267,
+      "step": 3730
+    },
+    {
+      "epoch": 0.27435840870652256,
+      "grad_norm": 1.691642165184021,
+      "learning_rate": 0.00017030111692813336,
+      "loss": 0.5159,
+      "step": 3731
+    },
+    {
+      "epoch": 0.27443194352525924,
+      "grad_norm": 1.2802506685256958,
+      "learning_rate": 0.0001702841773701455,
+      "loss": 0.5115,
+      "step": 3732
+    },
+    {
+      "epoch": 0.27450547834399586,
+      "grad_norm": 1.6899384260177612,
+      "learning_rate": 0.00017026723382556486,
+      "loss": 0.5147,
+      "step": 3733
+    },
+    {
+      "epoch": 0.27457901316273253,
+      "grad_norm": 1.3319584131240845,
+      "learning_rate": 0.0001702502862953526,
+      "loss": 0.537,
+      "step": 3734
+    },
+    {
+      "epoch": 0.2746525479814692,
+      "grad_norm": 1.6300195455551147,
+      "learning_rate": 0.00017023333478046997,
+      "loss": 0.497,
+      "step": 3735
+    },
+    {
+      "epoch": 0.2747260828002059,
+      "grad_norm": 1.3889424800872803,
+      "learning_rate": 0.00017021637928187844,
+      "loss": 0.5263,
+      "step": 3736
+    },
+    {
+      "epoch": 0.27479961761894256,
+      "grad_norm": 1.452985405921936,
+      "learning_rate": 0.0001701994198005398,
+      "loss": 0.5161,
+      "step": 3737
+    },
+    {
+      "epoch": 0.27487315243767924,
+      "grad_norm": 1.140074610710144,
+      "learning_rate": 0.00017018245633741597,
+      "loss": 0.5072,
+      "step": 3738
+    },
+    {
+      "epoch": 0.2749466872564159,
+      "grad_norm": 1.8563662767410278,
+      "learning_rate": 0.00017016548889346917,
+      "loss": 0.5402,
+      "step": 3739
+    },
+    {
+      "epoch": 0.2750202220751526,
+      "grad_norm": 1.1396527290344238,
+      "learning_rate": 0.00017014851746966175,
+      "loss": 0.5115,
+      "step": 3740
+    },
+    {
+      "epoch": 0.27509375689388926,
+      "grad_norm": 1.8224670886993408,
+      "learning_rate": 0.00017013154206695644,
+      "loss": 0.5055,
+      "step": 3741
+    },
+    {
+      "epoch": 0.27516729171262594,
+      "grad_norm": 1.3004552125930786,
+      "learning_rate": 0.00017011456268631602,
+      "loss": 0.5258,
+      "step": 3742
+    },
+    {
+      "epoch": 0.2752408265313626,
+      "grad_norm": 1.0637668371200562,
+      "learning_rate": 0.00017009757932870363,
+      "loss": 0.5038,
+      "step": 3743
+    },
+    {
+      "epoch": 0.2753143613500993,
+      "grad_norm": 0.8769312500953674,
+      "learning_rate": 0.00017008059199508254,
+      "loss": 0.5054,
+      "step": 3744
+    },
+    {
+      "epoch": 0.27538789616883597,
+      "grad_norm": 1.297065258026123,
+      "learning_rate": 0.00017006360068641633,
+      "loss": 0.4837,
+      "step": 3745
+    },
+    {
+      "epoch": 0.27546143098757264,
+      "grad_norm": 0.7307043671607971,
+      "learning_rate": 0.00017004660540366872,
+      "loss": 0.5047,
+      "step": 3746
+    },
+    {
+      "epoch": 0.27553496580630926,
+      "grad_norm": 1.2058576345443726,
+      "learning_rate": 0.00017002960614780378,
+      "loss": 0.5043,
+      "step": 3747
+    },
+    {
+      "epoch": 0.27560850062504594,
+      "grad_norm": 0.7041051983833313,
+      "learning_rate": 0.00017001260291978564,
+      "loss": 0.4989,
+      "step": 3748
+    },
+    {
+      "epoch": 0.2756820354437826,
+      "grad_norm": 0.9413318634033203,
+      "learning_rate": 0.00016999559572057878,
+      "loss": 0.51,
+      "step": 3749
+    },
+    {
+      "epoch": 0.2757555702625193,
+      "grad_norm": 1.1291922330856323,
+      "learning_rate": 0.00016997858455114785,
+      "loss": 0.5574,
+      "step": 3750
+    },
+    {
+      "epoch": 0.27582910508125597,
+      "grad_norm": 0.8130052089691162,
+      "learning_rate": 0.00016996156941245777,
+      "loss": 0.498,
+      "step": 3751
+    },
+    {
+      "epoch": 0.27590263989999264,
+      "grad_norm": 1.1036922931671143,
+      "learning_rate": 0.00016994455030547366,
+      "loss": 0.5378,
+      "step": 3752
+    },
+    {
+      "epoch": 0.2759761747187293,
+      "grad_norm": 0.6816931962966919,
+      "learning_rate": 0.0001699275272311608,
+      "loss": 0.459,
+      "step": 3753
+    },
+    {
+      "epoch": 0.276049709537466,
+      "grad_norm": 0.927787721157074,
+      "learning_rate": 0.00016991050019048482,
+      "loss": 0.5137,
+      "step": 3754
+    },
+    {
+      "epoch": 0.27612324435620267,
+      "grad_norm": 1.0556777715682983,
+      "learning_rate": 0.0001698934691844115,
+      "loss": 0.4909,
+      "step": 3755
+    },
+    {
+      "epoch": 0.27619677917493934,
+      "grad_norm": 0.6569586992263794,
+      "learning_rate": 0.0001698764342139068,
+      "loss": 0.5237,
+      "step": 3756
+    },
+    {
+      "epoch": 0.276270313993676,
+      "grad_norm": 0.9705957770347595,
+      "learning_rate": 0.00016985939527993705,
+      "loss": 0.4794,
+      "step": 3757
+    },
+    {
+      "epoch": 0.2763438488124127,
+      "grad_norm": 0.718802273273468,
+      "learning_rate": 0.00016984235238346864,
+      "loss": 0.5171,
+      "step": 3758
+    },
+    {
+      "epoch": 0.27641738363114937,
+      "grad_norm": 0.9245346188545227,
+      "learning_rate": 0.00016982530552546834,
+      "loss": 0.5078,
+      "step": 3759
+    },
+    {
+      "epoch": 0.27649091844988605,
+      "grad_norm": 0.9671157598495483,
+      "learning_rate": 0.000169808254706903,
+      "loss": 0.4926,
+      "step": 3760
+    },
+    {
+      "epoch": 0.27656445326862267,
+      "grad_norm": 0.6022995710372925,
+      "learning_rate": 0.00016979119992873977,
+      "loss": 0.4921,
+      "step": 3761
+    },
+    {
+      "epoch": 0.27663798808735934,
+      "grad_norm": 0.9312941431999207,
+      "learning_rate": 0.00016977414119194602,
+      "loss": 0.5263,
+      "step": 3762
+    },
+    {
+      "epoch": 0.276711522906096,
+      "grad_norm": 0.9181712865829468,
+      "learning_rate": 0.0001697570784974894,
+      "loss": 0.5074,
+      "step": 3763
+    },
+    {
+      "epoch": 0.2767850577248327,
+      "grad_norm": 0.6013015508651733,
+      "learning_rate": 0.00016974001184633762,
+      "loss": 0.481,
+      "step": 3764
+    },
+    {
+      "epoch": 0.27685859254356937,
+      "grad_norm": 1.1595017910003662,
+      "learning_rate": 0.00016972294123945874,
+      "loss": 0.5054,
+      "step": 3765
+    },
+    {
+      "epoch": 0.27693212736230605,
+      "grad_norm": 1.110661268234253,
+      "learning_rate": 0.00016970586667782112,
+      "loss": 0.5047,
+      "step": 3766
+    },
+    {
+      "epoch": 0.2770056621810427,
+      "grad_norm": 0.7793318629264832,
+      "learning_rate": 0.00016968878816239316,
+      "loss": 0.4606,
+      "step": 3767
+    },
+    {
+      "epoch": 0.2770791969997794,
+      "grad_norm": 0.8500633239746094,
+      "learning_rate": 0.00016967170569414357,
+      "loss": 0.5161,
+      "step": 3768
+    },
+    {
+      "epoch": 0.2771527318185161,
+      "grad_norm": 0.8294308185577393,
+      "learning_rate": 0.00016965461927404132,
+      "loss": 0.4473,
+      "step": 3769
+    },
+    {
+      "epoch": 0.27722626663725275,
+      "grad_norm": 0.9069722890853882,
+      "learning_rate": 0.00016963752890305554,
+      "loss": 0.4931,
+      "step": 3770
+    },
+    {
+      "epoch": 0.2772998014559894,
+      "grad_norm": 0.618799090385437,
+      "learning_rate": 0.00016962043458215563,
+      "loss": 0.5184,
+      "step": 3771
+    },
+    {
+      "epoch": 0.2773733362747261,
+      "grad_norm": 0.70931077003479,
+      "learning_rate": 0.00016960333631231122,
+      "loss": 0.4857,
+      "step": 3772
+    },
+    {
+      "epoch": 0.2774468710934628,
+      "grad_norm": 1.024071455001831,
+      "learning_rate": 0.00016958623409449214,
+      "loss": 0.5035,
+      "step": 3773
+    },
+    {
+      "epoch": 0.27752040591219945,
+      "grad_norm": 0.9760534763336182,
+      "learning_rate": 0.00016956912792966838,
+      "loss": 0.4958,
+      "step": 3774
+    },
+    {
+      "epoch": 0.2775939407309361,
+      "grad_norm": 0.6068201065063477,
+      "learning_rate": 0.00016955201781881026,
+      "loss": 0.4714,
+      "step": 3775
+    },
+    {
+      "epoch": 0.27766747554967275,
+      "grad_norm": 0.9181552529335022,
+      "learning_rate": 0.0001695349037628883,
+      "loss": 0.4988,
+      "step": 3776
+    },
+    {
+      "epoch": 0.2777410103684094,
+      "grad_norm": 1.019065022468567,
+      "learning_rate": 0.00016951778576287322,
+      "loss": 0.4944,
+      "step": 3777
+    },
+    {
+      "epoch": 0.2778145451871461,
+      "grad_norm": 0.8325920104980469,
+      "learning_rate": 0.00016950066381973597,
+      "loss": 0.4927,
+      "step": 3778
+    },
+    {
+      "epoch": 0.2778880800058828,
+      "grad_norm": 0.7647660374641418,
+      "learning_rate": 0.00016948353793444772,
+      "loss": 0.4982,
+      "step": 3779
+    },
+    {
+      "epoch": 0.27796161482461945,
+      "grad_norm": 0.806194007396698,
+      "learning_rate": 0.00016946640810797984,
+      "loss": 0.483,
+      "step": 3780
+    },
+    {
+      "epoch": 0.2780351496433561,
+      "grad_norm": 0.6721197366714478,
+      "learning_rate": 0.00016944927434130405,
+      "loss": 0.4967,
+      "step": 3781
+    },
+    {
+      "epoch": 0.2781086844620928,
+      "grad_norm": 0.9054644107818604,
+      "learning_rate": 0.00016943213663539206,
+      "loss": 0.472,
+      "step": 3782
+    },
+    {
+      "epoch": 0.2781822192808295,
+      "grad_norm": 1.4267768859863281,
+      "learning_rate": 0.00016941499499121605,
+      "loss": 0.4945,
+      "step": 3783
+    },
+    {
+      "epoch": 0.27825575409956615,
+      "grad_norm": 0.6576398015022278,
+      "learning_rate": 0.00016939784940974824,
+      "loss": 0.5478,
+      "step": 3784
+    },
+    {
+      "epoch": 0.27832928891830283,
+      "grad_norm": 1.606078863143921,
+      "learning_rate": 0.00016938069989196118,
+      "loss": 0.4835,
+      "step": 3785
+    },
+    {
+      "epoch": 0.2784028237370395,
+      "grad_norm": 0.7444186210632324,
+      "learning_rate": 0.0001693635464388276,
+      "loss": 0.4678,
+      "step": 3786
+    },
+    {
+      "epoch": 0.2784763585557762,
+      "grad_norm": 1.3784575462341309,
+      "learning_rate": 0.00016934638905132044,
+      "loss": 0.5059,
+      "step": 3787
+    },
+    {
+      "epoch": 0.27854989337451286,
+      "grad_norm": 0.7581371665000916,
+      "learning_rate": 0.00016932922773041294,
+      "loss": 0.4794,
+      "step": 3788
+    },
+    {
+      "epoch": 0.2786234281932495,
+      "grad_norm": 1.0061012506484985,
+      "learning_rate": 0.0001693120624770785,
+      "loss": 0.4933,
+      "step": 3789
+    },
+    {
+      "epoch": 0.27869696301198615,
+      "grad_norm": 1.5050722360610962,
+      "learning_rate": 0.00016929489329229068,
+      "loss": 0.485,
+      "step": 3790
+    },
+    {
+      "epoch": 0.27877049783072283,
+      "grad_norm": 0.7599188089370728,
+      "learning_rate": 0.00016927772017702344,
+      "loss": 0.4763,
+      "step": 3791
+    },
+    {
+      "epoch": 0.2788440326494595,
+      "grad_norm": 1.8443493843078613,
+      "learning_rate": 0.00016926054313225075,
+      "loss": 0.5143,
+      "step": 3792
+    },
+    {
+      "epoch": 0.2789175674681962,
+      "grad_norm": 1.008804440498352,
+      "learning_rate": 0.00016924336215894695,
+      "loss": 0.5127,
+      "step": 3793
+    },
+    {
+      "epoch": 0.27899110228693286,
+      "grad_norm": 3.0983638763427734,
+      "learning_rate": 0.0001692261772580866,
+      "loss": 0.4966,
+      "step": 3794
+    },
+    {
+      "epoch": 0.27906463710566953,
+      "grad_norm": 3.3106191158294678,
+      "learning_rate": 0.00016920898843064445,
+      "loss": 0.4742,
+      "step": 3795
+    },
+    {
+      "epoch": 0.2791381719244062,
+      "grad_norm": 0.8600287437438965,
+      "learning_rate": 0.00016919179567759539,
+      "loss": 0.5205,
+      "step": 3796
+    },
+    {
+      "epoch": 0.2792117067431429,
+      "grad_norm": 3.280297040939331,
+      "learning_rate": 0.00016917459899991466,
+      "loss": 0.5171,
+      "step": 3797
+    },
+    {
+      "epoch": 0.27928524156187956,
+      "grad_norm": 3.0831236839294434,
+      "learning_rate": 0.00016915739839857768,
+      "loss": 0.5133,
+      "step": 3798
+    },
+    {
+      "epoch": 0.27935877638061624,
+      "grad_norm": 1.2397141456604004,
+      "learning_rate": 0.0001691401938745601,
+      "loss": 0.4757,
+      "step": 3799
+    },
+    {
+      "epoch": 0.2794323111993529,
+      "grad_norm": 2.2558205127716064,
+      "learning_rate": 0.00016912298542883772,
+      "loss": 0.5192,
+      "step": 3800
+    },
+    {
+      "epoch": 0.2795058460180896,
+      "grad_norm": 1.8565986156463623,
+      "learning_rate": 0.00016910577306238666,
+      "loss": 0.5044,
+      "step": 3801
+    },
+    {
+      "epoch": 0.27957938083682626,
+      "grad_norm": 1.6774263381958008,
+      "learning_rate": 0.00016908855677618325,
+      "loss": 0.499,
+      "step": 3802
+    },
+    {
+      "epoch": 0.2796529156555629,
+      "grad_norm": 1.6500383615493774,
+      "learning_rate": 0.00016907133657120395,
+      "loss": 0.5322,
+      "step": 3803
+    },
+    {
+      "epoch": 0.27972645047429956,
+      "grad_norm": 1.1635632514953613,
+      "learning_rate": 0.0001690541124484256,
+      "loss": 0.5042,
+      "step": 3804
+    },
+    {
+      "epoch": 0.27979998529303624,
+      "grad_norm": 1.6270389556884766,
+      "learning_rate": 0.00016903688440882508,
+      "loss": 0.4987,
+      "step": 3805
+    },
+    {
+      "epoch": 0.2798735201117729,
+      "grad_norm": 1.4217513799667358,
+      "learning_rate": 0.00016901965245337962,
+      "loss": 0.477,
+      "step": 3806
+    },
+    {
+      "epoch": 0.2799470549305096,
+      "grad_norm": 1.0284348726272583,
+      "learning_rate": 0.00016900241658306665,
+      "loss": 0.4628,
+      "step": 3807
+    },
+    {
+      "epoch": 0.28002058974924626,
+      "grad_norm": 1.3058878183364868,
+      "learning_rate": 0.00016898517679886377,
+      "loss": 0.4907,
+      "step": 3808
+    },
+    {
+      "epoch": 0.28009412456798294,
+      "grad_norm": 1.1893832683563232,
+      "learning_rate": 0.00016896793310174885,
+      "loss": 0.4807,
+      "step": 3809
+    },
+    {
+      "epoch": 0.2801676593867196,
+      "grad_norm": 1.2588542699813843,
+      "learning_rate": 0.00016895068549270002,
+      "loss": 0.4974,
+      "step": 3810
+    },
+    {
+      "epoch": 0.2802411942054563,
+      "grad_norm": 1.0195170640945435,
+      "learning_rate": 0.00016893343397269552,
+      "loss": 0.5027,
+      "step": 3811
+    },
+    {
+      "epoch": 0.28031472902419297,
+      "grad_norm": 1.1702467203140259,
+      "learning_rate": 0.0001689161785427139,
+      "loss": 0.5081,
+      "step": 3812
+    },
+    {
+      "epoch": 0.28038826384292964,
+      "grad_norm": 1.1422303915023804,
+      "learning_rate": 0.0001688989192037339,
+      "loss": 0.4971,
+      "step": 3813
+    },
+    {
+      "epoch": 0.2804617986616663,
+      "grad_norm": 0.8672306537628174,
+      "learning_rate": 0.0001688816559567345,
+      "loss": 0.5039,
+      "step": 3814
+    },
+    {
+      "epoch": 0.280535333480403,
+      "grad_norm": 1.0595232248306274,
+      "learning_rate": 0.00016886438880269483,
+      "loss": 0.5474,
+      "step": 3815
+    },
+    {
+      "epoch": 0.28060886829913967,
+      "grad_norm": 0.7233802676200867,
+      "learning_rate": 0.00016884711774259442,
+      "loss": 0.4975,
+      "step": 3816
+    },
+    {
+      "epoch": 0.2806824031178763,
+      "grad_norm": 1.0852291584014893,
+      "learning_rate": 0.00016882984277741278,
+      "loss": 0.4857,
+      "step": 3817
+    },
+    {
+      "epoch": 0.28075593793661296,
+      "grad_norm": 0.6231606006622314,
+      "learning_rate": 0.00016881256390812986,
+      "loss": 0.491,
+      "step": 3818
+    },
+    {
+      "epoch": 0.28082947275534964,
+      "grad_norm": 1.2002453804016113,
+      "learning_rate": 0.0001687952811357257,
+      "loss": 0.4685,
+      "step": 3819
+    },
+    {
+      "epoch": 0.2809030075740863,
+      "grad_norm": 0.8536419868469238,
+      "learning_rate": 0.00016877799446118056,
+      "loss": 0.5307,
+      "step": 3820
+    },
+    {
+      "epoch": 0.280976542392823,
+      "grad_norm": 0.8386676907539368,
+      "learning_rate": 0.000168760703885475,
+      "loss": 0.4835,
+      "step": 3821
+    },
+    {
+      "epoch": 0.28105007721155967,
+      "grad_norm": 1.450599193572998,
+      "learning_rate": 0.00016874340940958976,
+      "loss": 0.4872,
+      "step": 3822
+    },
+    {
+      "epoch": 0.28112361203029634,
+      "grad_norm": 0.7631549835205078,
+      "learning_rate": 0.00016872611103450576,
+      "loss": 0.4639,
+      "step": 3823
+    },
+    {
+      "epoch": 0.281197146849033,
+      "grad_norm": 2.0491366386413574,
+      "learning_rate": 0.00016870880876120425,
+      "loss": 0.5025,
+      "step": 3824
+    },
+    {
+      "epoch": 0.2812706816677697,
+      "grad_norm": 1.3106803894042969,
+      "learning_rate": 0.00016869150259066657,
+      "loss": 0.4857,
+      "step": 3825
+    },
+    {
+      "epoch": 0.28134421648650637,
+      "grad_norm": 2.240690231323242,
+      "learning_rate": 0.00016867419252387437,
+      "loss": 0.4973,
+      "step": 3826
+    },
+    {
+      "epoch": 0.28141775130524305,
+      "grad_norm": 2.2369275093078613,
+      "learning_rate": 0.0001686568785618095,
+      "loss": 0.5023,
+      "step": 3827
+    },
+    {
+      "epoch": 0.2814912861239797,
+      "grad_norm": 0.9880452156066895,
+      "learning_rate": 0.00016863956070545404,
+      "loss": 0.4959,
+      "step": 3828
+    },
+    {
+      "epoch": 0.2815648209427164,
+      "grad_norm": 1.205361008644104,
+      "learning_rate": 0.00016862223895579024,
+      "loss": 0.5035,
+      "step": 3829
+    },
+    {
+      "epoch": 0.2816383557614531,
+      "grad_norm": 1.4465514421463013,
+      "learning_rate": 0.00016860491331380063,
+      "loss": 0.4843,
+      "step": 3830
+    },
+    {
+      "epoch": 0.2817118905801897,
+      "grad_norm": 0.850191056728363,
+      "learning_rate": 0.00016858758378046794,
+      "loss": 0.491,
+      "step": 3831
+    },
+    {
+      "epoch": 0.28178542539892637,
+      "grad_norm": 1.880473256111145,
+      "learning_rate": 0.00016857025035677513,
+      "loss": 0.5061,
+      "step": 3832
+    },
+    {
+      "epoch": 0.28185896021766305,
+      "grad_norm": 1.1297498941421509,
+      "learning_rate": 0.00016855291304370532,
+      "loss": 0.4588,
+      "step": 3833
+    },
+    {
+      "epoch": 0.2819324950363997,
+      "grad_norm": 2.036827802658081,
+      "learning_rate": 0.00016853557184224195,
+      "loss": 0.5149,
+      "step": 3834
+    },
+    {
+      "epoch": 0.2820060298551364,
+      "grad_norm": 1.7815383672714233,
+      "learning_rate": 0.00016851822675336862,
+      "loss": 0.4882,
+      "step": 3835
+    },
+    {
+      "epoch": 0.2820795646738731,
+      "grad_norm": 1.3091585636138916,
+      "learning_rate": 0.00016850087777806915,
+      "loss": 0.5153,
+      "step": 3836
+    },
+    {
+      "epoch": 0.28215309949260975,
+      "grad_norm": 1.0647560358047485,
+      "learning_rate": 0.00016848352491732763,
+      "loss": 0.5004,
+      "step": 3837
+    },
+    {
+      "epoch": 0.2822266343113464,
+      "grad_norm": 2.0053131580352783,
+      "learning_rate": 0.00016846616817212825,
+      "loss": 0.5084,
+      "step": 3838
+    },
+    {
+      "epoch": 0.2823001691300831,
+      "grad_norm": 1.3132367134094238,
+      "learning_rate": 0.0001684488075434556,
+      "loss": 0.4778,
+      "step": 3839
+    },
+    {
+      "epoch": 0.2823737039488198,
+      "grad_norm": 1.6292738914489746,
+      "learning_rate": 0.0001684314430322943,
+      "loss": 0.4883,
+      "step": 3840
+    },
+    {
+      "epoch": 0.28244723876755645,
+      "grad_norm": 1.3394725322723389,
+      "learning_rate": 0.0001684140746396294,
+      "loss": 0.5032,
+      "step": 3841
+    },
+    {
+      "epoch": 0.28252077358629313,
+      "grad_norm": 1.5366202592849731,
+      "learning_rate": 0.00016839670236644595,
+      "loss": 0.481,
+      "step": 3842
+    },
+    {
+      "epoch": 0.2825943084050298,
+      "grad_norm": 1.301266074180603,
+      "learning_rate": 0.0001683793262137294,
+      "loss": 0.4615,
+      "step": 3843
+    },
+    {
+      "epoch": 0.2826678432237665,
+      "grad_norm": 1.7865251302719116,
+      "learning_rate": 0.00016836194618246523,
+      "loss": 0.4869,
+      "step": 3844
+    },
+    {
+      "epoch": 0.2827413780425031,
+      "grad_norm": 1.9469773769378662,
+      "learning_rate": 0.00016834456227363936,
+      "loss": 0.5131,
+      "step": 3845
+    },
+    {
+      "epoch": 0.2828149128612398,
+      "grad_norm": 1.1690438985824585,
+      "learning_rate": 0.0001683271744882378,
+      "loss": 0.4967,
+      "step": 3846
+    },
+    {
+      "epoch": 0.28288844767997645,
+      "grad_norm": 1.162516713142395,
+      "learning_rate": 0.0001683097828272468,
+      "loss": 0.4842,
+      "step": 3847
+    },
+    {
+      "epoch": 0.2829619824987131,
+      "grad_norm": 1.350223183631897,
+      "learning_rate": 0.00016829238729165283,
+      "loss": 0.4918,
+      "step": 3848
+    },
+    {
+      "epoch": 0.2830355173174498,
+      "grad_norm": 1.1078466176986694,
+      "learning_rate": 0.00016827498788244257,
+      "loss": 0.4846,
+      "step": 3849
+    },
+    {
+      "epoch": 0.2831090521361865,
+      "grad_norm": 1.8199208974838257,
+      "learning_rate": 0.00016825758460060292,
+      "loss": 0.4947,
+      "step": 3850
+    },
+    {
+      "epoch": 0.28318258695492315,
+      "grad_norm": 1.7717530727386475,
+      "learning_rate": 0.00016824017744712107,
+      "loss": 0.488,
+      "step": 3851
+    },
+    {
+      "epoch": 0.28325612177365983,
+      "grad_norm": 1.4025018215179443,
+      "learning_rate": 0.0001682227664229843,
+      "loss": 0.4951,
+      "step": 3852
+    },
+    {
+      "epoch": 0.2833296565923965,
+      "grad_norm": 1.3372883796691895,
+      "learning_rate": 0.0001682053515291803,
+      "loss": 0.4605,
+      "step": 3853
+    },
+    {
+      "epoch": 0.2834031914111332,
+      "grad_norm": 1.4675410985946655,
+      "learning_rate": 0.00016818793276669668,
+      "loss": 0.4919,
+      "step": 3854
+    },
+    {
+      "epoch": 0.28347672622986986,
+      "grad_norm": 1.184800624847412,
+      "learning_rate": 0.00016817051013652163,
+      "loss": 0.506,
+      "step": 3855
+    },
+    {
+      "epoch": 0.28355026104860653,
+      "grad_norm": 1.5830122232437134,
+      "learning_rate": 0.00016815308363964327,
+      "loss": 0.4756,
+      "step": 3856
+    },
+    {
+      "epoch": 0.2836237958673432,
+      "grad_norm": 1.446945309638977,
+      "learning_rate": 0.00016813565327705008,
+      "loss": 0.4905,
+      "step": 3857
+    },
+    {
+      "epoch": 0.2836973306860799,
+      "grad_norm": 1.7619283199310303,
+      "learning_rate": 0.00016811821904973073,
+      "loss": 0.4925,
+      "step": 3858
+    },
+    {
+      "epoch": 0.2837708655048165,
+      "grad_norm": 1.8531080484390259,
+      "learning_rate": 0.00016810078095867413,
+      "loss": 0.5178,
+      "step": 3859
+    },
+    {
+      "epoch": 0.2838444003235532,
+      "grad_norm": 1.2041015625,
+      "learning_rate": 0.0001680833390048693,
+      "loss": 0.5029,
+      "step": 3860
+    },
+    {
+      "epoch": 0.28391793514228986,
+      "grad_norm": 1.1998194456100464,
+      "learning_rate": 0.00016806589318930573,
+      "loss": 0.4771,
+      "step": 3861
+    },
+    {
+      "epoch": 0.28399146996102653,
+      "grad_norm": 1.3878636360168457,
+      "learning_rate": 0.00016804844351297278,
+      "loss": 0.493,
+      "step": 3862
+    },
+    {
+      "epoch": 0.2840650047797632,
+      "grad_norm": 1.00780189037323,
+      "learning_rate": 0.00016803098997686036,
+      "loss": 0.4803,
+      "step": 3863
+    },
+    {
+      "epoch": 0.2841385395984999,
+      "grad_norm": 1.6254162788391113,
+      "learning_rate": 0.00016801353258195833,
+      "loss": 0.5054,
+      "step": 3864
+    },
+    {
+      "epoch": 0.28421207441723656,
+      "grad_norm": 1.1874650716781616,
+      "learning_rate": 0.00016799607132925702,
+      "loss": 0.5195,
+      "step": 3865
+    },
+    {
+      "epoch": 0.28428560923597324,
+      "grad_norm": 1.5156925916671753,
+      "learning_rate": 0.00016797860621974673,
+      "loss": 0.4794,
+      "step": 3866
+    },
+    {
+      "epoch": 0.2843591440547099,
+      "grad_norm": 1.43695867061615,
+      "learning_rate": 0.00016796113725441817,
+      "loss": 0.4727,
+      "step": 3867
+    },
+    {
+      "epoch": 0.2844326788734466,
+      "grad_norm": 1.6245036125183105,
+      "learning_rate": 0.00016794366443426218,
+      "loss": 0.4958,
+      "step": 3868
+    },
+    {
+      "epoch": 0.28450621369218326,
+      "grad_norm": 1.5398176908493042,
+      "learning_rate": 0.00016792618776026986,
+      "loss": 0.4769,
+      "step": 3869
+    },
+    {
+      "epoch": 0.28457974851091994,
+      "grad_norm": 1.4199116230010986,
+      "learning_rate": 0.0001679087072334325,
+      "loss": 0.5332,
+      "step": 3870
+    },
+    {
+      "epoch": 0.2846532833296566,
+      "grad_norm": 1.1752541065216064,
+      "learning_rate": 0.00016789122285474156,
+      "loss": 0.497,
+      "step": 3871
+    },
+    {
+      "epoch": 0.2847268181483933,
+      "grad_norm": 1.4500588178634644,
+      "learning_rate": 0.00016787373462518883,
+      "loss": 0.4724,
+      "step": 3872
+    },
+    {
+      "epoch": 0.2848003529671299,
+      "grad_norm": 1.2521796226501465,
+      "learning_rate": 0.00016785624254576624,
+      "loss": 0.479,
+      "step": 3873
+    },
+    {
+      "epoch": 0.2848738877858666,
+      "grad_norm": 1.6141328811645508,
+      "learning_rate": 0.00016783874661746595,
+      "loss": 0.4721,
+      "step": 3874
+    },
+    {
+      "epoch": 0.28494742260460326,
+      "grad_norm": 1.4918538331985474,
+      "learning_rate": 0.00016782124684128034,
+      "loss": 0.524,
+      "step": 3875
+    },
+    {
+      "epoch": 0.28502095742333994,
+      "grad_norm": 1.5394470691680908,
+      "learning_rate": 0.0001678037432182021,
+      "loss": 0.4837,
+      "step": 3876
+    },
+    {
+      "epoch": 0.2850944922420766,
+      "grad_norm": 1.6188768148422241,
+      "learning_rate": 0.00016778623574922394,
+      "loss": 0.4932,
+      "step": 3877
+    },
+    {
+      "epoch": 0.2851680270608133,
+      "grad_norm": 1.1438161134719849,
+      "learning_rate": 0.00016776872443533897,
+      "loss": 0.4863,
+      "step": 3878
+    },
+    {
+      "epoch": 0.28524156187954997,
+      "grad_norm": 1.1117298603057861,
+      "learning_rate": 0.0001677512092775404,
+      "loss": 0.4758,
+      "step": 3879
+    },
+    {
+      "epoch": 0.28531509669828664,
+      "grad_norm": 1.6951490640640259,
+      "learning_rate": 0.00016773369027682177,
+      "loss": 0.5159,
+      "step": 3880
+    },
+    {
+      "epoch": 0.2853886315170233,
+      "grad_norm": 1.5638628005981445,
+      "learning_rate": 0.00016771616743417673,
+      "loss": 0.4863,
+      "step": 3881
+    },
+    {
+      "epoch": 0.28546216633576,
+      "grad_norm": 1.5177173614501953,
+      "learning_rate": 0.00016769864075059923,
+      "loss": 0.4745,
+      "step": 3882
+    },
+    {
+      "epoch": 0.28553570115449667,
+      "grad_norm": 1.4265996217727661,
+      "learning_rate": 0.00016768111022708338,
+      "loss": 0.467,
+      "step": 3883
+    },
+    {
+      "epoch": 0.28560923597323334,
+      "grad_norm": 1.174867868423462,
+      "learning_rate": 0.00016766357586462348,
+      "loss": 0.4692,
+      "step": 3884
+    },
+    {
+      "epoch": 0.28568277079197,
+      "grad_norm": 0.7991251349449158,
+      "learning_rate": 0.0001676460376642142,
+      "loss": 0.4925,
+      "step": 3885
+    },
+    {
+      "epoch": 0.2857563056107067,
+      "grad_norm": 1.3809529542922974,
+      "learning_rate": 0.00016762849562685026,
+      "loss": 0.5062,
+      "step": 3886
+    },
+    {
+      "epoch": 0.2858298404294433,
+      "grad_norm": 0.6512576937675476,
+      "learning_rate": 0.00016761094975352667,
+      "loss": 0.4813,
+      "step": 3887
+    },
+    {
+      "epoch": 0.28590337524818,
+      "grad_norm": 1.4632203578948975,
+      "learning_rate": 0.0001675934000452387,
+      "loss": 0.4945,
+      "step": 3888
+    },
+    {
+      "epoch": 0.28597691006691667,
+      "grad_norm": 0.6968827843666077,
+      "learning_rate": 0.0001675758465029817,
+      "loss": 0.4658,
+      "step": 3889
+    },
+    {
+      "epoch": 0.28605044488565334,
+      "grad_norm": 1.7206695079803467,
+      "learning_rate": 0.00016755828912775136,
+      "loss": 0.4509,
+      "step": 3890
+    },
+    {
+      "epoch": 0.28612397970439,
+      "grad_norm": 1.0140941143035889,
+      "learning_rate": 0.0001675407279205436,
+      "loss": 0.491,
+      "step": 3891
+    },
+    {
+      "epoch": 0.2861975145231267,
+      "grad_norm": 2.2142131328582764,
+      "learning_rate": 0.00016752316288235446,
+      "loss": 0.4768,
+      "step": 3892
+    },
+    {
+      "epoch": 0.28627104934186337,
+      "grad_norm": 2.415309429168701,
+      "learning_rate": 0.00016750559401418024,
+      "loss": 0.4782,
+      "step": 3893
+    },
+    {
+      "epoch": 0.28634458416060005,
+      "grad_norm": 0.7315730452537537,
+      "learning_rate": 0.00016748802131701752,
+      "loss": 0.4964,
+      "step": 3894
+    },
+    {
+      "epoch": 0.2864181189793367,
+      "grad_norm": 2.1363441944122314,
+      "learning_rate": 0.000167470444791863,
+      "loss": 0.4636,
+      "step": 3895
+    },
+    {
+      "epoch": 0.2864916537980734,
+      "grad_norm": 1.668032169342041,
+      "learning_rate": 0.00016745286443971364,
+      "loss": 0.4883,
+      "step": 3896
+    },
+    {
+      "epoch": 0.2865651886168101,
+      "grad_norm": 1.7672983407974243,
+      "learning_rate": 0.00016743528026156662,
+      "loss": 0.474,
+      "step": 3897
+    },
+    {
+      "epoch": 0.28663872343554675,
+      "grad_norm": 2.253821849822998,
+      "learning_rate": 0.00016741769225841932,
+      "loss": 0.4857,
+      "step": 3898
+    },
+    {
+      "epoch": 0.2867122582542834,
+      "grad_norm": 0.9108845591545105,
+      "learning_rate": 0.0001674001004312694,
+      "loss": 0.5035,
+      "step": 3899
+    },
+    {
+      "epoch": 0.2867857930730201,
+      "grad_norm": 2.0449795722961426,
+      "learning_rate": 0.00016738250478111466,
+      "loss": 0.4659,
+      "step": 3900
+    },
+    {
+      "epoch": 0.2868593278917567,
+      "grad_norm": 2.038248300552368,
+      "learning_rate": 0.00016736490530895313,
+      "loss": 0.4921,
+      "step": 3901
+    },
+    {
+      "epoch": 0.2869328627104934,
+      "grad_norm": 0.8962936401367188,
+      "learning_rate": 0.00016734730201578302,
+      "loss": 0.4941,
+      "step": 3902
+    },
+    {
+      "epoch": 0.2870063975292301,
+      "grad_norm": 1.6845381259918213,
+      "learning_rate": 0.00016732969490260293,
+      "loss": 0.4847,
+      "step": 3903
+    },
+    {
+      "epoch": 0.28707993234796675,
+      "grad_norm": 1.4575388431549072,
+      "learning_rate": 0.00016731208397041146,
+      "loss": 0.5078,
+      "step": 3904
+    },
+    {
+      "epoch": 0.2871534671667034,
+      "grad_norm": 1.4688003063201904,
+      "learning_rate": 0.00016729446922020754,
+      "loss": 0.5088,
+      "step": 3905
+    },
+    {
+      "epoch": 0.2872270019854401,
+      "grad_norm": 2.1236109733581543,
+      "learning_rate": 0.00016727685065299035,
+      "loss": 0.5007,
+      "step": 3906
+    },
+    {
+      "epoch": 0.2873005368041768,
+      "grad_norm": 2.0285420417785645,
+      "learning_rate": 0.00016725922826975917,
+      "loss": 0.4646,
+      "step": 3907
+    },
+    {
+      "epoch": 0.28737407162291345,
+      "grad_norm": 2.1739213466644287,
+      "learning_rate": 0.00016724160207151357,
+      "loss": 0.5438,
+      "step": 3908
+    },
+    {
+      "epoch": 0.2874476064416501,
+      "grad_norm": 1.7348607778549194,
+      "learning_rate": 0.00016722397205925336,
+      "loss": 0.4998,
+      "step": 3909
+    },
+    {
+      "epoch": 0.2875211412603868,
+      "grad_norm": 1.9512079954147339,
+      "learning_rate": 0.0001672063382339785,
+      "loss": 0.4939,
+      "step": 3910
+    },
+    {
+      "epoch": 0.2875946760791235,
+      "grad_norm": 1.4393388032913208,
+      "learning_rate": 0.0001671887005966892,
+      "loss": 0.485,
+      "step": 3911
+    },
+    {
+      "epoch": 0.28766821089786015,
+      "grad_norm": 2.153761625289917,
+      "learning_rate": 0.0001671710591483859,
+      "loss": 0.5193,
+      "step": 3912
+    },
+    {
+      "epoch": 0.28774174571659683,
+      "grad_norm": 1.4736535549163818,
+      "learning_rate": 0.00016715341389006923,
+      "loss": 0.5157,
+      "step": 3913
+    },
+    {
+      "epoch": 0.2878152805353335,
+      "grad_norm": 1.729118824005127,
+      "learning_rate": 0.00016713576482274006,
+      "loss": 0.5004,
+      "step": 3914
+    },
+    {
+      "epoch": 0.2878888153540701,
+      "grad_norm": 1.4492710828781128,
+      "learning_rate": 0.00016711811194739947,
+      "loss": 0.4898,
+      "step": 3915
+    },
+    {
+      "epoch": 0.2879623501728068,
+      "grad_norm": 1.2077593803405762,
+      "learning_rate": 0.00016710045526504872,
+      "loss": 0.476,
+      "step": 3916
+    },
+    {
+      "epoch": 0.2880358849915435,
+      "grad_norm": 1.248091459274292,
+      "learning_rate": 0.00016708279477668935,
+      "loss": 0.5459,
+      "step": 3917
+    },
+    {
+      "epoch": 0.28810941981028015,
+      "grad_norm": 1.2795183658599854,
+      "learning_rate": 0.00016706513048332305,
+      "loss": 0.4965,
+      "step": 3918
+    },
+    {
+      "epoch": 0.28818295462901683,
+      "grad_norm": 0.9942612648010254,
+      "learning_rate": 0.0001670474623859518,
+      "loss": 0.4846,
+      "step": 3919
+    },
+    {
+      "epoch": 0.2882564894477535,
+      "grad_norm": 1.1391254663467407,
+      "learning_rate": 0.00016702979048557767,
+      "loss": 0.4804,
+      "step": 3920
+    },
+    {
+      "epoch": 0.2883300242664902,
+      "grad_norm": 0.8974140882492065,
+      "learning_rate": 0.00016701211478320314,
+      "loss": 0.468,
+      "step": 3921
+    },
+    {
+      "epoch": 0.28840355908522686,
+      "grad_norm": 1.1354339122772217,
+      "learning_rate": 0.00016699443527983071,
+      "loss": 0.4721,
+      "step": 3922
+    },
+    {
+      "epoch": 0.28847709390396353,
+      "grad_norm": 0.7518997192382812,
+      "learning_rate": 0.00016697675197646324,
+      "loss": 0.4905,
+      "step": 3923
+    },
+    {
+      "epoch": 0.2885506287227002,
+      "grad_norm": 1.5885963439941406,
+      "learning_rate": 0.0001669590648741037,
+      "loss": 0.5131,
+      "step": 3924
+    },
+    {
+      "epoch": 0.2886241635414369,
+      "grad_norm": 0.9759964942932129,
+      "learning_rate": 0.00016694137397375532,
+      "loss": 0.4937,
+      "step": 3925
+    },
+    {
+      "epoch": 0.28869769836017356,
+      "grad_norm": 2.286334991455078,
+      "learning_rate": 0.00016692367927642154,
+      "loss": 0.4916,
+      "step": 3926
+    },
+    {
+      "epoch": 0.28877123317891024,
+      "grad_norm": 2.2576072216033936,
+      "learning_rate": 0.00016690598078310609,
+      "loss": 0.4707,
+      "step": 3927
+    },
+    {
+      "epoch": 0.2888447679976469,
+      "grad_norm": 0.8425034880638123,
+      "learning_rate": 0.0001668882784948128,
+      "loss": 0.5118,
+      "step": 3928
+    },
+    {
+      "epoch": 0.28891830281638353,
+      "grad_norm": 1.771156668663025,
+      "learning_rate": 0.00016687057241254573,
+      "loss": 0.4828,
+      "step": 3929
+    },
+    {
+      "epoch": 0.2889918376351202,
+      "grad_norm": 1.1538059711456299,
+      "learning_rate": 0.00016685286253730923,
+      "loss": 0.4787,
+      "step": 3930
+    },
+    {
+      "epoch": 0.2890653724538569,
+      "grad_norm": 1.6275345087051392,
+      "learning_rate": 0.00016683514887010783,
+      "loss": 0.4832,
+      "step": 3931
+    },
+    {
+      "epoch": 0.28913890727259356,
+      "grad_norm": 1.5904288291931152,
+      "learning_rate": 0.00016681743141194625,
+      "loss": 0.468,
+      "step": 3932
+    },
+    {
+      "epoch": 0.28921244209133024,
+      "grad_norm": 1.148384690284729,
+      "learning_rate": 0.00016679971016382944,
+      "loss": 0.4651,
+      "step": 3933
+    },
+    {
+      "epoch": 0.2892859769100669,
+      "grad_norm": 1.253398060798645,
+      "learning_rate": 0.00016678198512676254,
+      "loss": 0.5008,
+      "step": 3934
+    },
+    {
+      "epoch": 0.2893595117288036,
+      "grad_norm": 1.4770537614822388,
+      "learning_rate": 0.00016676425630175103,
+      "loss": 0.4901,
+      "step": 3935
+    },
+    {
+      "epoch": 0.28943304654754026,
+      "grad_norm": 1.0323021411895752,
+      "learning_rate": 0.0001667465236898004,
+      "loss": 0.5023,
+      "step": 3936
+    },
+    {
+      "epoch": 0.28950658136627694,
+      "grad_norm": 1.427350640296936,
+      "learning_rate": 0.00016672878729191655,
+      "loss": 0.475,
+      "step": 3937
+    },
+    {
+      "epoch": 0.2895801161850136,
+      "grad_norm": 0.7827684283256531,
+      "learning_rate": 0.0001667110471091054,
+      "loss": 0.482,
+      "step": 3938
+    },
+    {
+      "epoch": 0.2896536510037503,
+      "grad_norm": 1.312795639038086,
+      "learning_rate": 0.00016669330314237328,
+      "loss": 0.4427,
+      "step": 3939
+    },
+    {
+      "epoch": 0.28972718582248697,
+      "grad_norm": 0.7890863418579102,
+      "learning_rate": 0.00016667555539272665,
+      "loss": 0.4854,
+      "step": 3940
+    },
+    {
+      "epoch": 0.28980072064122364,
+      "grad_norm": 1.4752572774887085,
+      "learning_rate": 0.00016665780386117214,
+      "loss": 0.5156,
+      "step": 3941
+    },
+    {
+      "epoch": 0.2898742554599603,
+      "grad_norm": 0.8016797304153442,
+      "learning_rate": 0.00016664004854871663,
+      "loss": 0.4839,
+      "step": 3942
+    },
+    {
+      "epoch": 0.28994779027869694,
+      "grad_norm": 1.0593185424804688,
+      "learning_rate": 0.00016662228945636725,
+      "loss": 0.4886,
+      "step": 3943
+    },
+    {
+      "epoch": 0.2900213250974336,
+      "grad_norm": 0.9003704786300659,
+      "learning_rate": 0.00016660452658513131,
+      "loss": 0.4638,
+      "step": 3944
+    },
+    {
+      "epoch": 0.2900948599161703,
+      "grad_norm": 1.0194860696792603,
+      "learning_rate": 0.0001665867599360163,
+      "loss": 0.4743,
+      "step": 3945
+    },
+    {
+      "epoch": 0.29016839473490696,
+      "grad_norm": 1.2885366678237915,
+      "learning_rate": 0.00016656898951003,
+      "loss": 0.5004,
+      "step": 3946
+    },
+    {
+      "epoch": 0.29024192955364364,
+      "grad_norm": 1.0486128330230713,
+      "learning_rate": 0.0001665512153081804,
+      "loss": 0.4736,
+      "step": 3947
+    },
+    {
+      "epoch": 0.2903154643723803,
+      "grad_norm": 1.1769037246704102,
+      "learning_rate": 0.00016653343733147563,
+      "loss": 0.4721,
+      "step": 3948
+    },
+    {
+      "epoch": 0.290388999191117,
+      "grad_norm": 1.0825755596160889,
+      "learning_rate": 0.00016651565558092405,
+      "loss": 0.5075,
+      "step": 3949
+    },
+    {
+      "epoch": 0.29046253400985367,
+      "grad_norm": 1.1731011867523193,
+      "learning_rate": 0.0001664978700575343,
+      "loss": 0.4644,
+      "step": 3950
+    },
+    {
+      "epoch": 0.29053606882859034,
+      "grad_norm": 1.0502817630767822,
+      "learning_rate": 0.00016648008076231515,
+      "loss": 0.5048,
+      "step": 3951
+    },
+    {
+      "epoch": 0.290609603647327,
+      "grad_norm": 0.771177351474762,
+      "learning_rate": 0.00016646228769627572,
+      "loss": 0.4888,
+      "step": 3952
+    },
+    {
+      "epoch": 0.2906831384660637,
+      "grad_norm": 1.2222900390625,
+      "learning_rate": 0.00016644449086042516,
+      "loss": 0.4931,
+      "step": 3953
+    },
+    {
+      "epoch": 0.29075667328480037,
+      "grad_norm": 0.922599196434021,
+      "learning_rate": 0.00016642669025577296,
+      "loss": 0.5117,
+      "step": 3954
+    },
+    {
+      "epoch": 0.29083020810353705,
+      "grad_norm": 1.287070631980896,
+      "learning_rate": 0.0001664088858833288,
+      "loss": 0.4769,
+      "step": 3955
+    },
+    {
+      "epoch": 0.2909037429222737,
+      "grad_norm": 0.975594162940979,
+      "learning_rate": 0.00016639107774410252,
+      "loss": 0.4679,
+      "step": 3956
+    },
+    {
+      "epoch": 0.29097727774101034,
+      "grad_norm": 1.4489250183105469,
+      "learning_rate": 0.00016637326583910427,
+      "loss": 0.4567,
+      "step": 3957
+    },
+    {
+      "epoch": 0.291050812559747,
+      "grad_norm": 1.0873459577560425,
+      "learning_rate": 0.00016635545016934432,
+      "loss": 0.4825,
+      "step": 3958
+    },
+    {
+      "epoch": 0.2911243473784837,
+      "grad_norm": 1.4296886920928955,
+      "learning_rate": 0.0001663376307358332,
+      "loss": 0.4644,
+      "step": 3959
+    },
+    {
+      "epoch": 0.29119788219722037,
+      "grad_norm": 1.0964332818984985,
+      "learning_rate": 0.00016631980753958168,
+      "loss": 0.4794,
+      "step": 3960
+    },
+    {
+      "epoch": 0.29127141701595705,
+      "grad_norm": 1.2162470817565918,
+      "learning_rate": 0.00016630198058160067,
+      "loss": 0.5081,
+      "step": 3961
+    },
+    {
+      "epoch": 0.2913449518346937,
+      "grad_norm": 0.9814333915710449,
+      "learning_rate": 0.00016628414986290135,
+      "loss": 0.4804,
+      "step": 3962
+    },
+    {
+      "epoch": 0.2914184866534304,
+      "grad_norm": 1.3956778049468994,
+      "learning_rate": 0.00016626631538449514,
+      "loss": 0.5073,
+      "step": 3963
+    },
+    {
+      "epoch": 0.2914920214721671,
+      "grad_norm": 0.986216127872467,
+      "learning_rate": 0.00016624847714739353,
+      "loss": 0.4663,
+      "step": 3964
+    },
+    {
+      "epoch": 0.29156555629090375,
+      "grad_norm": 1.1923093795776367,
+      "learning_rate": 0.0001662306351526084,
+      "loss": 0.5023,
+      "step": 3965
+    },
+    {
+      "epoch": 0.2916390911096404,
+      "grad_norm": 0.9055904150009155,
+      "learning_rate": 0.00016621278940115176,
+      "loss": 0.4719,
+      "step": 3966
+    },
+    {
+      "epoch": 0.2917126259283771,
+      "grad_norm": 1.1203372478485107,
+      "learning_rate": 0.0001661949398940358,
+      "loss": 0.5047,
+      "step": 3967
+    },
+    {
+      "epoch": 0.2917861607471138,
+      "grad_norm": 1.2655417919158936,
+      "learning_rate": 0.00016617708663227302,
+      "loss": 0.4926,
+      "step": 3968
+    },
+    {
+      "epoch": 0.29185969556585045,
+      "grad_norm": 0.6573495864868164,
+      "learning_rate": 0.00016615922961687606,
+      "loss": 0.4505,
+      "step": 3969
+    },
+    {
+      "epoch": 0.29193323038458713,
+      "grad_norm": 0.8819087147712708,
+      "learning_rate": 0.00016614136884885773,
+      "loss": 0.4918,
+      "step": 3970
+    },
+    {
+      "epoch": 0.29200676520332375,
+      "grad_norm": 0.6180109977722168,
+      "learning_rate": 0.00016612350432923116,
+      "loss": 0.4755,
+      "step": 3971
+    },
+    {
+      "epoch": 0.2920803000220604,
+      "grad_norm": 0.7238010168075562,
+      "learning_rate": 0.00016610563605900968,
+      "loss": 0.4831,
+      "step": 3972
+    },
+    {
+      "epoch": 0.2921538348407971,
+      "grad_norm": 0.8367425203323364,
+      "learning_rate": 0.00016608776403920675,
+      "loss": 0.5013,
+      "step": 3973
+    },
+    {
+      "epoch": 0.2922273696595338,
+      "grad_norm": 0.6738390922546387,
+      "learning_rate": 0.00016606988827083606,
+      "loss": 0.5015,
+      "step": 3974
+    },
+    {
+      "epoch": 0.29230090447827045,
+      "grad_norm": 1.045078992843628,
+      "learning_rate": 0.0001660520087549116,
+      "loss": 0.4959,
+      "step": 3975
+    },
+    {
+      "epoch": 0.2923744392970071,
+      "grad_norm": 0.7512903809547424,
+      "learning_rate": 0.00016603412549244753,
+      "loss": 0.48,
+      "step": 3976
+    },
+    {
+      "epoch": 0.2924479741157438,
+      "grad_norm": 1.0069949626922607,
+      "learning_rate": 0.00016601623848445812,
+      "loss": 0.4964,
+      "step": 3977
+    },
+    {
+      "epoch": 0.2925215089344805,
+      "grad_norm": 1.1169507503509521,
+      "learning_rate": 0.00016599834773195805,
+      "loss": 0.4862,
+      "step": 3978
+    },
+    {
+      "epoch": 0.29259504375321715,
+      "grad_norm": 0.8313238620758057,
+      "learning_rate": 0.000165980453235962,
+      "loss": 0.4425,
+      "step": 3979
+    },
+    {
+      "epoch": 0.29266857857195383,
+      "grad_norm": 0.9662755727767944,
+      "learning_rate": 0.00016596255499748503,
+      "loss": 0.4788,
+      "step": 3980
+    },
+    {
+      "epoch": 0.2927421133906905,
+      "grad_norm": 1.2546521425247192,
+      "learning_rate": 0.0001659446530175423,
+      "loss": 0.4589,
+      "step": 3981
+    },
+    {
+      "epoch": 0.2928156482094272,
+      "grad_norm": 0.7718207836151123,
+      "learning_rate": 0.00016592674729714932,
+      "loss": 0.4821,
+      "step": 3982
+    },
+    {
+      "epoch": 0.29288918302816386,
+      "grad_norm": 1.6346189975738525,
+      "learning_rate": 0.00016590883783732161,
+      "loss": 0.4776,
+      "step": 3983
+    },
+    {
+      "epoch": 0.29296271784690053,
+      "grad_norm": 0.8829125761985779,
+      "learning_rate": 0.00016589092463907508,
+      "loss": 0.4586,
+      "step": 3984
+    },
+    {
+      "epoch": 0.29303625266563715,
+      "grad_norm": 2.22739839553833,
+      "learning_rate": 0.00016587300770342576,
+      "loss": 0.503,
+      "step": 3985
+    },
+    {
+      "epoch": 0.29310978748437383,
+      "grad_norm": 2.0824241638183594,
+      "learning_rate": 0.00016585508703138992,
+      "loss": 0.4755,
+      "step": 3986
+    },
+    {
+      "epoch": 0.2931833223031105,
+      "grad_norm": 0.990425169467926,
+      "learning_rate": 0.00016583716262398403,
+      "loss": 0.5059,
+      "step": 3987
+    },
+    {
+      "epoch": 0.2932568571218472,
+      "grad_norm": 1.2694857120513916,
+      "learning_rate": 0.00016581923448222483,
+      "loss": 0.5284,
+      "step": 3988
+    },
+    {
+      "epoch": 0.29333039194058386,
+      "grad_norm": 1.3578261137008667,
+      "learning_rate": 0.0001658013026071292,
+      "loss": 0.4983,
+      "step": 3989
+    },
+    {
+      "epoch": 0.29340392675932053,
+      "grad_norm": 0.6327720880508423,
+      "learning_rate": 0.00016578336699971417,
+      "loss": 0.4521,
+      "step": 3990
+    },
+    {
+      "epoch": 0.2934774615780572,
+      "grad_norm": 1.2872252464294434,
+      "learning_rate": 0.00016576542766099717,
+      "loss": 0.4826,
+      "step": 3991
+    },
+    {
+      "epoch": 0.2935509963967939,
+      "grad_norm": 0.6018351912498474,
+      "learning_rate": 0.00016574748459199573,
+      "loss": 0.5017,
+      "step": 3992
+    },
+    {
+      "epoch": 0.29362453121553056,
+      "grad_norm": 1.3501354455947876,
+      "learning_rate": 0.00016572953779372759,
+      "loss": 0.4664,
+      "step": 3993
+    },
+    {
+      "epoch": 0.29369806603426724,
+      "grad_norm": 0.6996738314628601,
+      "learning_rate": 0.0001657115872672107,
+      "loss": 0.4938,
+      "step": 3994
+    },
+    {
+      "epoch": 0.2937716008530039,
+      "grad_norm": 0.9381628632545471,
+      "learning_rate": 0.00016569363301346317,
+      "loss": 0.4699,
+      "step": 3995
+    },
+    {
+      "epoch": 0.2938451356717406,
+      "grad_norm": 1.100730299949646,
+      "learning_rate": 0.00016567567503350352,
+      "loss": 0.5077,
+      "step": 3996
+    },
+    {
+      "epoch": 0.29391867049047726,
+      "grad_norm": 0.6493743658065796,
+      "learning_rate": 0.0001656577133283502,
+      "loss": 0.4691,
+      "step": 3997
+    },
+    {
+      "epoch": 0.29399220530921394,
+      "grad_norm": 1.230330228805542,
+      "learning_rate": 0.00016563974789902215,
+      "loss": 0.4853,
+      "step": 3998
+    },
+    {
+      "epoch": 0.29406574012795056,
+      "grad_norm": 0.7331894040107727,
+      "learning_rate": 0.0001656217787465383,
+      "loss": 0.509,
+      "step": 3999
+    },
+    {
+      "epoch": 0.29413927494668723,
+      "grad_norm": 1.095902442932129,
+      "learning_rate": 0.0001656038058719179,
+      "loss": 0.4935,
+      "step": 4000
+    },
+    {
+      "epoch": 0.2942128097654239,
+      "grad_norm": 1.0273445844650269,
+      "learning_rate": 0.00016558582927618043,
+      "loss": 0.5035,
+      "step": 4001
+    },
+    {
+      "epoch": 0.2942863445841606,
+      "grad_norm": 0.623331606388092,
+      "learning_rate": 0.00016556784896034547,
+      "loss": 0.5047,
+      "step": 4002
+    },
+    {
+      "epoch": 0.29435987940289726,
+      "grad_norm": 0.7919434309005737,
+      "learning_rate": 0.00016554986492543292,
+      "loss": 0.5277,
+      "step": 4003
+    },
+    {
+      "epoch": 0.29443341422163394,
+      "grad_norm": 0.5588165521621704,
+      "learning_rate": 0.0001655318771724629,
+      "loss": 0.4962,
+      "step": 4004
+    },
+    {
+      "epoch": 0.2945069490403706,
+      "grad_norm": 0.9268544316291809,
+      "learning_rate": 0.0001655138857024556,
+      "loss": 0.4987,
+      "step": 4005
+    },
+    {
+      "epoch": 0.2945804838591073,
+      "grad_norm": 0.7482399344444275,
+      "learning_rate": 0.0001654958905164316,
+      "loss": 0.4877,
+      "step": 4006
+    },
+    {
+      "epoch": 0.29465401867784397,
+      "grad_norm": 0.7629526257514954,
+      "learning_rate": 0.00016547789161541155,
+      "loss": 0.5073,
+      "step": 4007
+    },
+    {
+      "epoch": 0.29472755349658064,
+      "grad_norm": 0.8034685850143433,
+      "learning_rate": 0.00016545988900041638,
+      "loss": 0.4953,
+      "step": 4008
+    },
+    {
+      "epoch": 0.2948010883153173,
+      "grad_norm": 1.0710382461547852,
+      "learning_rate": 0.0001654418826724672,
+      "loss": 0.5105,
+      "step": 4009
+    },
+    {
+      "epoch": 0.294874623134054,
+      "grad_norm": 0.6375782489776611,
+      "learning_rate": 0.0001654238726325854,
+      "loss": 0.4578,
+      "step": 4010
+    },
+    {
+      "epoch": 0.29494815795279067,
+      "grad_norm": 0.7787714004516602,
+      "learning_rate": 0.00016540585888179252,
+      "loss": 0.5351,
+      "step": 4011
+    },
+    {
+      "epoch": 0.29502169277152734,
+      "grad_norm": 0.7147220373153687,
+      "learning_rate": 0.00016538784142111027,
+      "loss": 0.4962,
+      "step": 4012
+    },
+    {
+      "epoch": 0.29509522759026396,
+      "grad_norm": 0.7772040963172913,
+      "learning_rate": 0.00016536982025156067,
+      "loss": 0.4743,
+      "step": 4013
+    },
+    {
+      "epoch": 0.29516876240900064,
+      "grad_norm": 1.1730377674102783,
+      "learning_rate": 0.00016535179537416586,
+      "loss": 0.5074,
+      "step": 4014
+    },
+    {
+      "epoch": 0.2952422972277373,
+      "grad_norm": 0.8446546792984009,
+      "learning_rate": 0.00016533376678994825,
+      "loss": 0.4905,
+      "step": 4015
+    },
+    {
+      "epoch": 0.295315832046474,
+      "grad_norm": 0.7236410975456238,
+      "learning_rate": 0.00016531573449993044,
+      "loss": 0.4788,
+      "step": 4016
+    },
+    {
+      "epoch": 0.29538936686521067,
+      "grad_norm": 1.0691238641738892,
+      "learning_rate": 0.00016529769850513523,
+      "loss": 0.4688,
+      "step": 4017
+    },
+    {
+      "epoch": 0.29546290168394734,
+      "grad_norm": 0.6055645942687988,
+      "learning_rate": 0.0001652796588065857,
+      "loss": 0.4807,
+      "step": 4018
+    },
+    {
+      "epoch": 0.295536436502684,
+      "grad_norm": 1.0710723400115967,
+      "learning_rate": 0.00016526161540530496,
+      "loss": 0.4737,
+      "step": 4019
+    },
+    {
+      "epoch": 0.2956099713214207,
+      "grad_norm": 1.2520232200622559,
+      "learning_rate": 0.00016524356830231656,
+      "loss": 0.4584,
+      "step": 4020
+    },
+    {
+      "epoch": 0.29568350614015737,
+      "grad_norm": 0.662691056728363,
+      "learning_rate": 0.00016522551749864408,
+      "loss": 0.4897,
+      "step": 4021
+    },
+    {
+      "epoch": 0.29575704095889405,
+      "grad_norm": 0.9432357549667358,
+      "learning_rate": 0.00016520746299531145,
+      "loss": 0.4873,
+      "step": 4022
+    },
+    {
+      "epoch": 0.2958305757776307,
+      "grad_norm": 1.2052909135818481,
+      "learning_rate": 0.0001651894047933427,
+      "loss": 0.4678,
+      "step": 4023
+    },
+    {
+      "epoch": 0.2959041105963674,
+      "grad_norm": 0.6343536972999573,
+      "learning_rate": 0.0001651713428937621,
+      "loss": 0.4724,
+      "step": 4024
+    },
+    {
+      "epoch": 0.2959776454151041,
+      "grad_norm": 0.7663953304290771,
+      "learning_rate": 0.00016515327729759414,
+      "loss": 0.4847,
+      "step": 4025
+    },
+    {
+      "epoch": 0.29605118023384075,
+      "grad_norm": 1.0064215660095215,
+      "learning_rate": 0.00016513520800586355,
+      "loss": 0.4557,
+      "step": 4026
+    },
+    {
+      "epoch": 0.29612471505257737,
+      "grad_norm": 0.7006927728652954,
+      "learning_rate": 0.0001651171350195952,
+      "loss": 0.4822,
+      "step": 4027
+    },
+    {
+      "epoch": 0.29619824987131405,
+      "grad_norm": 0.7510364651679993,
+      "learning_rate": 0.00016509905833981425,
+      "loss": 0.45,
+      "step": 4028
+    },
+    {
+      "epoch": 0.2962717846900507,
+      "grad_norm": 1.2822514772415161,
+      "learning_rate": 0.00016508097796754601,
+      "loss": 0.4819,
+      "step": 4029
+    },
+    {
+      "epoch": 0.2963453195087874,
+      "grad_norm": 0.6758435368537903,
+      "learning_rate": 0.00016506289390381604,
+      "loss": 0.4645,
+      "step": 4030
+    },
+    {
+      "epoch": 0.2964188543275241,
+      "grad_norm": 0.991280198097229,
+      "learning_rate": 0.00016504480614965003,
+      "loss": 0.4533,
+      "step": 4031
+    },
+    {
+      "epoch": 0.29649238914626075,
+      "grad_norm": 1.6326483488082886,
+      "learning_rate": 0.00016502671470607395,
+      "loss": 0.4856,
+      "step": 4032
+    },
+    {
+      "epoch": 0.2965659239649974,
+      "grad_norm": 0.8233618140220642,
+      "learning_rate": 0.000165008619574114,
+      "loss": 0.4687,
+      "step": 4033
+    },
+    {
+      "epoch": 0.2966394587837341,
+      "grad_norm": 1.808336615562439,
+      "learning_rate": 0.0001649905207547966,
+      "loss": 0.4775,
+      "step": 4034
+    },
+    {
+      "epoch": 0.2967129936024708,
+      "grad_norm": 0.9362839460372925,
+      "learning_rate": 0.00016497241824914825,
+      "loss": 0.4484,
+      "step": 4035
+    },
+    {
+      "epoch": 0.29678652842120745,
+      "grad_norm": 1.9210366010665894,
+      "learning_rate": 0.00016495431205819572,
+      "loss": 0.4662,
+      "step": 4036
+    },
+    {
+      "epoch": 0.2968600632399441,
+      "grad_norm": 1.7087767124176025,
+      "learning_rate": 0.0001649362021829661,
+      "loss": 0.5075,
+      "step": 4037
+    },
+    {
+      "epoch": 0.2969335980586808,
+      "grad_norm": 1.728674292564392,
+      "learning_rate": 0.00016491808862448656,
+      "loss": 0.4981,
+      "step": 4038
+    },
+    {
+      "epoch": 0.2970071328774175,
+      "grad_norm": 1.7036902904510498,
+      "learning_rate": 0.0001648999713837845,
+      "loss": 0.4827,
+      "step": 4039
+    },
+    {
+      "epoch": 0.29708066769615415,
+      "grad_norm": 1.3314744234085083,
+      "learning_rate": 0.0001648818504618876,
+      "loss": 0.497,
+      "step": 4040
+    },
+    {
+      "epoch": 0.2971542025148908,
+      "grad_norm": 1.1268997192382812,
+      "learning_rate": 0.00016486372585982365,
+      "loss": 0.4596,
+      "step": 4041
+    },
+    {
+      "epoch": 0.29722773733362745,
+      "grad_norm": 1.651979923248291,
+      "learning_rate": 0.00016484559757862073,
+      "loss": 0.4793,
+      "step": 4042
+    },
+    {
+      "epoch": 0.2973012721523641,
+      "grad_norm": 1.1786456108093262,
+      "learning_rate": 0.0001648274656193071,
+      "loss": 0.4715,
+      "step": 4043
+    },
+    {
+      "epoch": 0.2973748069711008,
+      "grad_norm": 1.7437801361083984,
+      "learning_rate": 0.00016480932998291117,
+      "loss": 0.4686,
+      "step": 4044
+    },
+    {
+      "epoch": 0.2974483417898375,
+      "grad_norm": 1.3382220268249512,
+      "learning_rate": 0.00016479119067046164,
+      "loss": 0.475,
+      "step": 4045
+    },
+    {
+      "epoch": 0.29752187660857415,
+      "grad_norm": 1.7869176864624023,
+      "learning_rate": 0.00016477304768298742,
+      "loss": 0.4743,
+      "step": 4046
+    },
+    {
+      "epoch": 0.29759541142731083,
+      "grad_norm": 1.5022248029708862,
+      "learning_rate": 0.00016475490102151757,
+      "loss": 0.4979,
+      "step": 4047
+    },
+    {
+      "epoch": 0.2976689462460475,
+      "grad_norm": 1.6318644285202026,
+      "learning_rate": 0.00016473675068708142,
+      "loss": 0.4776,
+      "step": 4048
+    },
+    {
+      "epoch": 0.2977424810647842,
+      "grad_norm": 1.5303605794906616,
+      "learning_rate": 0.00016471859668070843,
+      "loss": 0.4573,
+      "step": 4049
+    },
+    {
+      "epoch": 0.29781601588352086,
+      "grad_norm": 1.4122116565704346,
+      "learning_rate": 0.0001647004390034283,
+      "loss": 0.4553,
+      "step": 4050
+    },
+    {
+      "epoch": 0.29788955070225753,
+      "grad_norm": 1.5065429210662842,
+      "learning_rate": 0.00016468227765627106,
+      "loss": 0.4639,
+      "step": 4051
+    },
+    {
+      "epoch": 0.2979630855209942,
+      "grad_norm": 1.473036527633667,
+      "learning_rate": 0.00016466411264026674,
+      "loss": 0.4778,
+      "step": 4052
+    },
+    {
+      "epoch": 0.2980366203397309,
+      "grad_norm": 1.3690768480300903,
+      "learning_rate": 0.0001646459439564457,
+      "loss": 0.4783,
+      "step": 4053
+    },
+    {
+      "epoch": 0.29811015515846756,
+      "grad_norm": 1.4849756956100464,
+      "learning_rate": 0.00016462777160583848,
+      "loss": 0.5043,
+      "step": 4054
+    },
+    {
+      "epoch": 0.2981836899772042,
+      "grad_norm": 0.9628124237060547,
+      "learning_rate": 0.00016460959558947587,
+      "loss": 0.4876,
+      "step": 4055
+    },
+    {
+      "epoch": 0.29825722479594086,
+      "grad_norm": 1.7026469707489014,
+      "learning_rate": 0.00016459141590838882,
+      "loss": 0.4521,
+      "step": 4056
+    },
+    {
+      "epoch": 0.29833075961467753,
+      "grad_norm": 1.2482730150222778,
+      "learning_rate": 0.00016457323256360845,
+      "loss": 0.4786,
+      "step": 4057
+    },
+    {
+      "epoch": 0.2984042944334142,
+      "grad_norm": 1.6816673278808594,
+      "learning_rate": 0.00016455504555616622,
+      "loss": 0.4726,
+      "step": 4058
+    },
+    {
+      "epoch": 0.2984778292521509,
+      "grad_norm": 1.4776170253753662,
+      "learning_rate": 0.0001645368548870937,
+      "loss": 0.4619,
+      "step": 4059
+    },
+    {
+      "epoch": 0.29855136407088756,
+      "grad_norm": 1.5142205953598022,
+      "learning_rate": 0.00016451866055742264,
+      "loss": 0.4507,
+      "step": 4060
+    },
+    {
+      "epoch": 0.29862489888962424,
+      "grad_norm": 1.4071942567825317,
+      "learning_rate": 0.00016450046256818505,
+      "loss": 0.4652,
+      "step": 4061
+    },
+    {
+      "epoch": 0.2986984337083609,
+      "grad_norm": 1.6587491035461426,
+      "learning_rate": 0.00016448226092041318,
+      "loss": 0.4711,
+      "step": 4062
+    },
+    {
+      "epoch": 0.2987719685270976,
+      "grad_norm": 1.397538661956787,
+      "learning_rate": 0.00016446405561513938,
+      "loss": 0.4823,
+      "step": 4063
+    },
+    {
+      "epoch": 0.29884550334583426,
+      "grad_norm": 1.2922557592391968,
+      "learning_rate": 0.00016444584665339636,
+      "loss": 0.4659,
+      "step": 4064
+    },
+    {
+      "epoch": 0.29891903816457094,
+      "grad_norm": 0.9409567713737488,
+      "learning_rate": 0.00016442763403621687,
+      "loss": 0.4816,
+      "step": 4065
+    },
+    {
+      "epoch": 0.2989925729833076,
+      "grad_norm": 1.204483151435852,
+      "learning_rate": 0.00016440941776463402,
+      "loss": 0.4502,
+      "step": 4066
+    },
+    {
+      "epoch": 0.2990661078020443,
+      "grad_norm": 0.7119899988174438,
+      "learning_rate": 0.00016439119783968098,
+      "loss": 0.4915,
+      "step": 4067
+    },
+    {
+      "epoch": 0.29913964262078097,
+      "grad_norm": 1.016917109489441,
+      "learning_rate": 0.00016437297426239125,
+      "loss": 0.4941,
+      "step": 4068
+    },
+    {
+      "epoch": 0.2992131774395176,
+      "grad_norm": 0.9210928082466125,
+      "learning_rate": 0.00016435474703379853,
+      "loss": 0.465,
+      "step": 4069
+    },
+    {
+      "epoch": 0.29928671225825426,
+      "grad_norm": 0.6635233759880066,
+      "learning_rate": 0.0001643365161549366,
+      "loss": 0.507,
+      "step": 4070
+    },
+    {
+      "epoch": 0.29936024707699094,
+      "grad_norm": 1.0080891847610474,
+      "learning_rate": 0.00016431828162683957,
+      "loss": 0.4641,
+      "step": 4071
+    },
+    {
+      "epoch": 0.2994337818957276,
+      "grad_norm": 0.862221896648407,
+      "learning_rate": 0.00016430004345054176,
+      "loss": 0.4542,
+      "step": 4072
+    },
+    {
+      "epoch": 0.2995073167144643,
+      "grad_norm": 0.541054368019104,
+      "learning_rate": 0.0001642818016270776,
+      "loss": 0.4911,
+      "step": 4073
+    },
+    {
+      "epoch": 0.29958085153320096,
+      "grad_norm": 0.8342422842979431,
+      "learning_rate": 0.00016426355615748183,
+      "loss": 0.4366,
+      "step": 4074
+    },
+    {
+      "epoch": 0.29965438635193764,
+      "grad_norm": 0.5923808813095093,
+      "learning_rate": 0.00016424530704278934,
+      "loss": 0.4697,
+      "step": 4075
+    },
+    {
+      "epoch": 0.2997279211706743,
+      "grad_norm": 0.8654104471206665,
+      "learning_rate": 0.00016422705428403522,
+      "loss": 0.464,
+      "step": 4076
+    },
+    {
+      "epoch": 0.299801455989411,
+      "grad_norm": 0.9707460999488831,
+      "learning_rate": 0.00016420879788225478,
+      "loss": 0.4512,
+      "step": 4077
+    },
+    {
+      "epoch": 0.29987499080814767,
+      "grad_norm": 0.77558833360672,
+      "learning_rate": 0.0001641905378384836,
+      "loss": 0.4891,
+      "step": 4078
+    },
+    {
+      "epoch": 0.29994852562688434,
+      "grad_norm": 0.8037930130958557,
+      "learning_rate": 0.00016417227415375734,
+      "loss": 0.4727,
+      "step": 4079
+    },
+    {
+      "epoch": 0.300022060445621,
+      "grad_norm": 1.1428978443145752,
+      "learning_rate": 0.00016415400682911197,
+      "loss": 0.4683,
+      "step": 4080
+    },
+    {
+      "epoch": 0.3000955952643577,
+      "grad_norm": 1.0636868476867676,
+      "learning_rate": 0.00016413573586558365,
+      "loss": 0.4919,
+      "step": 4081
+    },
+    {
+      "epoch": 0.30016913008309437,
+      "grad_norm": 0.6523507833480835,
+      "learning_rate": 0.0001641174612642087,
+      "loss": 0.4926,
+      "step": 4082
+    },
+    {
+      "epoch": 0.300242664901831,
+      "grad_norm": 1.0817303657531738,
+      "learning_rate": 0.00016409918302602366,
+      "loss": 0.5232,
+      "step": 4083
+    },
+    {
+      "epoch": 0.30031619972056767,
+      "grad_norm": 1.23158860206604,
+      "learning_rate": 0.00016408090115206534,
+      "loss": 0.4754,
+      "step": 4084
+    },
+    {
+      "epoch": 0.30038973453930434,
+      "grad_norm": 0.57576584815979,
+      "learning_rate": 0.00016406261564337068,
+      "loss": 0.5162,
+      "step": 4085
+    },
+    {
+      "epoch": 0.300463269358041,
+      "grad_norm": 1.1294108629226685,
+      "learning_rate": 0.00016404432650097683,
+      "loss": 0.4796,
+      "step": 4086
+    },
+    {
+      "epoch": 0.3005368041767777,
+      "grad_norm": 0.7896365523338318,
+      "learning_rate": 0.0001640260337259212,
+      "loss": 0.4623,
+      "step": 4087
+    },
+    {
+      "epoch": 0.30061033899551437,
+      "grad_norm": 1.0218861103057861,
+      "learning_rate": 0.00016400773731924136,
+      "loss": 0.4956,
+      "step": 4088
+    },
+    {
+      "epoch": 0.30068387381425105,
+      "grad_norm": 1.3996071815490723,
+      "learning_rate": 0.00016398943728197513,
+      "loss": 0.452,
+      "step": 4089
+    },
+    {
+      "epoch": 0.3007574086329877,
+      "grad_norm": 0.8085007667541504,
+      "learning_rate": 0.00016397113361516045,
+      "loss": 0.4749,
+      "step": 4090
+    },
+    {
+      "epoch": 0.3008309434517244,
+      "grad_norm": 1.477413296699524,
+      "learning_rate": 0.0001639528263198356,
+      "loss": 0.4884,
+      "step": 4091
+    },
+    {
+      "epoch": 0.3009044782704611,
+      "grad_norm": 0.7536367177963257,
+      "learning_rate": 0.00016393451539703895,
+      "loss": 0.4679,
+      "step": 4092
+    },
+    {
+      "epoch": 0.30097801308919775,
+      "grad_norm": 1.4738478660583496,
+      "learning_rate": 0.0001639162008478091,
+      "loss": 0.4486,
+      "step": 4093
+    },
+    {
+      "epoch": 0.3010515479079344,
+      "grad_norm": 0.935185432434082,
+      "learning_rate": 0.00016389788267318488,
+      "loss": 0.4907,
+      "step": 4094
+    },
+    {
+      "epoch": 0.3011250827266711,
+      "grad_norm": 2.096895456314087,
+      "learning_rate": 0.00016387956087420534,
+      "loss": 0.473,
+      "step": 4095
+    },
+    {
+      "epoch": 0.3011986175454078,
+      "grad_norm": 2.013341188430786,
+      "learning_rate": 0.00016386123545190966,
+      "loss": 0.4682,
+      "step": 4096
+    },
+    {
+      "epoch": 0.3012721523641444,
+      "grad_norm": 1.6970038414001465,
+      "learning_rate": 0.00016384290640733732,
+      "loss": 0.4816,
+      "step": 4097
+    },
+    {
+      "epoch": 0.3013456871828811,
+      "grad_norm": 1.8003673553466797,
+      "learning_rate": 0.00016382457374152794,
+      "loss": 0.4505,
+      "step": 4098
+    },
+    {
+      "epoch": 0.30141922200161775,
+      "grad_norm": 1.0768725872039795,
+      "learning_rate": 0.00016380623745552143,
+      "loss": 0.4476,
+      "step": 4099
+    },
+    {
+      "epoch": 0.3014927568203544,
+      "grad_norm": 1.2851226329803467,
+      "learning_rate": 0.00016378789755035775,
+      "loss": 0.4654,
+      "step": 4100
+    },
+    {
+      "epoch": 0.3015662916390911,
+      "grad_norm": 1.2232463359832764,
+      "learning_rate": 0.00016376955402707722,
+      "loss": 0.4567,
+      "step": 4101
+    },
+    {
+      "epoch": 0.3016398264578278,
+      "grad_norm": 0.873881995677948,
+      "learning_rate": 0.00016375120688672028,
+      "loss": 0.4662,
+      "step": 4102
+    },
+    {
+      "epoch": 0.30171336127656445,
+      "grad_norm": 1.5675560235977173,
+      "learning_rate": 0.0001637328561303276,
+      "loss": 0.4754,
+      "step": 4103
+    },
+    {
+      "epoch": 0.3017868960953011,
+      "grad_norm": 0.9666794538497925,
+      "learning_rate": 0.00016371450175894009,
+      "loss": 0.4258,
+      "step": 4104
+    },
+    {
+      "epoch": 0.3018604309140378,
+      "grad_norm": 2.1662437915802,
+      "learning_rate": 0.00016369614377359878,
+      "loss": 0.4864,
+      "step": 4105
+    },
+    {
+      "epoch": 0.3019339657327745,
+      "grad_norm": 1.7167857885360718,
+      "learning_rate": 0.00016367778217534499,
+      "loss": 0.4559,
+      "step": 4106
+    },
+    {
+      "epoch": 0.30200750055151115,
+      "grad_norm": 1.2666889429092407,
+      "learning_rate": 0.00016365941696522014,
+      "loss": 0.4471,
+      "step": 4107
+    },
+    {
+      "epoch": 0.30208103537024783,
+      "grad_norm": 1.259871482849121,
+      "learning_rate": 0.000163641048144266,
+      "loss": 0.4518,
+      "step": 4108
+    },
+    {
+      "epoch": 0.3021545701889845,
+      "grad_norm": 1.4042062759399414,
+      "learning_rate": 0.0001636226757135245,
+      "loss": 0.4421,
+      "step": 4109
+    },
+    {
+      "epoch": 0.3022281050077212,
+      "grad_norm": 1.0427985191345215,
+      "learning_rate": 0.00016360429967403762,
+      "loss": 0.4439,
+      "step": 4110
+    },
+    {
+      "epoch": 0.3023016398264578,
+      "grad_norm": 1.6941567659378052,
+      "learning_rate": 0.00016358592002684778,
+      "loss": 0.4448,
+      "step": 4111
+    },
+    {
+      "epoch": 0.3023751746451945,
+      "grad_norm": 1.4518945217132568,
+      "learning_rate": 0.00016356753677299746,
+      "loss": 0.4644,
+      "step": 4112
+    },
+    {
+      "epoch": 0.30244870946393115,
+      "grad_norm": 1.5594053268432617,
+      "learning_rate": 0.00016354914991352932,
+      "loss": 0.4492,
+      "step": 4113
+    },
+    {
+      "epoch": 0.30252224428266783,
+      "grad_norm": 1.4215949773788452,
+      "learning_rate": 0.00016353075944948637,
+      "loss": 0.4357,
+      "step": 4114
+    },
+    {
+      "epoch": 0.3025957791014045,
+      "grad_norm": 1.709458589553833,
+      "learning_rate": 0.00016351236538191167,
+      "loss": 0.4513,
+      "step": 4115
+    },
+    {
+      "epoch": 0.3026693139201412,
+      "grad_norm": 1.4862632751464844,
+      "learning_rate": 0.00016349396771184863,
+      "loss": 0.4469,
+      "step": 4116
+    },
+    {
+      "epoch": 0.30274284873887786,
+      "grad_norm": 1.3762661218643188,
+      "learning_rate": 0.00016347556644034068,
+      "loss": 0.4371,
+      "step": 4117
+    },
+    {
+      "epoch": 0.30281638355761453,
+      "grad_norm": 1.3540276288986206,
+      "learning_rate": 0.00016345716156843164,
+      "loss": 0.4623,
+      "step": 4118
+    },
+    {
+      "epoch": 0.3028899183763512,
+      "grad_norm": 1.1975075006484985,
+      "learning_rate": 0.0001634387530971654,
+      "loss": 0.4487,
+      "step": 4119
+    },
+    {
+      "epoch": 0.3029634531950879,
+      "grad_norm": 1.1728171110153198,
+      "learning_rate": 0.00016342034102758617,
+      "loss": 0.4509,
+      "step": 4120
+    },
+    {
+      "epoch": 0.30303698801382456,
+      "grad_norm": 1.4867773056030273,
+      "learning_rate": 0.0001634019253607382,
+      "loss": 0.4201,
+      "step": 4121
+    },
+    {
+      "epoch": 0.30311052283256124,
+      "grad_norm": 1.3294686079025269,
+      "learning_rate": 0.0001633835060976662,
+      "loss": 0.4327,
+      "step": 4122
+    },
+    {
+      "epoch": 0.3031840576512979,
+      "grad_norm": 1.83938729763031,
+      "learning_rate": 0.00016336508323941482,
+      "loss": 0.4327,
+      "step": 4123
+    },
+    {
+      "epoch": 0.3032575924700346,
+      "grad_norm": 1.7036106586456299,
+      "learning_rate": 0.00016334665678702904,
+      "loss": 0.4501,
+      "step": 4124
+    },
+    {
+      "epoch": 0.3033311272887712,
+      "grad_norm": 1.173882007598877,
+      "learning_rate": 0.000163328226741554,
+      "loss": 0.4536,
+      "step": 4125
+    },
+    {
+      "epoch": 0.3034046621075079,
+      "grad_norm": 1.189680814743042,
+      "learning_rate": 0.00016330979310403513,
+      "loss": 0.4439,
+      "step": 4126
+    },
+    {
+      "epoch": 0.30347819692624456,
+      "grad_norm": 1.5253885984420776,
+      "learning_rate": 0.00016329135587551797,
+      "loss": 0.4284,
+      "step": 4127
+    },
+    {
+      "epoch": 0.30355173174498123,
+      "grad_norm": 1.2449208498001099,
+      "learning_rate": 0.00016327291505704837,
+      "loss": 0.456,
+      "step": 4128
+    },
+    {
+      "epoch": 0.3036252665637179,
+      "grad_norm": 1.5662304162979126,
+      "learning_rate": 0.0001632544706496722,
+      "loss": 0.4315,
+      "step": 4129
+    },
+    {
+      "epoch": 0.3036988013824546,
+      "grad_norm": 1.4366717338562012,
+      "learning_rate": 0.0001632360226544357,
+      "loss": 0.4449,
+      "step": 4130
+    },
+    {
+      "epoch": 0.30377233620119126,
+      "grad_norm": 1.2113827466964722,
+      "learning_rate": 0.00016321757107238524,
+      "loss": 0.4645,
+      "step": 4131
+    },
+    {
+      "epoch": 0.30384587101992794,
+      "grad_norm": 0.9963471293449402,
+      "learning_rate": 0.00016319911590456746,
+      "loss": 0.459,
+      "step": 4132
+    },
+    {
+      "epoch": 0.3039194058386646,
+      "grad_norm": 1.3947728872299194,
+      "learning_rate": 0.00016318065715202916,
+      "loss": 0.4578,
+      "step": 4133
+    },
+    {
+      "epoch": 0.3039929406574013,
+      "grad_norm": 1.0621788501739502,
+      "learning_rate": 0.00016316219481581727,
+      "loss": 0.4596,
+      "step": 4134
+    },
+    {
+      "epoch": 0.30406647547613797,
+      "grad_norm": 1.2732822895050049,
+      "learning_rate": 0.00016314372889697905,
+      "loss": 0.4424,
+      "step": 4135
+    },
+    {
+      "epoch": 0.30414001029487464,
+      "grad_norm": 1.0720270872116089,
+      "learning_rate": 0.00016312525939656187,
+      "loss": 0.4537,
+      "step": 4136
+    },
+    {
+      "epoch": 0.3042135451136113,
+      "grad_norm": 1.413480520248413,
+      "learning_rate": 0.00016310678631561336,
+      "loss": 0.4158,
+      "step": 4137
+    },
+    {
+      "epoch": 0.304287079932348,
+      "grad_norm": 1.0149235725402832,
+      "learning_rate": 0.00016308830965518133,
+      "loss": 0.4594,
+      "step": 4138
+    },
+    {
+      "epoch": 0.3043606147510846,
+      "grad_norm": 1.665711522102356,
+      "learning_rate": 0.00016306982941631382,
+      "loss": 0.4413,
+      "step": 4139
+    },
+    {
+      "epoch": 0.3044341495698213,
+      "grad_norm": 1.368626594543457,
+      "learning_rate": 0.000163051345600059,
+      "loss": 0.4563,
+      "step": 4140
+    },
+    {
+      "epoch": 0.30450768438855796,
+      "grad_norm": 1.5273698568344116,
+      "learning_rate": 0.00016303285820746535,
+      "loss": 0.454,
+      "step": 4141
+    },
+    {
+      "epoch": 0.30458121920729464,
+      "grad_norm": 1.4362424612045288,
+      "learning_rate": 0.00016301436723958145,
+      "loss": 0.4804,
+      "step": 4142
+    },
+    {
+      "epoch": 0.3046547540260313,
+      "grad_norm": 1.3119878768920898,
+      "learning_rate": 0.00016299587269745608,
+      "loss": 0.4304,
+      "step": 4143
+    },
+    {
+      "epoch": 0.304728288844768,
+      "grad_norm": 1.1439580917358398,
+      "learning_rate": 0.0001629773745821384,
+      "loss": 0.4571,
+      "step": 4144
+    },
+    {
+      "epoch": 0.30480182366350467,
+      "grad_norm": 1.5458954572677612,
+      "learning_rate": 0.00016295887289467754,
+      "loss": 0.4499,
+      "step": 4145
+    },
+    {
+      "epoch": 0.30487535848224134,
+      "grad_norm": 1.254917860031128,
+      "learning_rate": 0.00016294036763612297,
+      "loss": 0.4624,
+      "step": 4146
+    },
+    {
+      "epoch": 0.304948893300978,
+      "grad_norm": 1.6065398454666138,
+      "learning_rate": 0.00016292185880752434,
+      "loss": 0.4598,
+      "step": 4147
+    },
+    {
+      "epoch": 0.3050224281197147,
+      "grad_norm": 1.4873071908950806,
+      "learning_rate": 0.00016290334640993147,
+      "loss": 0.4161,
+      "step": 4148
+    },
+    {
+      "epoch": 0.30509596293845137,
+      "grad_norm": 1.0563849210739136,
+      "learning_rate": 0.0001628848304443944,
+      "loss": 0.4368,
+      "step": 4149
+    },
+    {
+      "epoch": 0.30516949775718805,
+      "grad_norm": 1.2378889322280884,
+      "learning_rate": 0.00016286631091196337,
+      "loss": 0.4561,
+      "step": 4150
+    },
+    {
+      "epoch": 0.3052430325759247,
+      "grad_norm": 1.097695231437683,
+      "learning_rate": 0.00016284778781368883,
+      "loss": 0.4814,
+      "step": 4151
+    },
+    {
+      "epoch": 0.3053165673946614,
+      "grad_norm": 1.2276848554611206,
+      "learning_rate": 0.00016282926115062147,
+      "loss": 0.4581,
+      "step": 4152
+    },
+    {
+      "epoch": 0.305390102213398,
+      "grad_norm": 0.8539280891418457,
+      "learning_rate": 0.00016281073092381208,
+      "loss": 0.4783,
+      "step": 4153
+    },
+    {
+      "epoch": 0.3054636370321347,
+      "grad_norm": 1.3891030550003052,
+      "learning_rate": 0.0001627921971343118,
+      "loss": 0.4434,
+      "step": 4154
+    },
+    {
+      "epoch": 0.30553717185087137,
+      "grad_norm": 0.8928256630897522,
+      "learning_rate": 0.00016277365978317182,
+      "loss": 0.453,
+      "step": 4155
+    },
+    {
+      "epoch": 0.30561070666960805,
+      "grad_norm": 1.6141053438186646,
+      "learning_rate": 0.00016275511887144355,
+      "loss": 0.4819,
+      "step": 4156
+    },
+    {
+      "epoch": 0.3056842414883447,
+      "grad_norm": 1.181441307067871,
+      "learning_rate": 0.00016273657440017876,
+      "loss": 0.4404,
+      "step": 4157
+    },
+    {
+      "epoch": 0.3057577763070814,
+      "grad_norm": 1.4124315977096558,
+      "learning_rate": 0.00016271802637042923,
+      "loss": 0.4644,
+      "step": 4158
+    },
+    {
+      "epoch": 0.3058313111258181,
+      "grad_norm": 1.1270520687103271,
+      "learning_rate": 0.0001626994747832471,
+      "loss": 0.4578,
+      "step": 4159
+    },
+    {
+      "epoch": 0.30590484594455475,
+      "grad_norm": 1.0523431301116943,
+      "learning_rate": 0.00016268091963968453,
+      "loss": 0.4401,
+      "step": 4160
+    },
+    {
+      "epoch": 0.3059783807632914,
+      "grad_norm": 1.0527286529541016,
+      "learning_rate": 0.0001626623609407941,
+      "loss": 0.468,
+      "step": 4161
+    },
+    {
+      "epoch": 0.3060519155820281,
+      "grad_norm": 1.0796469449996948,
+      "learning_rate": 0.0001626437986876284,
+      "loss": 0.458,
+      "step": 4162
+    },
+    {
+      "epoch": 0.3061254504007648,
+      "grad_norm": 0.6752147674560547,
+      "learning_rate": 0.00016262523288124037,
+      "loss": 0.4398,
+      "step": 4163
+    },
+    {
+      "epoch": 0.30619898521950145,
+      "grad_norm": 1.0130952596664429,
+      "learning_rate": 0.00016260666352268299,
+      "loss": 0.425,
+      "step": 4164
+    },
+    {
+      "epoch": 0.3062725200382381,
+      "grad_norm": 0.7473921179771423,
+      "learning_rate": 0.0001625880906130096,
+      "loss": 0.4699,
+      "step": 4165
+    },
+    {
+      "epoch": 0.3063460548569748,
+      "grad_norm": 1.2307355403900146,
+      "learning_rate": 0.00016256951415327366,
+      "loss": 0.4422,
+      "step": 4166
+    },
+    {
+      "epoch": 0.3064195896757114,
+      "grad_norm": 0.6285234689712524,
+      "learning_rate": 0.00016255093414452881,
+      "loss": 0.4362,
+      "step": 4167
+    },
+    {
+      "epoch": 0.3064931244944481,
+      "grad_norm": 1.0858805179595947,
+      "learning_rate": 0.00016253235058782904,
+      "loss": 0.4348,
+      "step": 4168
+    },
+    {
+      "epoch": 0.3065666593131848,
+      "grad_norm": 0.6790233850479126,
+      "learning_rate": 0.0001625137634842283,
+      "loss": 0.4105,
+      "step": 4169
+    },
+    {
+      "epoch": 0.30664019413192145,
+      "grad_norm": 1.0587186813354492,
+      "learning_rate": 0.00016249517283478093,
+      "loss": 0.4527,
+      "step": 4170
+    },
+    {
+      "epoch": 0.3067137289506581,
+      "grad_norm": 0.5956820249557495,
+      "learning_rate": 0.0001624765786405414,
+      "loss": 0.4514,
+      "step": 4171
+    },
+    {
+      "epoch": 0.3067872637693948,
+      "grad_norm": 0.9287803769111633,
+      "learning_rate": 0.0001624579809025644,
+      "loss": 0.4424,
+      "step": 4172
+    },
+    {
+      "epoch": 0.3068607985881315,
+      "grad_norm": 0.5248487591743469,
+      "learning_rate": 0.0001624393796219048,
+      "loss": 0.4284,
+      "step": 4173
+    },
+    {
+      "epoch": 0.30693433340686815,
+      "grad_norm": 0.9756878018379211,
+      "learning_rate": 0.0001624207747996177,
+      "loss": 0.4506,
+      "step": 4174
+    },
+    {
+      "epoch": 0.30700786822560483,
+      "grad_norm": 0.5983788371086121,
+      "learning_rate": 0.0001624021664367584,
+      "loss": 0.4201,
+      "step": 4175
+    },
+    {
+      "epoch": 0.3070814030443415,
+      "grad_norm": 0.996790885925293,
+      "learning_rate": 0.00016238355453438233,
+      "loss": 0.4239,
+      "step": 4176
+    },
+    {
+      "epoch": 0.3071549378630782,
+      "grad_norm": 0.9168099761009216,
+      "learning_rate": 0.00016236493909354522,
+      "loss": 0.4366,
+      "step": 4177
+    },
+    {
+      "epoch": 0.30722847268181486,
+      "grad_norm": 0.7353887557983398,
+      "learning_rate": 0.00016234632011530295,
+      "loss": 0.457,
+      "step": 4178
+    },
+    {
+      "epoch": 0.30730200750055153,
+      "grad_norm": 0.7494170665740967,
+      "learning_rate": 0.0001623276976007116,
+      "loss": 0.4332,
+      "step": 4179
+    },
+    {
+      "epoch": 0.3073755423192882,
+      "grad_norm": 0.7106029391288757,
+      "learning_rate": 0.00016230907155082749,
+      "loss": 0.4273,
+      "step": 4180
+    },
+    {
+      "epoch": 0.30744907713802483,
+      "grad_norm": 1.1592515707015991,
+      "learning_rate": 0.00016229044196670707,
+      "loss": 0.475,
+      "step": 4181
+    },
+    {
+      "epoch": 0.3075226119567615,
+      "grad_norm": 0.8390259742736816,
+      "learning_rate": 0.000162271808849407,
+      "loss": 0.4668,
+      "step": 4182
+    },
+    {
+      "epoch": 0.3075961467754982,
+      "grad_norm": 0.8350958228111267,
+      "learning_rate": 0.0001622531721999843,
+      "loss": 0.4542,
+      "step": 4183
+    },
+    {
+      "epoch": 0.30766968159423486,
+      "grad_norm": 0.6728086471557617,
+      "learning_rate": 0.00016223453201949593,
+      "loss": 0.4592,
+      "step": 4184
+    },
+    {
+      "epoch": 0.30774321641297153,
+      "grad_norm": 0.9165151715278625,
+      "learning_rate": 0.0001622158883089992,
+      "loss": 0.4392,
+      "step": 4185
+    },
+    {
+      "epoch": 0.3078167512317082,
+      "grad_norm": 0.6141794323921204,
+      "learning_rate": 0.00016219724106955168,
+      "loss": 0.4658,
+      "step": 4186
+    },
+    {
+      "epoch": 0.3078902860504449,
+      "grad_norm": 0.9238064885139465,
+      "learning_rate": 0.00016217859030221097,
+      "loss": 0.4395,
+      "step": 4187
+    },
+    {
+      "epoch": 0.30796382086918156,
+      "grad_norm": 0.8111700415611267,
+      "learning_rate": 0.000162159936008035,
+      "loss": 0.443,
+      "step": 4188
+    },
+    {
+      "epoch": 0.30803735568791824,
+      "grad_norm": 0.6610586047172546,
+      "learning_rate": 0.00016214127818808187,
+      "loss": 0.433,
+      "step": 4189
+    },
+    {
+      "epoch": 0.3081108905066549,
+      "grad_norm": 0.6914655566215515,
+      "learning_rate": 0.00016212261684340982,
+      "loss": 0.4319,
+      "step": 4190
+    },
+    {
+      "epoch": 0.3081844253253916,
+      "grad_norm": 0.5693867206573486,
+      "learning_rate": 0.00016210395197507744,
+      "loss": 0.4576,
+      "step": 4191
+    },
+    {
+      "epoch": 0.30825796014412826,
+      "grad_norm": 0.6683956384658813,
+      "learning_rate": 0.00016208528358414335,
+      "loss": 0.4699,
+      "step": 4192
+    },
+    {
+      "epoch": 0.30833149496286494,
+      "grad_norm": 0.8712949752807617,
+      "learning_rate": 0.00016206661167166646,
+      "loss": 0.4442,
+      "step": 4193
+    },
+    {
+      "epoch": 0.3084050297816016,
+      "grad_norm": 1.0093327760696411,
+      "learning_rate": 0.00016204793623870583,
+      "loss": 0.4487,
+      "step": 4194
+    },
+    {
+      "epoch": 0.30847856460033823,
+      "grad_norm": 0.6850917339324951,
+      "learning_rate": 0.00016202925728632078,
+      "loss": 0.4572,
+      "step": 4195
+    },
+    {
+      "epoch": 0.3085520994190749,
+      "grad_norm": 0.7471334934234619,
+      "learning_rate": 0.0001620105748155708,
+      "loss": 0.447,
+      "step": 4196
+    },
+    {
+      "epoch": 0.3086256342378116,
+      "grad_norm": 1.0820047855377197,
+      "learning_rate": 0.00016199188882751558,
+      "loss": 0.4419,
+      "step": 4197
+    },
+    {
+      "epoch": 0.30869916905654826,
+      "grad_norm": 0.5756519436836243,
+      "learning_rate": 0.000161973199323215,
+      "loss": 0.4739,
+      "step": 4198
+    },
+    {
+      "epoch": 0.30877270387528494,
+      "grad_norm": 1.1001005172729492,
+      "learning_rate": 0.00016195450630372915,
+      "loss": 0.4604,
+      "step": 4199
+    },
+    {
+      "epoch": 0.3088462386940216,
+      "grad_norm": 1.0235342979431152,
+      "learning_rate": 0.0001619358097701183,
+      "loss": 0.4569,
+      "step": 4200
+    },
+    {
+      "epoch": 0.3089197735127583,
+      "grad_norm": 0.6369898319244385,
+      "learning_rate": 0.00016191710972344298,
+      "loss": 0.4666,
+      "step": 4201
+    },
+    {
+      "epoch": 0.30899330833149496,
+      "grad_norm": 0.88750821352005,
+      "learning_rate": 0.00016189840616476385,
+      "loss": 0.4341,
+      "step": 4202
+    },
+    {
+      "epoch": 0.30906684315023164,
+      "grad_norm": 1.1891928911209106,
+      "learning_rate": 0.0001618796990951418,
+      "loss": 0.4309,
+      "step": 4203
+    },
+    {
+      "epoch": 0.3091403779689683,
+      "grad_norm": 0.5617215037345886,
+      "learning_rate": 0.00016186098851563791,
+      "loss": 0.4355,
+      "step": 4204
+    },
+    {
+      "epoch": 0.309213912787705,
+      "grad_norm": 1.101022720336914,
+      "learning_rate": 0.00016184227442731346,
+      "loss": 0.4559,
+      "step": 4205
+    },
+    {
+      "epoch": 0.30928744760644167,
+      "grad_norm": 1.167011022567749,
+      "learning_rate": 0.00016182355683122993,
+      "loss": 0.4626,
+      "step": 4206
+    },
+    {
+      "epoch": 0.30936098242517834,
+      "grad_norm": 0.7199912071228027,
+      "learning_rate": 0.00016180483572844905,
+      "loss": 0.4474,
+      "step": 4207
+    },
+    {
+      "epoch": 0.309434517243915,
+      "grad_norm": 0.9045674800872803,
+      "learning_rate": 0.00016178611112003264,
+      "loss": 0.4676,
+      "step": 4208
+    },
+    {
+      "epoch": 0.30950805206265164,
+      "grad_norm": 1.1348133087158203,
+      "learning_rate": 0.0001617673830070428,
+      "loss": 0.4595,
+      "step": 4209
+    },
+    {
+      "epoch": 0.3095815868813883,
+      "grad_norm": 0.6441000699996948,
+      "learning_rate": 0.00016174865139054182,
+      "loss": 0.4353,
+      "step": 4210
+    },
+    {
+      "epoch": 0.309655121700125,
+      "grad_norm": 0.7989594340324402,
+      "learning_rate": 0.00016172991627159215,
+      "loss": 0.4433,
+      "step": 4211
+    },
+    {
+      "epoch": 0.30972865651886167,
+      "grad_norm": 1.398503065109253,
+      "learning_rate": 0.0001617111776512565,
+      "loss": 0.462,
+      "step": 4212
+    },
+    {
+      "epoch": 0.30980219133759834,
+      "grad_norm": 0.5224777460098267,
+      "learning_rate": 0.0001616924355305977,
+      "loss": 0.4496,
+      "step": 4213
+    },
+    {
+      "epoch": 0.309875726156335,
+      "grad_norm": 1.3117369413375854,
+      "learning_rate": 0.00016167368991067886,
+      "loss": 0.4401,
+      "step": 4214
+    },
+    {
+      "epoch": 0.3099492609750717,
+      "grad_norm": 0.6871164441108704,
+      "learning_rate": 0.00016165494079256325,
+      "loss": 0.4459,
+      "step": 4215
+    },
+    {
+      "epoch": 0.31002279579380837,
+      "grad_norm": 1.1995311975479126,
+      "learning_rate": 0.00016163618817731433,
+      "loss": 0.471,
+      "step": 4216
+    },
+    {
+      "epoch": 0.31009633061254505,
+      "grad_norm": 1.0272560119628906,
+      "learning_rate": 0.00016161743206599577,
+      "loss": 0.4641,
+      "step": 4217
+    },
+    {
+      "epoch": 0.3101698654312817,
+      "grad_norm": 0.9552346467971802,
+      "learning_rate": 0.00016159867245967144,
+      "loss": 0.4423,
+      "step": 4218
+    },
+    {
+      "epoch": 0.3102434002500184,
+      "grad_norm": 1.5050556659698486,
+      "learning_rate": 0.0001615799093594054,
+      "loss": 0.469,
+      "step": 4219
+    },
+    {
+      "epoch": 0.3103169350687551,
+      "grad_norm": 1.0123496055603027,
+      "learning_rate": 0.0001615611427662619,
+      "loss": 0.4551,
+      "step": 4220
+    },
+    {
+      "epoch": 0.31039046988749175,
+      "grad_norm": 1.3164246082305908,
+      "learning_rate": 0.00016154237268130545,
+      "loss": 0.4678,
+      "step": 4221
+    },
+    {
+      "epoch": 0.3104640047062284,
+      "grad_norm": 1.2646297216415405,
+      "learning_rate": 0.00016152359910560064,
+      "loss": 0.4603,
+      "step": 4222
+    },
+    {
+      "epoch": 0.31053753952496504,
+      "grad_norm": 0.8999269604682922,
+      "learning_rate": 0.00016150482204021235,
+      "loss": 0.4911,
+      "step": 4223
+    },
+    {
+      "epoch": 0.3106110743437017,
+      "grad_norm": 1.5285078287124634,
+      "learning_rate": 0.00016148604148620568,
+      "loss": 0.441,
+      "step": 4224
+    },
+    {
+      "epoch": 0.3106846091624384,
+      "grad_norm": 0.8054295778274536,
+      "learning_rate": 0.0001614672574446458,
+      "loss": 0.4679,
+      "step": 4225
+    },
+    {
+      "epoch": 0.3107581439811751,
+      "grad_norm": 1.592097282409668,
+      "learning_rate": 0.00016144846991659824,
+      "loss": 0.4638,
+      "step": 4226
+    },
+    {
+      "epoch": 0.31083167879991175,
+      "grad_norm": 0.848820686340332,
+      "learning_rate": 0.00016142967890312866,
+      "loss": 0.4378,
+      "step": 4227
+    },
+    {
+      "epoch": 0.3109052136186484,
+      "grad_norm": 1.9518998861312866,
+      "learning_rate": 0.00016141088440530278,
+      "loss": 0.4763,
+      "step": 4228
+    },
+    {
+      "epoch": 0.3109787484373851,
+      "grad_norm": 1.434719204902649,
+      "learning_rate": 0.0001613920864241868,
+      "loss": 0.4747,
+      "step": 4229
+    },
+    {
+      "epoch": 0.3110522832561218,
+      "grad_norm": 2.046329975128174,
+      "learning_rate": 0.00016137328496084683,
+      "loss": 0.4446,
+      "step": 4230
+    },
+    {
+      "epoch": 0.31112581807485845,
+      "grad_norm": 1.896889090538025,
+      "learning_rate": 0.0001613544800163494,
+      "loss": 0.4421,
+      "step": 4231
+    },
+    {
+      "epoch": 0.3111993528935951,
+      "grad_norm": 1.1972570419311523,
+      "learning_rate": 0.0001613356715917611,
+      "loss": 0.4935,
+      "step": 4232
+    },
+    {
+      "epoch": 0.3112728877123318,
+      "grad_norm": 1.1166102886199951,
+      "learning_rate": 0.0001613168596881488,
+      "loss": 0.4548,
+      "step": 4233
+    },
+    {
+      "epoch": 0.3113464225310685,
+      "grad_norm": 1.6253892183303833,
+      "learning_rate": 0.0001612980443065795,
+      "loss": 0.4636,
+      "step": 4234
+    },
+    {
+      "epoch": 0.31141995734980515,
+      "grad_norm": 1.1187347173690796,
+      "learning_rate": 0.00016127922544812042,
+      "loss": 0.4597,
+      "step": 4235
+    },
+    {
+      "epoch": 0.31149349216854183,
+      "grad_norm": 1.9967730045318604,
+      "learning_rate": 0.00016126040311383898,
+      "loss": 0.4806,
+      "step": 4236
+    },
+    {
+      "epoch": 0.31156702698727845,
+      "grad_norm": 1.6540663242340088,
+      "learning_rate": 0.00016124157730480288,
+      "loss": 0.4566,
+      "step": 4237
+    },
+    {
+      "epoch": 0.3116405618060151,
+      "grad_norm": 1.6189906597137451,
+      "learning_rate": 0.00016122274802207986,
+      "loss": 0.4544,
+      "step": 4238
+    },
+    {
+      "epoch": 0.3117140966247518,
+      "grad_norm": 1.5429075956344604,
+      "learning_rate": 0.00016120391526673798,
+      "loss": 0.4657,
+      "step": 4239
+    },
+    {
+      "epoch": 0.3117876314434885,
+      "grad_norm": 1.4600998163223267,
+      "learning_rate": 0.00016118507903984538,
+      "loss": 0.44,
+      "step": 4240
+    },
+    {
+      "epoch": 0.31186116626222515,
+      "grad_norm": 1.2202861309051514,
+      "learning_rate": 0.00016116623934247056,
+      "loss": 0.4869,
+      "step": 4241
+    },
+    {
+      "epoch": 0.31193470108096183,
+      "grad_norm": 1.8223652839660645,
+      "learning_rate": 0.0001611473961756821,
+      "loss": 0.4762,
+      "step": 4242
+    },
+    {
+      "epoch": 0.3120082358996985,
+      "grad_norm": 1.4273614883422852,
+      "learning_rate": 0.0001611285495405488,
+      "loss": 0.4426,
+      "step": 4243
+    },
+    {
+      "epoch": 0.3120817707184352,
+      "grad_norm": 1.4206193685531616,
+      "learning_rate": 0.00016110969943813964,
+      "loss": 0.4472,
+      "step": 4244
+    },
+    {
+      "epoch": 0.31215530553717186,
+      "grad_norm": 1.3439115285873413,
+      "learning_rate": 0.00016109084586952386,
+      "loss": 0.4827,
+      "step": 4245
+    },
+    {
+      "epoch": 0.31222884035590853,
+      "grad_norm": 1.4745954275131226,
+      "learning_rate": 0.00016107198883577082,
+      "loss": 0.4352,
+      "step": 4246
+    },
+    {
+      "epoch": 0.3123023751746452,
+      "grad_norm": 1.1882961988449097,
+      "learning_rate": 0.0001610531283379501,
+      "loss": 0.4595,
+      "step": 4247
+    },
+    {
+      "epoch": 0.3123759099933819,
+      "grad_norm": 2.0872576236724854,
+      "learning_rate": 0.00016103426437713155,
+      "loss": 0.4603,
+      "step": 4248
+    },
+    {
+      "epoch": 0.31244944481211856,
+      "grad_norm": 1.4406657218933105,
+      "learning_rate": 0.00016101539695438508,
+      "loss": 0.4584,
+      "step": 4249
+    },
+    {
+      "epoch": 0.31252297963085524,
+      "grad_norm": 1.9070197343826294,
+      "learning_rate": 0.0001609965260707809,
+      "loss": 0.4618,
+      "step": 4250
+    },
+    {
+      "epoch": 0.31259651444959186,
+      "grad_norm": 1.860802173614502,
+      "learning_rate": 0.00016097765172738942,
+      "loss": 0.4437,
+      "step": 4251
+    },
+    {
+      "epoch": 0.31267004926832853,
+      "grad_norm": 1.0366183519363403,
+      "learning_rate": 0.00016095877392528117,
+      "loss": 0.4678,
+      "step": 4252
+    },
+    {
+      "epoch": 0.3127435840870652,
+      "grad_norm": 1.047163724899292,
+      "learning_rate": 0.00016093989266552692,
+      "loss": 0.4451,
+      "step": 4253
+    },
+    {
+      "epoch": 0.3128171189058019,
+      "grad_norm": 1.3992069959640503,
+      "learning_rate": 0.00016092100794919766,
+      "loss": 0.4341,
+      "step": 4254
+    },
+    {
+      "epoch": 0.31289065372453856,
+      "grad_norm": 0.8757414817810059,
+      "learning_rate": 0.0001609021197773645,
+      "loss": 0.4533,
+      "step": 4255
+    },
+    {
+      "epoch": 0.31296418854327523,
+      "grad_norm": 1.5579979419708252,
+      "learning_rate": 0.00016088322815109888,
+      "loss": 0.4757,
+      "step": 4256
+    },
+    {
+      "epoch": 0.3130377233620119,
+      "grad_norm": 0.8138765096664429,
+      "learning_rate": 0.0001608643330714723,
+      "loss": 0.4299,
+      "step": 4257
+    },
+    {
+      "epoch": 0.3131112581807486,
+      "grad_norm": 1.333807349205017,
+      "learning_rate": 0.00016084543453955647,
+      "loss": 0.4471,
+      "step": 4258
+    },
+    {
+      "epoch": 0.31318479299948526,
+      "grad_norm": 0.8290607929229736,
+      "learning_rate": 0.00016082653255642342,
+      "loss": 0.4716,
+      "step": 4259
+    },
+    {
+      "epoch": 0.31325832781822194,
+      "grad_norm": 1.6007872819900513,
+      "learning_rate": 0.00016080762712314525,
+      "loss": 0.4661,
+      "step": 4260
+    },
+    {
+      "epoch": 0.3133318626369586,
+      "grad_norm": 1.0164114236831665,
+      "learning_rate": 0.0001607887182407943,
+      "loss": 0.456,
+      "step": 4261
+    },
+    {
+      "epoch": 0.3134053974556953,
+      "grad_norm": 2.0310893058776855,
+      "learning_rate": 0.0001607698059104431,
+      "loss": 0.43,
+      "step": 4262
+    },
+    {
+      "epoch": 0.31347893227443197,
+      "grad_norm": 2.1255669593811035,
+      "learning_rate": 0.00016075089013316434,
+      "loss": 0.4368,
+      "step": 4263
+    },
+    {
+      "epoch": 0.31355246709316864,
+      "grad_norm": 0.961737334728241,
+      "learning_rate": 0.000160731970910031,
+      "loss": 0.4407,
+      "step": 4264
+    },
+    {
+      "epoch": 0.31362600191190526,
+      "grad_norm": 1.2326757907867432,
+      "learning_rate": 0.00016071304824211618,
+      "loss": 0.4399,
+      "step": 4265
+    },
+    {
+      "epoch": 0.31369953673064194,
+      "grad_norm": 1.0756665468215942,
+      "learning_rate": 0.00016069412213049318,
+      "loss": 0.4455,
+      "step": 4266
+    },
+    {
+      "epoch": 0.3137730715493786,
+      "grad_norm": 0.8929990530014038,
+      "learning_rate": 0.00016067519257623555,
+      "loss": 0.4629,
+      "step": 4267
+    },
+    {
+      "epoch": 0.3138466063681153,
+      "grad_norm": 1.1046075820922852,
+      "learning_rate": 0.00016065625958041694,
+      "loss": 0.4356,
+      "step": 4268
+    },
+    {
+      "epoch": 0.31392014118685196,
+      "grad_norm": 0.854168176651001,
+      "learning_rate": 0.00016063732314411126,
+      "loss": 0.4457,
+      "step": 4269
+    },
+    {
+      "epoch": 0.31399367600558864,
+      "grad_norm": 1.0407685041427612,
+      "learning_rate": 0.00016061838326839265,
+      "loss": 0.4654,
+      "step": 4270
+    },
+    {
+      "epoch": 0.3140672108243253,
+      "grad_norm": 1.2921258211135864,
+      "learning_rate": 0.00016059943995433534,
+      "loss": 0.4568,
+      "step": 4271
+    },
+    {
+      "epoch": 0.314140745643062,
+      "grad_norm": 0.6333062052726746,
+      "learning_rate": 0.00016058049320301384,
+      "loss": 0.4352,
+      "step": 4272
+    },
+    {
+      "epoch": 0.31421428046179867,
+      "grad_norm": 1.4957746267318726,
+      "learning_rate": 0.00016056154301550288,
+      "loss": 0.4638,
+      "step": 4273
+    },
+    {
+      "epoch": 0.31428781528053534,
+      "grad_norm": 0.6304306983947754,
+      "learning_rate": 0.00016054258939287726,
+      "loss": 0.4786,
+      "step": 4274
+    },
+    {
+      "epoch": 0.314361350099272,
+      "grad_norm": 0.9806075692176819,
+      "learning_rate": 0.00016052363233621207,
+      "loss": 0.4369,
+      "step": 4275
+    },
+    {
+      "epoch": 0.3144348849180087,
+      "grad_norm": 0.8126814961433411,
+      "learning_rate": 0.00016050467184658258,
+      "loss": 0.4407,
+      "step": 4276
+    },
+    {
+      "epoch": 0.31450841973674537,
+      "grad_norm": 0.5778896808624268,
+      "learning_rate": 0.00016048570792506425,
+      "loss": 0.4506,
+      "step": 4277
+    },
+    {
+      "epoch": 0.31458195455548205,
+      "grad_norm": 0.794487476348877,
+      "learning_rate": 0.00016046674057273273,
+      "loss": 0.4456,
+      "step": 4278
+    },
+    {
+      "epoch": 0.31465548937421867,
+      "grad_norm": 0.8226993680000305,
+      "learning_rate": 0.00016044776979066386,
+      "loss": 0.4421,
+      "step": 4279
+    },
+    {
+      "epoch": 0.31472902419295534,
+      "grad_norm": 0.4966588318347931,
+      "learning_rate": 0.00016042879557993372,
+      "loss": 0.4735,
+      "step": 4280
+    },
+    {
+      "epoch": 0.314802559011692,
+      "grad_norm": 0.8597519993782043,
+      "learning_rate": 0.00016040981794161854,
+      "loss": 0.4642,
+      "step": 4281
+    },
+    {
+      "epoch": 0.3148760938304287,
+      "grad_norm": 0.6643438339233398,
+      "learning_rate": 0.0001603908368767947,
+      "loss": 0.461,
+      "step": 4282
+    },
+    {
+      "epoch": 0.31494962864916537,
+      "grad_norm": 0.7052252888679504,
+      "learning_rate": 0.00016037185238653887,
+      "loss": 0.4936,
+      "step": 4283
+    },
+    {
+      "epoch": 0.31502316346790205,
+      "grad_norm": 0.7394566535949707,
+      "learning_rate": 0.00016035286447192782,
+      "loss": 0.4598,
+      "step": 4284
+    },
+    {
+      "epoch": 0.3150966982866387,
+      "grad_norm": 0.570057213306427,
+      "learning_rate": 0.00016033387313403865,
+      "loss": 0.4611,
+      "step": 4285
+    },
+    {
+      "epoch": 0.3151702331053754,
+      "grad_norm": 0.6555688977241516,
+      "learning_rate": 0.00016031487837394855,
+      "loss": 0.4575,
+      "step": 4286
+    },
+    {
+      "epoch": 0.3152437679241121,
+      "grad_norm": 0.6366528272628784,
+      "learning_rate": 0.00016029588019273483,
+      "loss": 0.4543,
+      "step": 4287
+    },
+    {
+      "epoch": 0.31531730274284875,
+      "grad_norm": 0.6434847116470337,
+      "learning_rate": 0.0001602768785914752,
+      "loss": 0.4691,
+      "step": 4288
+    },
+    {
+      "epoch": 0.3153908375615854,
+      "grad_norm": 0.582440197467804,
+      "learning_rate": 0.00016025787357124737,
+      "loss": 0.4552,
+      "step": 4289
+    },
+    {
+      "epoch": 0.3154643723803221,
+      "grad_norm": 0.7820794582366943,
+      "learning_rate": 0.00016023886513312938,
+      "loss": 0.4683,
+      "step": 4290
+    },
+    {
+      "epoch": 0.3155379071990588,
+      "grad_norm": 0.6961053609848022,
+      "learning_rate": 0.00016021985327819936,
+      "loss": 0.4623,
+      "step": 4291
+    },
+    {
+      "epoch": 0.31561144201779545,
+      "grad_norm": 0.6278983950614929,
+      "learning_rate": 0.00016020083800753575,
+      "loss": 0.4497,
+      "step": 4292
+    },
+    {
+      "epoch": 0.31568497683653207,
+      "grad_norm": 0.8574775457382202,
+      "learning_rate": 0.00016018181932221705,
+      "loss": 0.4492,
+      "step": 4293
+    },
+    {
+      "epoch": 0.31575851165526875,
+      "grad_norm": 0.7788645625114441,
+      "learning_rate": 0.00016016279722332207,
+      "loss": 0.4364,
+      "step": 4294
+    },
+    {
+      "epoch": 0.3158320464740054,
+      "grad_norm": 0.5716366767883301,
+      "learning_rate": 0.00016014377171192973,
+      "loss": 0.4474,
+      "step": 4295
+    },
+    {
+      "epoch": 0.3159055812927421,
+      "grad_norm": 0.8948233127593994,
+      "learning_rate": 0.00016012474278911913,
+      "loss": 0.5028,
+      "step": 4296
+    },
+    {
+      "epoch": 0.3159791161114788,
+      "grad_norm": 0.9603826403617859,
+      "learning_rate": 0.00016010571045596973,
+      "loss": 0.4745,
+      "step": 4297
+    },
+    {
+      "epoch": 0.31605265093021545,
+      "grad_norm": 0.7175876498222351,
+      "learning_rate": 0.000160086674713561,
+      "loss": 0.4642,
+      "step": 4298
+    },
+    {
+      "epoch": 0.3161261857489521,
+      "grad_norm": 1.5774115324020386,
+      "learning_rate": 0.00016006763556297265,
+      "loss": 0.4592,
+      "step": 4299
+    },
+    {
+      "epoch": 0.3161997205676888,
+      "grad_norm": 0.8975028395652771,
+      "learning_rate": 0.00016004859300528464,
+      "loss": 0.452,
+      "step": 4300
+    },
+    {
+      "epoch": 0.3162732553864255,
+      "grad_norm": 1.6950775384902954,
+      "learning_rate": 0.00016002954704157703,
+      "loss": 0.4432,
+      "step": 4301
+    },
+    {
+      "epoch": 0.31634679020516215,
+      "grad_norm": 1.2402896881103516,
+      "learning_rate": 0.00016001049767293016,
+      "loss": 0.4705,
+      "step": 4302
+    },
+    {
+      "epoch": 0.31642032502389883,
+      "grad_norm": 1.8617264032363892,
+      "learning_rate": 0.00015999144490042456,
+      "loss": 0.4754,
+      "step": 4303
+    },
+    {
+      "epoch": 0.3164938598426355,
+      "grad_norm": 1.751065731048584,
+      "learning_rate": 0.0001599723887251409,
+      "loss": 0.4619,
+      "step": 4304
+    },
+    {
+      "epoch": 0.3165673946613722,
+      "grad_norm": 1.9620331525802612,
+      "learning_rate": 0.00015995332914816,
+      "loss": 0.4685,
+      "step": 4305
+    },
+    {
+      "epoch": 0.31664092948010886,
+      "grad_norm": 2.35394024848938,
+      "learning_rate": 0.00015993426617056305,
+      "loss": 0.4615,
+      "step": 4306
+    },
+    {
+      "epoch": 0.3167144642988455,
+      "grad_norm": 1.5836467742919922,
+      "learning_rate": 0.00015991519979343126,
+      "loss": 0.4387,
+      "step": 4307
+    },
+    {
+      "epoch": 0.31678799911758215,
+      "grad_norm": 1.6535468101501465,
+      "learning_rate": 0.0001598961300178461,
+      "loss": 0.4588,
+      "step": 4308
+    },
+    {
+      "epoch": 0.31686153393631883,
+      "grad_norm": 1.797429084777832,
+      "learning_rate": 0.00015987705684488926,
+      "loss": 0.4569,
+      "step": 4309
+    },
+    {
+      "epoch": 0.3169350687550555,
+      "grad_norm": 1.4844897985458374,
+      "learning_rate": 0.00015985798027564254,
+      "loss": 0.4919,
+      "step": 4310
+    },
+    {
+      "epoch": 0.3170086035737922,
+      "grad_norm": 1.582093358039856,
+      "learning_rate": 0.00015983890031118802,
+      "loss": 0.4687,
+      "step": 4311
+    },
+    {
+      "epoch": 0.31708213839252886,
+      "grad_norm": 1.340837836265564,
+      "learning_rate": 0.0001598198169526079,
+      "loss": 0.4621,
+      "step": 4312
+    },
+    {
+      "epoch": 0.31715567321126553,
+      "grad_norm": 1.830580234527588,
+      "learning_rate": 0.00015980073020098465,
+      "loss": 0.4551,
+      "step": 4313
+    },
+    {
+      "epoch": 0.3172292080300022,
+      "grad_norm": 1.5851370096206665,
+      "learning_rate": 0.00015978164005740086,
+      "loss": 0.459,
+      "step": 4314
+    },
+    {
+      "epoch": 0.3173027428487389,
+      "grad_norm": 1.8970234394073486,
+      "learning_rate": 0.0001597625465229394,
+      "loss": 0.4397,
+      "step": 4315
+    },
+    {
+      "epoch": 0.31737627766747556,
+      "grad_norm": 1.5691145658493042,
+      "learning_rate": 0.00015974344959868319,
+      "loss": 0.4642,
+      "step": 4316
+    },
+    {
+      "epoch": 0.31744981248621224,
+      "grad_norm": 1.5026893615722656,
+      "learning_rate": 0.00015972434928571548,
+      "loss": 0.4686,
+      "step": 4317
+    },
+    {
+      "epoch": 0.3175233473049489,
+      "grad_norm": 1.3414164781570435,
+      "learning_rate": 0.00015970524558511965,
+      "loss": 0.4661,
+      "step": 4318
+    },
+    {
+      "epoch": 0.3175968821236856,
+      "grad_norm": 1.5631722211837769,
+      "learning_rate": 0.00015968613849797925,
+      "loss": 0.4777,
+      "step": 4319
+    },
+    {
+      "epoch": 0.31767041694242226,
+      "grad_norm": 1.3427236080169678,
+      "learning_rate": 0.00015966702802537813,
+      "loss": 0.4677,
+      "step": 4320
+    },
+    {
+      "epoch": 0.3177439517611589,
+      "grad_norm": 1.2699166536331177,
+      "learning_rate": 0.00015964791416840022,
+      "loss": 0.4327,
+      "step": 4321
+    },
+    {
+      "epoch": 0.31781748657989556,
+      "grad_norm": 1.2626985311508179,
+      "learning_rate": 0.00015962879692812964,
+      "loss": 0.4533,
+      "step": 4322
+    },
+    {
+      "epoch": 0.31789102139863223,
+      "grad_norm": 1.645742416381836,
+      "learning_rate": 0.00015960967630565076,
+      "loss": 0.4617,
+      "step": 4323
+    },
+    {
+      "epoch": 0.3179645562173689,
+      "grad_norm": 1.3491004705429077,
+      "learning_rate": 0.00015959055230204818,
+      "loss": 0.4468,
+      "step": 4324
+    },
+    {
+      "epoch": 0.3180380910361056,
+      "grad_norm": 1.3848789930343628,
+      "learning_rate": 0.00015957142491840654,
+      "loss": 0.4438,
+      "step": 4325
+    },
+    {
+      "epoch": 0.31811162585484226,
+      "grad_norm": 1.2370988130569458,
+      "learning_rate": 0.00015955229415581083,
+      "loss": 0.4835,
+      "step": 4326
+    },
+    {
+      "epoch": 0.31818516067357894,
+      "grad_norm": 1.5170114040374756,
+      "learning_rate": 0.00015953316001534614,
+      "loss": 0.46,
+      "step": 4327
+    },
+    {
+      "epoch": 0.3182586954923156,
+      "grad_norm": 1.340915560722351,
+      "learning_rate": 0.00015951402249809776,
+      "loss": 0.4715,
+      "step": 4328
+    },
+    {
+      "epoch": 0.3183322303110523,
+      "grad_norm": 1.8131386041641235,
+      "learning_rate": 0.00015949488160515128,
+      "loss": 0.4564,
+      "step": 4329
+    },
+    {
+      "epoch": 0.31840576512978896,
+      "grad_norm": 1.7087870836257935,
+      "learning_rate": 0.00015947573733759226,
+      "loss": 0.4603,
+      "step": 4330
+    },
+    {
+      "epoch": 0.31847929994852564,
+      "grad_norm": 1.244613528251648,
+      "learning_rate": 0.00015945658969650667,
+      "loss": 0.4991,
+      "step": 4331
+    },
+    {
+      "epoch": 0.3185528347672623,
+      "grad_norm": 1.1209495067596436,
+      "learning_rate": 0.00015943743868298058,
+      "loss": 0.459,
+      "step": 4332
+    },
+    {
+      "epoch": 0.318626369585999,
+      "grad_norm": 1.3378818035125732,
+      "learning_rate": 0.00015941828429810024,
+      "loss": 0.43,
+      "step": 4333
+    },
+    {
+      "epoch": 0.31869990440473567,
+      "grad_norm": 0.943024754524231,
+      "learning_rate": 0.0001593991265429521,
+      "loss": 0.4557,
+      "step": 4334
+    },
+    {
+      "epoch": 0.3187734392234723,
+      "grad_norm": 1.8545790910720825,
+      "learning_rate": 0.0001593799654186228,
+      "loss": 0.445,
+      "step": 4335
+    },
+    {
+      "epoch": 0.31884697404220896,
+      "grad_norm": 1.543213129043579,
+      "learning_rate": 0.00015936080092619922,
+      "loss": 0.4329,
+      "step": 4336
+    },
+    {
+      "epoch": 0.31892050886094564,
+      "grad_norm": 1.364163875579834,
+      "learning_rate": 0.00015934163306676838,
+      "loss": 0.4175,
+      "step": 4337
+    },
+    {
+      "epoch": 0.3189940436796823,
+      "grad_norm": 1.7789310216903687,
+      "learning_rate": 0.00015932246184141748,
+      "loss": 0.4592,
+      "step": 4338
+    },
+    {
+      "epoch": 0.319067578498419,
+      "grad_norm": 0.9617238640785217,
+      "learning_rate": 0.00015930328725123394,
+      "loss": 0.4633,
+      "step": 4339
+    },
+    {
+      "epoch": 0.31914111331715567,
+      "grad_norm": 1.0349581241607666,
+      "learning_rate": 0.00015928410929730533,
+      "loss": 0.4485,
+      "step": 4340
+    },
+    {
+      "epoch": 0.31921464813589234,
+      "grad_norm": 1.4778461456298828,
+      "learning_rate": 0.0001592649279807195,
+      "loss": 0.457,
+      "step": 4341
+    },
+    {
+      "epoch": 0.319288182954629,
+      "grad_norm": 0.8652521371841431,
+      "learning_rate": 0.00015924574330256438,
+      "loss": 0.445,
+      "step": 4342
+    },
+    {
+      "epoch": 0.3193617177733657,
+      "grad_norm": 1.5680525302886963,
+      "learning_rate": 0.00015922655526392823,
+      "loss": 0.4765,
+      "step": 4343
+    },
+    {
+      "epoch": 0.31943525259210237,
+      "grad_norm": 1.0485107898712158,
+      "learning_rate": 0.00015920736386589935,
+      "loss": 0.4908,
+      "step": 4344
+    },
+    {
+      "epoch": 0.31950878741083905,
+      "grad_norm": 2.081782579421997,
+      "learning_rate": 0.00015918816910956627,
+      "loss": 0.4474,
+      "step": 4345
+    },
+    {
+      "epoch": 0.3195823222295757,
+      "grad_norm": 2.14949631690979,
+      "learning_rate": 0.0001591689709960178,
+      "loss": 0.5178,
+      "step": 4346
+    },
+    {
+      "epoch": 0.3196558570483124,
+      "grad_norm": 0.8461922407150269,
+      "learning_rate": 0.00015914976952634282,
+      "loss": 0.4592,
+      "step": 4347
+    },
+    {
+      "epoch": 0.3197293918670491,
+      "grad_norm": 1.5217198133468628,
+      "learning_rate": 0.0001591305647016305,
+      "loss": 0.4382,
+      "step": 4348
+    },
+    {
+      "epoch": 0.3198029266857857,
+      "grad_norm": 0.870021402835846,
+      "learning_rate": 0.0001591113565229702,
+      "loss": 0.4474,
+      "step": 4349
+    },
+    {
+      "epoch": 0.31987646150452237,
+      "grad_norm": 0.913559079170227,
+      "learning_rate": 0.00015909214499145133,
+      "loss": 0.4361,
+      "step": 4350
+    },
+    {
+      "epoch": 0.31994999632325904,
+      "grad_norm": 1.4251657724380493,
+      "learning_rate": 0.00015907293010816362,
+      "loss": 0.4692,
+      "step": 4351
+    },
+    {
+      "epoch": 0.3200235311419957,
+      "grad_norm": 0.7470573782920837,
+      "learning_rate": 0.000159053711874197,
+      "loss": 0.4687,
+      "step": 4352
+    },
+    {
+      "epoch": 0.3200970659607324,
+      "grad_norm": 1.4267934560775757,
+      "learning_rate": 0.0001590344902906415,
+      "loss": 0.4976,
+      "step": 4353
+    },
+    {
+      "epoch": 0.3201706007794691,
+      "grad_norm": 0.9521996974945068,
+      "learning_rate": 0.0001590152653585874,
+      "loss": 0.4999,
+      "step": 4354
+    },
+    {
+      "epoch": 0.32024413559820575,
+      "grad_norm": 2.1687140464782715,
+      "learning_rate": 0.0001589960370791252,
+      "loss": 0.4819,
+      "step": 4355
+    },
+    {
+      "epoch": 0.3203176704169424,
+      "grad_norm": 2.003159999847412,
+      "learning_rate": 0.00015897680545334556,
+      "loss": 0.4483,
+      "step": 4356
+    },
+    {
+      "epoch": 0.3203912052356791,
+      "grad_norm": 1.579583764076233,
+      "learning_rate": 0.0001589575704823392,
+      "loss": 0.4621,
+      "step": 4357
+    },
+    {
+      "epoch": 0.3204647400544158,
+      "grad_norm": 1.573593258857727,
+      "learning_rate": 0.00015893833216719725,
+      "loss": 0.4621,
+      "step": 4358
+    },
+    {
+      "epoch": 0.32053827487315245,
+      "grad_norm": 1.0415915250778198,
+      "learning_rate": 0.00015891909050901086,
+      "loss": 0.4683,
+      "step": 4359
+    },
+    {
+      "epoch": 0.3206118096918891,
+      "grad_norm": 1.1750942468643188,
+      "learning_rate": 0.00015889984550887152,
+      "loss": 0.4763,
+      "step": 4360
+    },
+    {
+      "epoch": 0.3206853445106258,
+      "grad_norm": 1.431119441986084,
+      "learning_rate": 0.00015888059716787076,
+      "loss": 0.465,
+      "step": 4361
+    },
+    {
+      "epoch": 0.3207588793293625,
+      "grad_norm": 0.9782944321632385,
+      "learning_rate": 0.0001588613454871004,
+      "loss": 0.4663,
+      "step": 4362
+    },
+    {
+      "epoch": 0.3208324141480991,
+      "grad_norm": 1.5466028451919556,
+      "learning_rate": 0.0001588420904676524,
+      "loss": 0.4911,
+      "step": 4363
+    },
+    {
+      "epoch": 0.3209059489668358,
+      "grad_norm": 1.0101583003997803,
+      "learning_rate": 0.00015882283211061897,
+      "loss": 0.4607,
+      "step": 4364
+    },
+    {
+      "epoch": 0.32097948378557245,
+      "grad_norm": 1.8614888191223145,
+      "learning_rate": 0.00015880357041709236,
+      "loss": 0.4864,
+      "step": 4365
+    },
+    {
+      "epoch": 0.3210530186043091,
+      "grad_norm": 1.369871973991394,
+      "learning_rate": 0.00015878430538816523,
+      "loss": 0.4758,
+      "step": 4366
+    },
+    {
+      "epoch": 0.3211265534230458,
+      "grad_norm": 2.006481170654297,
+      "learning_rate": 0.00015876503702493024,
+      "loss": 0.4853,
+      "step": 4367
+    },
+    {
+      "epoch": 0.3212000882417825,
+      "grad_norm": 1.7531843185424805,
+      "learning_rate": 0.00015874576532848034,
+      "loss": 0.4533,
+      "step": 4368
+    },
+    {
+      "epoch": 0.32127362306051915,
+      "grad_norm": 1.2652257680892944,
+      "learning_rate": 0.0001587264902999086,
+      "loss": 0.4857,
+      "step": 4369
+    },
+    {
+      "epoch": 0.32134715787925583,
+      "grad_norm": 1.0898898839950562,
+      "learning_rate": 0.00015870721194030839,
+      "loss": 0.4968,
+      "step": 4370
+    },
+    {
+      "epoch": 0.3214206926979925,
+      "grad_norm": 1.368786334991455,
+      "learning_rate": 0.00015868793025077316,
+      "loss": 0.4497,
+      "step": 4371
+    },
+    {
+      "epoch": 0.3214942275167292,
+      "grad_norm": 0.9971004128456116,
+      "learning_rate": 0.00015866864523239658,
+      "loss": 0.4603,
+      "step": 4372
+    },
+    {
+      "epoch": 0.32156776233546586,
+      "grad_norm": 1.5492208003997803,
+      "learning_rate": 0.00015864935688627253,
+      "loss": 0.4632,
+      "step": 4373
+    },
+    {
+      "epoch": 0.32164129715420253,
+      "grad_norm": 1.1447831392288208,
+      "learning_rate": 0.00015863006521349505,
+      "loss": 0.4651,
+      "step": 4374
+    },
+    {
+      "epoch": 0.3217148319729392,
+      "grad_norm": 1.6919400691986084,
+      "learning_rate": 0.00015861077021515842,
+      "loss": 0.5008,
+      "step": 4375
+    },
+    {
+      "epoch": 0.3217883667916759,
+      "grad_norm": 1.5741161108016968,
+      "learning_rate": 0.00015859147189235703,
+      "loss": 0.4461,
+      "step": 4376
+    },
+    {
+      "epoch": 0.3218619016104125,
+      "grad_norm": 1.2639480829238892,
+      "learning_rate": 0.0001585721702461855,
+      "loss": 0.4371,
+      "step": 4377
+    },
+    {
+      "epoch": 0.3219354364291492,
+      "grad_norm": 1.5169405937194824,
+      "learning_rate": 0.00015855286527773868,
+      "loss": 0.4465,
+      "step": 4378
+    },
+    {
+      "epoch": 0.32200897124788586,
+      "grad_norm": 1.2665475606918335,
+      "learning_rate": 0.0001585335569881116,
+      "loss": 0.4578,
+      "step": 4379
+    },
+    {
+      "epoch": 0.32208250606662253,
+      "grad_norm": 1.3176567554473877,
+      "learning_rate": 0.00015851424537839933,
+      "loss": 0.4518,
+      "step": 4380
+    },
+    {
+      "epoch": 0.3221560408853592,
+      "grad_norm": 1.55814790725708,
+      "learning_rate": 0.00015849493044969734,
+      "loss": 0.4304,
+      "step": 4381
+    },
+    {
+      "epoch": 0.3222295757040959,
+      "grad_norm": 1.311453938484192,
+      "learning_rate": 0.00015847561220310114,
+      "loss": 0.4265,
+      "step": 4382
+    },
+    {
+      "epoch": 0.32230311052283256,
+      "grad_norm": 1.5871241092681885,
+      "learning_rate": 0.00015845629063970652,
+      "loss": 0.4487,
+      "step": 4383
+    },
+    {
+      "epoch": 0.32237664534156923,
+      "grad_norm": 1.1542423963546753,
+      "learning_rate": 0.00015843696576060937,
+      "loss": 0.4762,
+      "step": 4384
+    },
+    {
+      "epoch": 0.3224501801603059,
+      "grad_norm": 1.4475300312042236,
+      "learning_rate": 0.00015841763756690592,
+      "loss": 0.4527,
+      "step": 4385
+    },
+    {
+      "epoch": 0.3225237149790426,
+      "grad_norm": 1.0338889360427856,
+      "learning_rate": 0.00015839830605969237,
+      "loss": 0.453,
+      "step": 4386
+    },
+    {
+      "epoch": 0.32259724979777926,
+      "grad_norm": 1.1555240154266357,
+      "learning_rate": 0.00015837897124006527,
+      "loss": 0.4357,
+      "step": 4387
+    },
+    {
+      "epoch": 0.32267078461651594,
+      "grad_norm": 0.8086439371109009,
+      "learning_rate": 0.00015835963310912131,
+      "loss": 0.4434,
+      "step": 4388
+    },
+    {
+      "epoch": 0.3227443194352526,
+      "grad_norm": 1.2250441312789917,
+      "learning_rate": 0.0001583402916679574,
+      "loss": 0.441,
+      "step": 4389
+    },
+    {
+      "epoch": 0.3228178542539893,
+      "grad_norm": 0.8307499289512634,
+      "learning_rate": 0.00015832094691767054,
+      "loss": 0.4643,
+      "step": 4390
+    },
+    {
+      "epoch": 0.3228913890727259,
+      "grad_norm": 1.0720051527023315,
+      "learning_rate": 0.00015830159885935806,
+      "loss": 0.4353,
+      "step": 4391
+    },
+    {
+      "epoch": 0.3229649238914626,
+      "grad_norm": 1.107048511505127,
+      "learning_rate": 0.00015828224749411734,
+      "loss": 0.4631,
+      "step": 4392
+    },
+    {
+      "epoch": 0.32303845871019926,
+      "grad_norm": 0.6703454256057739,
+      "learning_rate": 0.00015826289282304601,
+      "loss": 0.4672,
+      "step": 4393
+    },
+    {
+      "epoch": 0.32311199352893594,
+      "grad_norm": 1.214545726776123,
+      "learning_rate": 0.00015824353484724196,
+      "loss": 0.4708,
+      "step": 4394
+    },
+    {
+      "epoch": 0.3231855283476726,
+      "grad_norm": 0.7221100926399231,
+      "learning_rate": 0.00015822417356780314,
+      "loss": 0.4738,
+      "step": 4395
+    },
+    {
+      "epoch": 0.3232590631664093,
+      "grad_norm": 1.4270522594451904,
+      "learning_rate": 0.00015820480898582771,
+      "loss": 0.4673,
+      "step": 4396
+    },
+    {
+      "epoch": 0.32333259798514596,
+      "grad_norm": 0.7018361687660217,
+      "learning_rate": 0.00015818544110241412,
+      "loss": 0.4582,
+      "step": 4397
+    },
+    {
+      "epoch": 0.32340613280388264,
+      "grad_norm": 0.9313506484031677,
+      "learning_rate": 0.0001581660699186609,
+      "loss": 0.4706,
+      "step": 4398
+    },
+    {
+      "epoch": 0.3234796676226193,
+      "grad_norm": 0.8699434995651245,
+      "learning_rate": 0.00015814669543566683,
+      "loss": 0.4706,
+      "step": 4399
+    },
+    {
+      "epoch": 0.323553202441356,
+      "grad_norm": 0.9012919664382935,
+      "learning_rate": 0.00015812731765453078,
+      "loss": 0.5068,
+      "step": 4400
+    },
+    {
+      "epoch": 0.32362673726009267,
+      "grad_norm": 0.7722523212432861,
+      "learning_rate": 0.00015810793657635195,
+      "loss": 0.4974,
+      "step": 4401
+    },
+    {
+      "epoch": 0.32370027207882934,
+      "grad_norm": 0.5688024163246155,
+      "learning_rate": 0.00015808855220222966,
+      "loss": 0.4526,
+      "step": 4402
+    },
+    {
+      "epoch": 0.323773806897566,
+      "grad_norm": 0.6473749876022339,
+      "learning_rate": 0.00015806916453326336,
+      "loss": 0.4584,
+      "step": 4403
+    },
+    {
+      "epoch": 0.3238473417163027,
+      "grad_norm": 0.5907595157623291,
+      "learning_rate": 0.00015804977357055278,
+      "loss": 0.4855,
+      "step": 4404
+    },
+    {
+      "epoch": 0.3239208765350393,
+      "grad_norm": 0.6777446269989014,
+      "learning_rate": 0.00015803037931519777,
+      "loss": 0.4995,
+      "step": 4405
+    },
+    {
+      "epoch": 0.323994411353776,
+      "grad_norm": 0.5833470821380615,
+      "learning_rate": 0.00015801098176829837,
+      "loss": 0.4741,
+      "step": 4406
+    },
+    {
+      "epoch": 0.32406794617251267,
+      "grad_norm": 0.7454569339752197,
+      "learning_rate": 0.0001579915809309549,
+      "loss": 0.4853,
+      "step": 4407
+    },
+    {
+      "epoch": 0.32414148099124934,
+      "grad_norm": 0.9458895921707153,
+      "learning_rate": 0.0001579721768042678,
+      "loss": 0.4364,
+      "step": 4408
+    },
+    {
+      "epoch": 0.324215015809986,
+      "grad_norm": 0.8469191789627075,
+      "learning_rate": 0.00015795276938933758,
+      "loss": 0.4773,
+      "step": 4409
+    },
+    {
+      "epoch": 0.3242885506287227,
+      "grad_norm": 0.9135574698448181,
+      "learning_rate": 0.00015793335868726516,
+      "loss": 0.457,
+      "step": 4410
+    },
+    {
+      "epoch": 0.32436208544745937,
+      "grad_norm": 0.7982253432273865,
+      "learning_rate": 0.00015791394469915148,
+      "loss": 0.4287,
+      "step": 4411
+    },
+    {
+      "epoch": 0.32443562026619605,
+      "grad_norm": 0.7422060370445251,
+      "learning_rate": 0.00015789452742609773,
+      "loss": 0.4638,
+      "step": 4412
+    },
+    {
+      "epoch": 0.3245091550849327,
+      "grad_norm": 0.9668279886245728,
+      "learning_rate": 0.0001578751068692053,
+      "loss": 0.4944,
+      "step": 4413
+    },
+    {
+      "epoch": 0.3245826899036694,
+      "grad_norm": 0.6301364898681641,
+      "learning_rate": 0.00015785568302957573,
+      "loss": 0.4651,
+      "step": 4414
+    },
+    {
+      "epoch": 0.3246562247224061,
+      "grad_norm": 0.871698260307312,
+      "learning_rate": 0.00015783625590831078,
+      "loss": 0.4495,
+      "step": 4415
+    },
+    {
+      "epoch": 0.32472975954114275,
+      "grad_norm": 0.7834815979003906,
+      "learning_rate": 0.00015781682550651234,
+      "loss": 0.4741,
+      "step": 4416
+    },
+    {
+      "epoch": 0.3248032943598794,
+      "grad_norm": 0.5882390737533569,
+      "learning_rate": 0.00015779739182528253,
+      "loss": 0.4757,
+      "step": 4417
+    },
+    {
+      "epoch": 0.3248768291786161,
+      "grad_norm": 0.6305413246154785,
+      "learning_rate": 0.00015777795486572366,
+      "loss": 0.4421,
+      "step": 4418
+    },
+    {
+      "epoch": 0.3249503639973527,
+      "grad_norm": 0.8217159509658813,
+      "learning_rate": 0.0001577585146289382,
+      "loss": 0.435,
+      "step": 4419
+    },
+    {
+      "epoch": 0.3250238988160894,
+      "grad_norm": 0.623734176158905,
+      "learning_rate": 0.0001577390711160289,
+      "loss": 0.4654,
+      "step": 4420
+    },
+    {
+      "epoch": 0.32509743363482607,
+      "grad_norm": 0.7001421451568604,
+      "learning_rate": 0.0001577196243280985,
+      "loss": 0.4421,
+      "step": 4421
+    },
+    {
+      "epoch": 0.32517096845356275,
+      "grad_norm": 0.9363953471183777,
+      "learning_rate": 0.0001577001742662501,
+      "loss": 0.4592,
+      "step": 4422
+    },
+    {
+      "epoch": 0.3252445032722994,
+      "grad_norm": 0.6759556531906128,
+      "learning_rate": 0.0001576807209315869,
+      "loss": 0.4829,
+      "step": 4423
+    },
+    {
+      "epoch": 0.3253180380910361,
+      "grad_norm": 0.7528856992721558,
+      "learning_rate": 0.00015766126432521235,
+      "loss": 0.4662,
+      "step": 4424
+    },
+    {
+      "epoch": 0.3253915729097728,
+      "grad_norm": 0.8388892412185669,
+      "learning_rate": 0.00015764180444823007,
+      "loss": 0.4773,
+      "step": 4425
+    },
+    {
+      "epoch": 0.32546510772850945,
+      "grad_norm": 0.7472540140151978,
+      "learning_rate": 0.0001576223413017438,
+      "loss": 0.4829,
+      "step": 4426
+    },
+    {
+      "epoch": 0.3255386425472461,
+      "grad_norm": 0.6801843643188477,
+      "learning_rate": 0.0001576028748868575,
+      "loss": 0.4571,
+      "step": 4427
+    },
+    {
+      "epoch": 0.3256121773659828,
+      "grad_norm": 0.5567529201507568,
+      "learning_rate": 0.00015758340520467533,
+      "loss": 0.4572,
+      "step": 4428
+    },
+    {
+      "epoch": 0.3256857121847195,
+      "grad_norm": 0.6567419767379761,
+      "learning_rate": 0.00015756393225630167,
+      "loss": 0.4209,
+      "step": 4429
+    },
+    {
+      "epoch": 0.32575924700345615,
+      "grad_norm": 0.5778314471244812,
+      "learning_rate": 0.00015754445604284102,
+      "loss": 0.4505,
+      "step": 4430
+    },
+    {
+      "epoch": 0.32583278182219283,
+      "grad_norm": 0.7949357628822327,
+      "learning_rate": 0.0001575249765653981,
+      "loss": 0.4732,
+      "step": 4431
+    },
+    {
+      "epoch": 0.3259063166409295,
+      "grad_norm": 1.0321369171142578,
+      "learning_rate": 0.0001575054938250778,
+      "loss": 0.4743,
+      "step": 4432
+    },
+    {
+      "epoch": 0.3259798514596661,
+      "grad_norm": 0.6878182291984558,
+      "learning_rate": 0.0001574860078229852,
+      "loss": 0.4547,
+      "step": 4433
+    },
+    {
+      "epoch": 0.3260533862784028,
+      "grad_norm": 0.8250906467437744,
+      "learning_rate": 0.0001574665185602256,
+      "loss": 0.4555,
+      "step": 4434
+    },
+    {
+      "epoch": 0.3261269210971395,
+      "grad_norm": 1.1946690082550049,
+      "learning_rate": 0.00015744702603790435,
+      "loss": 0.4626,
+      "step": 4435
+    },
+    {
+      "epoch": 0.32620045591587615,
+      "grad_norm": 0.5892102718353271,
+      "learning_rate": 0.00015742753025712718,
+      "loss": 0.434,
+      "step": 4436
+    },
+    {
+      "epoch": 0.32627399073461283,
+      "grad_norm": 1.2277836799621582,
+      "learning_rate": 0.00015740803121899993,
+      "loss": 0.4557,
+      "step": 4437
+    },
+    {
+      "epoch": 0.3263475255533495,
+      "grad_norm": 0.9942957758903503,
+      "learning_rate": 0.00015738852892462855,
+      "loss": 0.432,
+      "step": 4438
+    },
+    {
+      "epoch": 0.3264210603720862,
+      "grad_norm": 0.5280231237411499,
+      "learning_rate": 0.00015736902337511923,
+      "loss": 0.4111,
+      "step": 4439
+    },
+    {
+      "epoch": 0.32649459519082286,
+      "grad_norm": 0.6544248461723328,
+      "learning_rate": 0.00015734951457157834,
+      "loss": 0.4509,
+      "step": 4440
+    },
+    {
+      "epoch": 0.32656813000955953,
+      "grad_norm": 0.7309231758117676,
+      "learning_rate": 0.00015733000251511248,
+      "loss": 0.4547,
+      "step": 4441
+    },
+    {
+      "epoch": 0.3266416648282962,
+      "grad_norm": 0.5644868016242981,
+      "learning_rate": 0.00015731048720682837,
+      "loss": 0.4491,
+      "step": 4442
+    },
+    {
+      "epoch": 0.3267151996470329,
+      "grad_norm": 0.8553183674812317,
+      "learning_rate": 0.00015729096864783296,
+      "loss": 0.4499,
+      "step": 4443
+    },
+    {
+      "epoch": 0.32678873446576956,
+      "grad_norm": 1.2239938974380493,
+      "learning_rate": 0.00015727144683923335,
+      "loss": 0.4182,
+      "step": 4444
+    },
+    {
+      "epoch": 0.32686226928450623,
+      "grad_norm": 0.5751139521598816,
+      "learning_rate": 0.0001572519217821368,
+      "loss": 0.4589,
+      "step": 4445
+    },
+    {
+      "epoch": 0.3269358041032429,
+      "grad_norm": 1.1334850788116455,
+      "learning_rate": 0.00015723239347765087,
+      "loss": 0.4511,
+      "step": 4446
+    },
+    {
+      "epoch": 0.32700933892197953,
+      "grad_norm": 1.1159865856170654,
+      "learning_rate": 0.00015721286192688314,
+      "loss": 0.488,
+      "step": 4447
+    },
+    {
+      "epoch": 0.3270828737407162,
+      "grad_norm": 0.6527647972106934,
+      "learning_rate": 0.0001571933271309415,
+      "loss": 0.4305,
+      "step": 4448
+    },
+    {
+      "epoch": 0.3271564085594529,
+      "grad_norm": 0.8912895917892456,
+      "learning_rate": 0.00015717378909093404,
+      "loss": 0.4631,
+      "step": 4449
+    },
+    {
+      "epoch": 0.32722994337818956,
+      "grad_norm": 1.4844844341278076,
+      "learning_rate": 0.00015715424780796887,
+      "loss": 0.4334,
+      "step": 4450
+    },
+    {
+      "epoch": 0.32730347819692623,
+      "grad_norm": 0.680453360080719,
+      "learning_rate": 0.0001571347032831545,
+      "loss": 0.4327,
+      "step": 4451
+    },
+    {
+      "epoch": 0.3273770130156629,
+      "grad_norm": 2.155369758605957,
+      "learning_rate": 0.0001571151555175994,
+      "loss": 0.4768,
+      "step": 4452
+    },
+    {
+      "epoch": 0.3274505478343996,
+      "grad_norm": 1.1909452676773071,
+      "learning_rate": 0.00015709560451241246,
+      "loss": 0.4633,
+      "step": 4453
+    },
+    {
+      "epoch": 0.32752408265313626,
+      "grad_norm": 2.599792003631592,
+      "learning_rate": 0.00015707605026870253,
+      "loss": 0.4645,
+      "step": 4454
+    },
+    {
+      "epoch": 0.32759761747187294,
+      "grad_norm": 3.0779075622558594,
+      "learning_rate": 0.00015705649278757885,
+      "loss": 0.4654,
+      "step": 4455
+    },
+    {
+      "epoch": 0.3276711522906096,
+      "grad_norm": 0.5951907634735107,
+      "learning_rate": 0.00015703693207015063,
+      "loss": 0.4735,
+      "step": 4456
+    },
+    {
+      "epoch": 0.3277446871093463,
+      "grad_norm": 3.2632033824920654,
+      "learning_rate": 0.00015701736811752745,
+      "loss": 0.4807,
+      "step": 4457
+    },
+    {
+      "epoch": 0.32781822192808296,
+      "grad_norm": 3.1154747009277344,
+      "learning_rate": 0.00015699780093081898,
+      "loss": 0.4694,
+      "step": 4458
+    },
+    {
+      "epoch": 0.32789175674681964,
+      "grad_norm": 0.7922355532646179,
+      "learning_rate": 0.00015697823051113508,
+      "loss": 0.4609,
+      "step": 4459
+    },
+    {
+      "epoch": 0.3279652915655563,
+      "grad_norm": 2.8593997955322266,
+      "learning_rate": 0.0001569586568595858,
+      "loss": 0.4729,
+      "step": 4460
+    },
+    {
+      "epoch": 0.32803882638429294,
+      "grad_norm": 3.0133838653564453,
+      "learning_rate": 0.00015693907997728145,
+      "loss": 0.4656,
+      "step": 4461
+    },
+    {
+      "epoch": 0.3281123612030296,
+      "grad_norm": 1.176746129989624,
+      "learning_rate": 0.00015691949986533236,
+      "loss": 0.4667,
+      "step": 4462
+    },
+    {
+      "epoch": 0.3281858960217663,
+      "grad_norm": 2.0701904296875,
+      "learning_rate": 0.00015689991652484916,
+      "loss": 0.4561,
+      "step": 4463
+    },
+    {
+      "epoch": 0.32825943084050296,
+      "grad_norm": 1.2554813623428345,
+      "learning_rate": 0.00015688032995694266,
+      "loss": 0.4845,
+      "step": 4464
+    },
+    {
+      "epoch": 0.32833296565923964,
+      "grad_norm": 1.8889119625091553,
+      "learning_rate": 0.0001568607401627238,
+      "loss": 0.4537,
+      "step": 4465
+    },
+    {
+      "epoch": 0.3284065004779763,
+      "grad_norm": 1.7119276523590088,
+      "learning_rate": 0.00015684114714330377,
+      "loss": 0.4916,
+      "step": 4466
+    },
+    {
+      "epoch": 0.328480035296713,
+      "grad_norm": 1.6454579830169678,
+      "learning_rate": 0.00015682155089979388,
+      "loss": 0.45,
+      "step": 4467
+    },
+    {
+      "epoch": 0.32855357011544967,
+      "grad_norm": 1.644046664237976,
+      "learning_rate": 0.00015680195143330563,
+      "loss": 0.4593,
+      "step": 4468
+    },
+    {
+      "epoch": 0.32862710493418634,
+      "grad_norm": 1.3461312055587769,
+      "learning_rate": 0.00015678234874495078,
+      "loss": 0.4848,
+      "step": 4469
+    },
+    {
+      "epoch": 0.328700639752923,
+      "grad_norm": 1.0259512662887573,
+      "learning_rate": 0.0001567627428358412,
+      "loss": 0.4352,
+      "step": 4470
+    },
+    {
+      "epoch": 0.3287741745716597,
+      "grad_norm": 1.1119985580444336,
+      "learning_rate": 0.0001567431337070889,
+      "loss": 0.442,
+      "step": 4471
+    },
+    {
+      "epoch": 0.32884770939039637,
+      "grad_norm": 0.8046624064445496,
+      "learning_rate": 0.00015672352135980618,
+      "loss": 0.4356,
+      "step": 4472
+    },
+    {
+      "epoch": 0.32892124420913305,
+      "grad_norm": 1.0149346590042114,
+      "learning_rate": 0.0001567039057951055,
+      "loss": 0.4684,
+      "step": 4473
+    },
+    {
+      "epoch": 0.3289947790278697,
+      "grad_norm": 1.0962061882019043,
+      "learning_rate": 0.0001566842870140994,
+      "loss": 0.4384,
+      "step": 4474
+    },
+    {
+      "epoch": 0.32906831384660634,
+      "grad_norm": 0.8162040114402771,
+      "learning_rate": 0.00015666466501790076,
+      "loss": 0.4533,
+      "step": 4475
+    },
+    {
+      "epoch": 0.329141848665343,
+      "grad_norm": 1.124591588973999,
+      "learning_rate": 0.00015664503980762246,
+      "loss": 0.5329,
+      "step": 4476
+    },
+    {
+      "epoch": 0.3292153834840797,
+      "grad_norm": 0.7125924229621887,
+      "learning_rate": 0.00015662541138437776,
+      "loss": 0.4376,
+      "step": 4477
+    },
+    {
+      "epoch": 0.32928891830281637,
+      "grad_norm": 0.7879989147186279,
+      "learning_rate": 0.00015660577974928,
+      "loss": 0.4384,
+      "step": 4478
+    },
+    {
+      "epoch": 0.32936245312155304,
+      "grad_norm": 0.9322797060012817,
+      "learning_rate": 0.0001565861449034426,
+      "loss": 0.4657,
+      "step": 4479
+    },
+    {
+      "epoch": 0.3294359879402897,
+      "grad_norm": 0.6887032389640808,
+      "learning_rate": 0.0001565665068479794,
+      "loss": 0.4496,
+      "step": 4480
+    },
+    {
+      "epoch": 0.3295095227590264,
+      "grad_norm": 0.7075858116149902,
+      "learning_rate": 0.00015654686558400421,
+      "loss": 0.4454,
+      "step": 4481
+    },
+    {
+      "epoch": 0.32958305757776307,
+      "grad_norm": 0.7245625257492065,
+      "learning_rate": 0.00015652722111263114,
+      "loss": 0.4549,
+      "step": 4482
+    },
+    {
+      "epoch": 0.32965659239649975,
+      "grad_norm": 0.7372053265571594,
+      "learning_rate": 0.00015650757343497443,
+      "loss": 0.4595,
+      "step": 4483
+    },
+    {
+      "epoch": 0.3297301272152364,
+      "grad_norm": 0.8082761168479919,
+      "learning_rate": 0.00015648792255214855,
+      "loss": 0.4525,
+      "step": 4484
+    },
+    {
+      "epoch": 0.3298036620339731,
+      "grad_norm": 0.6847712993621826,
+      "learning_rate": 0.00015646826846526802,
+      "loss": 0.4441,
+      "step": 4485
+    },
+    {
+      "epoch": 0.3298771968527098,
+      "grad_norm": 0.6898791193962097,
+      "learning_rate": 0.00015644861117544777,
+      "loss": 0.4841,
+      "step": 4486
+    },
+    {
+      "epoch": 0.32995073167144645,
+      "grad_norm": 0.600948691368103,
+      "learning_rate": 0.0001564289506838027,
+      "loss": 0.4674,
+      "step": 4487
+    },
+    {
+      "epoch": 0.3300242664901831,
+      "grad_norm": 0.5766618847846985,
+      "learning_rate": 0.00015640928699144798,
+      "loss": 0.508,
+      "step": 4488
+    },
+    {
+      "epoch": 0.33009780130891975,
+      "grad_norm": 0.6894156336784363,
+      "learning_rate": 0.00015638962009949898,
+      "loss": 0.4447,
+      "step": 4489
+    },
+    {
+      "epoch": 0.3301713361276564,
+      "grad_norm": 0.589521586894989,
+      "learning_rate": 0.00015636995000907125,
+      "loss": 0.4762,
+      "step": 4490
+    },
+    {
+      "epoch": 0.3302448709463931,
+      "grad_norm": 0.8688734769821167,
+      "learning_rate": 0.00015635027672128042,
+      "loss": 0.4237,
+      "step": 4491
+    },
+    {
+      "epoch": 0.3303184057651298,
+      "grad_norm": 1.1137841939926147,
+      "learning_rate": 0.00015633060023724244,
+      "loss": 0.4481,
+      "step": 4492
+    },
+    {
+      "epoch": 0.33039194058386645,
+      "grad_norm": 0.7081375122070312,
+      "learning_rate": 0.00015631092055807337,
+      "loss": 0.4212,
+      "step": 4493
+    },
+    {
+      "epoch": 0.3304654754026031,
+      "grad_norm": 0.5863112211227417,
+      "learning_rate": 0.00015629123768488948,
+      "loss": 0.45,
+      "step": 4494
+    },
+    {
+      "epoch": 0.3305390102213398,
+      "grad_norm": 1.0832560062408447,
+      "learning_rate": 0.00015627155161880715,
+      "loss": 0.4276,
+      "step": 4495
+    },
+    {
+      "epoch": 0.3306125450400765,
+      "grad_norm": 0.7302224040031433,
+      "learning_rate": 0.00015625186236094304,
+      "loss": 0.4256,
+      "step": 4496
+    },
+    {
+      "epoch": 0.33068607985881315,
+      "grad_norm": 0.7928889393806458,
+      "learning_rate": 0.0001562321699124139,
+      "loss": 0.4267,
+      "step": 4497
+    },
+    {
+      "epoch": 0.33075961467754983,
+      "grad_norm": 1.427424669265747,
+      "learning_rate": 0.00015621247427433682,
+      "loss": 0.4562,
+      "step": 4498
+    },
+    {
+      "epoch": 0.3308331494962865,
+      "grad_norm": 0.5676271915435791,
+      "learning_rate": 0.00015619277544782876,
+      "loss": 0.4582,
+      "step": 4499
+    },
+    {
+      "epoch": 0.3309066843150232,
+      "grad_norm": 1.4638501405715942,
+      "learning_rate": 0.00015617307343400726,
+      "loss": 0.4811,
+      "step": 4500
+    },
+    {
+      "epoch": 0.33098021913375986,
+      "grad_norm": 0.7544583082199097,
+      "learning_rate": 0.00015615336823398974,
+      "loss": 0.4508,
+      "step": 4501
+    },
+    {
+      "epoch": 0.33105375395249653,
+      "grad_norm": 0.9350889921188354,
+      "learning_rate": 0.00015613365984889393,
+      "loss": 0.4407,
+      "step": 4502
+    },
+    {
+      "epoch": 0.33112728877123315,
+      "grad_norm": 1.739711046218872,
+      "learning_rate": 0.00015611394827983764,
+      "loss": 0.4546,
+      "step": 4503
+    },
+    {
+      "epoch": 0.33120082358996983,
+      "grad_norm": 0.8836866617202759,
+      "learning_rate": 0.00015609423352793902,
+      "loss": 0.4262,
+      "step": 4504
+    },
+    {
+      "epoch": 0.3312743584087065,
+      "grad_norm": 2.234393835067749,
+      "learning_rate": 0.00015607451559431626,
+      "loss": 0.4673,
+      "step": 4505
+    },
+    {
+      "epoch": 0.3313478932274432,
+      "grad_norm": 1.7108135223388672,
+      "learning_rate": 0.0001560547944800878,
+      "loss": 0.4375,
+      "step": 4506
+    },
+    {
+      "epoch": 0.33142142804617986,
+      "grad_norm": 1.2572671175003052,
+      "learning_rate": 0.00015603507018637224,
+      "loss": 0.4385,
+      "step": 4507
+    },
+    {
+      "epoch": 0.33149496286491653,
+      "grad_norm": 1.257860779762268,
+      "learning_rate": 0.0001560153427142884,
+      "loss": 0.4276,
+      "step": 4508
+    },
+    {
+      "epoch": 0.3315684976836532,
+      "grad_norm": 1.605824589729309,
+      "learning_rate": 0.00015599561206495514,
+      "loss": 0.4315,
+      "step": 4509
+    },
+    {
+      "epoch": 0.3316420325023899,
+      "grad_norm": 1.23849618434906,
+      "learning_rate": 0.00015597587823949172,
+      "loss": 0.4711,
+      "step": 4510
+    },
+    {
+      "epoch": 0.33171556732112656,
+      "grad_norm": 1.8237717151641846,
+      "learning_rate": 0.00015595614123901737,
+      "loss": 0.4301,
+      "step": 4511
+    },
+    {
+      "epoch": 0.33178910213986323,
+      "grad_norm": 1.9720951318740845,
+      "learning_rate": 0.00015593640106465167,
+      "loss": 0.4995,
+      "step": 4512
+    },
+    {
+      "epoch": 0.3318626369585999,
+      "grad_norm": 1.308905005455017,
+      "learning_rate": 0.00015591665771751427,
+      "loss": 0.465,
+      "step": 4513
+    },
+    {
+      "epoch": 0.3319361717773366,
+      "grad_norm": 1.2941428422927856,
+      "learning_rate": 0.00015589691119872502,
+      "loss": 0.4358,
+      "step": 4514
+    },
+    {
+      "epoch": 0.33200970659607326,
+      "grad_norm": 1.243644118309021,
+      "learning_rate": 0.00015587716150940398,
+      "loss": 0.4791,
+      "step": 4515
+    },
+    {
+      "epoch": 0.33208324141480994,
+      "grad_norm": 0.8654093146324158,
+      "learning_rate": 0.00015585740865067135,
+      "loss": 0.4693,
+      "step": 4516
+    },
+    {
+      "epoch": 0.33215677623354656,
+      "grad_norm": 1.0821815729141235,
+      "learning_rate": 0.00015583765262364757,
+      "loss": 0.4562,
+      "step": 4517
+    },
+    {
+      "epoch": 0.33223031105228323,
+      "grad_norm": 0.7485228776931763,
+      "learning_rate": 0.00015581789342945324,
+      "loss": 0.4221,
+      "step": 4518
+    },
+    {
+      "epoch": 0.3323038458710199,
+      "grad_norm": 1.616570234298706,
+      "learning_rate": 0.00015579813106920908,
+      "loss": 0.4707,
+      "step": 4519
+    },
+    {
+      "epoch": 0.3323773806897566,
+      "grad_norm": 0.8171940445899963,
+      "learning_rate": 0.00015577836554403598,
+      "loss": 0.4286,
+      "step": 4520
+    },
+    {
+      "epoch": 0.33245091550849326,
+      "grad_norm": 1.6286182403564453,
+      "learning_rate": 0.00015575859685505516,
+      "loss": 0.4169,
+      "step": 4521
+    },
+    {
+      "epoch": 0.33252445032722994,
+      "grad_norm": 1.099626898765564,
+      "learning_rate": 0.0001557388250033879,
+      "loss": 0.4389,
+      "step": 4522
+    },
+    {
+      "epoch": 0.3325979851459666,
+      "grad_norm": 2.0095691680908203,
+      "learning_rate": 0.00015571904999015561,
+      "loss": 0.4458,
+      "step": 4523
+    },
+    {
+      "epoch": 0.3326715199647033,
+      "grad_norm": 2.03708815574646,
+      "learning_rate": 0.00015569927181648002,
+      "loss": 0.4121,
+      "step": 4524
+    },
+    {
+      "epoch": 0.33274505478343996,
+      "grad_norm": 0.7792876958847046,
+      "learning_rate": 0.00015567949048348297,
+      "loss": 0.4519,
+      "step": 4525
+    },
+    {
+      "epoch": 0.33281858960217664,
+      "grad_norm": 1.6078158617019653,
+      "learning_rate": 0.00015565970599228646,
+      "loss": 0.424,
+      "step": 4526
+    },
+    {
+      "epoch": 0.3328921244209133,
+      "grad_norm": 0.8211187720298767,
+      "learning_rate": 0.00015563991834401268,
+      "loss": 0.4141,
+      "step": 4527
+    },
+    {
+      "epoch": 0.33296565923965,
+      "grad_norm": 1.9791572093963623,
+      "learning_rate": 0.00015562012753978402,
+      "loss": 0.4275,
+      "step": 4528
+    },
+    {
+      "epoch": 0.33303919405838667,
+      "grad_norm": 1.4691458940505981,
+      "learning_rate": 0.000155600333580723,
+      "loss": 0.4271,
+      "step": 4529
+    },
+    {
+      "epoch": 0.33311272887712334,
+      "grad_norm": 1.4822373390197754,
+      "learning_rate": 0.00015558053646795242,
+      "loss": 0.4538,
+      "step": 4530
+    },
+    {
+      "epoch": 0.33318626369585996,
+      "grad_norm": 1.504319429397583,
+      "learning_rate": 0.00015556073620259512,
+      "loss": 0.4418,
+      "step": 4531
+    },
+    {
+      "epoch": 0.33325979851459664,
+      "grad_norm": 0.8845163583755493,
+      "learning_rate": 0.00015554093278577423,
+      "loss": 0.4274,
+      "step": 4532
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 1.42852783203125,
+      "learning_rate": 0.00015552112621861305,
+      "loss": 0.4934,
+      "step": 4533
+    },
+    {
+      "epoch": 0.33340686815207,
+      "grad_norm": 1.021019458770752,
+      "learning_rate": 0.00015550131650223496,
+      "loss": 0.4306,
+      "step": 4534
+    },
+    {
+      "epoch": 0.33348040297080667,
+      "grad_norm": 1.8559014797210693,
+      "learning_rate": 0.00015548150363776367,
+      "loss": 0.4188,
+      "step": 4535
+    },
+    {
+      "epoch": 0.33355393778954334,
+      "grad_norm": 1.645843267440796,
+      "learning_rate": 0.0001554616876263229,
+      "loss": 0.4519,
+      "step": 4536
+    },
+    {
+      "epoch": 0.33362747260828,
+      "grad_norm": 1.370821237564087,
+      "learning_rate": 0.00015544186846903668,
+      "loss": 0.4341,
+      "step": 4537
+    },
+    {
+      "epoch": 0.3337010074270167,
+      "grad_norm": 1.345973253250122,
+      "learning_rate": 0.0001554220461670292,
+      "loss": 0.4276,
+      "step": 4538
+    },
+    {
+      "epoch": 0.33377454224575337,
+      "grad_norm": 1.569287896156311,
+      "learning_rate": 0.00015540222072142474,
+      "loss": 0.4701,
+      "step": 4539
+    },
+    {
+      "epoch": 0.33384807706449005,
+      "grad_norm": 1.156673550605774,
+      "learning_rate": 0.00015538239213334787,
+      "loss": 0.4499,
+      "step": 4540
+    },
+    {
+      "epoch": 0.3339216118832267,
+      "grad_norm": 1.440953016281128,
+      "learning_rate": 0.00015536256040392328,
+      "loss": 0.4472,
+      "step": 4541
+    },
+    {
+      "epoch": 0.3339951467019634,
+      "grad_norm": 0.9855929613113403,
+      "learning_rate": 0.0001553427255342758,
+      "loss": 0.48,
+      "step": 4542
+    },
+    {
+      "epoch": 0.3340686815207001,
+      "grad_norm": 1.0602442026138306,
+      "learning_rate": 0.00015532288752553055,
+      "loss": 0.4313,
+      "step": 4543
+    },
+    {
+      "epoch": 0.33414221633943675,
+      "grad_norm": 1.2035086154937744,
+      "learning_rate": 0.00015530304637881274,
+      "loss": 0.4399,
+      "step": 4544
+    },
+    {
+      "epoch": 0.33421575115817337,
+      "grad_norm": 0.6688458919525146,
+      "learning_rate": 0.00015528320209524775,
+      "loss": 0.448,
+      "step": 4545
+    },
+    {
+      "epoch": 0.33428928597691004,
+      "grad_norm": 1.5855494737625122,
+      "learning_rate": 0.0001552633546759612,
+      "loss": 0.4547,
+      "step": 4546
+    },
+    {
+      "epoch": 0.3343628207956467,
+      "grad_norm": 0.7329311966896057,
+      "learning_rate": 0.00015524350412207886,
+      "loss": 0.4405,
+      "step": 4547
+    },
+    {
+      "epoch": 0.3344363556143834,
+      "grad_norm": 1.6837658882141113,
+      "learning_rate": 0.00015522365043472668,
+      "loss": 0.4556,
+      "step": 4548
+    },
+    {
+      "epoch": 0.33450989043312007,
+      "grad_norm": 0.9831632971763611,
+      "learning_rate": 0.00015520379361503079,
+      "loss": 0.4524,
+      "step": 4549
+    },
+    {
+      "epoch": 0.33458342525185675,
+      "grad_norm": 2.3433947563171387,
+      "learning_rate": 0.0001551839336641174,
+      "loss": 0.4229,
+      "step": 4550
+    },
+    {
+      "epoch": 0.3346569600705934,
+      "grad_norm": 2.321359634399414,
+      "learning_rate": 0.0001551640705831131,
+      "loss": 0.4257,
+      "step": 4551
+    },
+    {
+      "epoch": 0.3347304948893301,
+      "grad_norm": 1.2717615365982056,
+      "learning_rate": 0.00015514420437314448,
+      "loss": 0.4196,
+      "step": 4552
+    },
+    {
+      "epoch": 0.3348040297080668,
+      "grad_norm": 1.3869922161102295,
+      "learning_rate": 0.00015512433503533842,
+      "loss": 0.4316,
+      "step": 4553
+    },
+    {
+      "epoch": 0.33487756452680345,
+      "grad_norm": 1.4322497844696045,
+      "learning_rate": 0.00015510446257082188,
+      "loss": 0.451,
+      "step": 4554
+    },
+    {
+      "epoch": 0.3349510993455401,
+      "grad_norm": 0.8288789391517639,
+      "learning_rate": 0.00015508458698072205,
+      "loss": 0.4757,
+      "step": 4555
+    },
+    {
+      "epoch": 0.3350246341642768,
+      "grad_norm": 1.6400619745254517,
+      "learning_rate": 0.00015506470826616634,
+      "loss": 0.4338,
+      "step": 4556
+    },
+    {
+      "epoch": 0.3350981689830135,
+      "grad_norm": 0.7433746457099915,
+      "learning_rate": 0.00015504482642828227,
+      "loss": 0.4806,
+      "step": 4557
+    },
+    {
+      "epoch": 0.33517170380175015,
+      "grad_norm": 1.5862596035003662,
+      "learning_rate": 0.00015502494146819754,
+      "loss": 0.4759,
+      "step": 4558
+    },
+    {
+      "epoch": 0.3352452386204868,
+      "grad_norm": 0.9375681281089783,
+      "learning_rate": 0.00015500505338704007,
+      "loss": 0.4143,
+      "step": 4559
+    },
+    {
+      "epoch": 0.33531877343922345,
+      "grad_norm": 1.6000820398330688,
+      "learning_rate": 0.0001549851621859379,
+      "loss": 0.4223,
+      "step": 4560
+    },
+    {
+      "epoch": 0.3353923082579601,
+      "grad_norm": 1.3570843935012817,
+      "learning_rate": 0.00015496526786601931,
+      "loss": 0.4505,
+      "step": 4561
+    },
+    {
+      "epoch": 0.3354658430766968,
+      "grad_norm": 1.2368438243865967,
+      "learning_rate": 0.00015494537042841272,
+      "loss": 0.4505,
+      "step": 4562
+    },
+    {
+      "epoch": 0.3355393778954335,
+      "grad_norm": 1.4399843215942383,
+      "learning_rate": 0.00015492546987424675,
+      "loss": 0.463,
+      "step": 4563
+    },
+    {
+      "epoch": 0.33561291271417015,
+      "grad_norm": 1.1565272808074951,
+      "learning_rate": 0.00015490556620465014,
+      "loss": 0.4688,
+      "step": 4564
+    },
+    {
+      "epoch": 0.33568644753290683,
+      "grad_norm": 1.6211107969284058,
+      "learning_rate": 0.0001548856594207519,
+      "loss": 0.4414,
+      "step": 4565
+    },
+    {
+      "epoch": 0.3357599823516435,
+      "grad_norm": 1.4235578775405884,
+      "learning_rate": 0.00015486574952368112,
+      "loss": 0.4616,
+      "step": 4566
+    },
+    {
+      "epoch": 0.3358335171703802,
+      "grad_norm": 1.6430044174194336,
+      "learning_rate": 0.00015484583651456711,
+      "loss": 0.4376,
+      "step": 4567
+    },
+    {
+      "epoch": 0.33590705198911686,
+      "grad_norm": 1.1114084720611572,
+      "learning_rate": 0.00015482592039453937,
+      "loss": 0.4562,
+      "step": 4568
+    },
+    {
+      "epoch": 0.33598058680785353,
+      "grad_norm": 1.9819087982177734,
+      "learning_rate": 0.00015480600116472758,
+      "loss": 0.4381,
+      "step": 4569
+    },
+    {
+      "epoch": 0.3360541216265902,
+      "grad_norm": 1.3566405773162842,
+      "learning_rate": 0.00015478607882626156,
+      "loss": 0.4336,
+      "step": 4570
+    },
+    {
+      "epoch": 0.3361276564453269,
+      "grad_norm": 1.7128435373306274,
+      "learning_rate": 0.00015476615338027135,
+      "loss": 0.4465,
+      "step": 4571
+    },
+    {
+      "epoch": 0.33620119126406356,
+      "grad_norm": 1.4359526634216309,
+      "learning_rate": 0.00015474622482788711,
+      "loss": 0.4332,
+      "step": 4572
+    },
+    {
+      "epoch": 0.3362747260828002,
+      "grad_norm": 1.774552822113037,
+      "learning_rate": 0.00015472629317023925,
+      "loss": 0.4473,
+      "step": 4573
+    },
+    {
+      "epoch": 0.33634826090153686,
+      "grad_norm": 1.6387779712677002,
+      "learning_rate": 0.00015470635840845822,
+      "loss": 0.4746,
+      "step": 4574
+    },
+    {
+      "epoch": 0.33642179572027353,
+      "grad_norm": 1.519067645072937,
+      "learning_rate": 0.00015468642054367482,
+      "loss": 0.4563,
+      "step": 4575
+    },
+    {
+      "epoch": 0.3364953305390102,
+      "grad_norm": 1.2952555418014526,
+      "learning_rate": 0.00015466647957702,
+      "loss": 0.4583,
+      "step": 4576
+    },
+    {
+      "epoch": 0.3365688653577469,
+      "grad_norm": 1.5806496143341064,
+      "learning_rate": 0.00015464653550962472,
+      "loss": 0.4452,
+      "step": 4577
+    },
+    {
+      "epoch": 0.33664240017648356,
+      "grad_norm": 1.1570371389389038,
+      "learning_rate": 0.0001546265883426203,
+      "loss": 0.4151,
+      "step": 4578
+    },
+    {
+      "epoch": 0.33671593499522023,
+      "grad_norm": 1.9829641580581665,
+      "learning_rate": 0.00015460663807713809,
+      "loss": 0.4429,
+      "step": 4579
+    },
+    {
+      "epoch": 0.3367894698139569,
+      "grad_norm": 1.864923119544983,
+      "learning_rate": 0.00015458668471430976,
+      "loss": 0.465,
+      "step": 4580
+    },
+    {
+      "epoch": 0.3368630046326936,
+      "grad_norm": 1.101876139640808,
+      "learning_rate": 0.0001545667282552671,
+      "loss": 0.4014,
+      "step": 4581
+    },
+    {
+      "epoch": 0.33693653945143026,
+      "grad_norm": 1.2601494789123535,
+      "learning_rate": 0.00015454676870114197,
+      "loss": 0.4313,
+      "step": 4582
+    },
+    {
+      "epoch": 0.33701007427016694,
+      "grad_norm": 1.6878739595413208,
+      "learning_rate": 0.0001545268060530666,
+      "loss": 0.4518,
+      "step": 4583
+    },
+    {
+      "epoch": 0.3370836090889036,
+      "grad_norm": 1.095000147819519,
+      "learning_rate": 0.00015450684031217325,
+      "loss": 0.4547,
+      "step": 4584
+    },
+    {
+      "epoch": 0.3371571439076403,
+      "grad_norm": 1.9591667652130127,
+      "learning_rate": 0.00015448687147959436,
+      "loss": 0.4336,
+      "step": 4585
+    },
+    {
+      "epoch": 0.33723067872637696,
+      "grad_norm": 1.6219918727874756,
+      "learning_rate": 0.00015446689955646265,
+      "loss": 0.4512,
+      "step": 4586
+    },
+    {
+      "epoch": 0.3373042135451136,
+      "grad_norm": 1.3705319166183472,
+      "learning_rate": 0.0001544469245439109,
+      "loss": 0.452,
+      "step": 4587
+    },
+    {
+      "epoch": 0.33737774836385026,
+      "grad_norm": 1.3446571826934814,
+      "learning_rate": 0.0001544269464430721,
+      "loss": 0.4583,
+      "step": 4588
+    },
+    {
+      "epoch": 0.33745128318258694,
+      "grad_norm": 1.5215964317321777,
+      "learning_rate": 0.0001544069652550795,
+      "loss": 0.4409,
+      "step": 4589
+    },
+    {
+      "epoch": 0.3375248180013236,
+      "grad_norm": 1.1947667598724365,
+      "learning_rate": 0.0001543869809810664,
+      "loss": 0.4119,
+      "step": 4590
+    },
+    {
+      "epoch": 0.3375983528200603,
+      "grad_norm": 2.179154634475708,
+      "learning_rate": 0.00015436699362216636,
+      "loss": 0.4185,
+      "step": 4591
+    },
+    {
+      "epoch": 0.33767188763879696,
+      "grad_norm": 1.8216681480407715,
+      "learning_rate": 0.00015434700317951304,
+      "loss": 0.4346,
+      "step": 4592
+    },
+    {
+      "epoch": 0.33774542245753364,
+      "grad_norm": 1.3173481225967407,
+      "learning_rate": 0.00015432700965424033,
+      "loss": 0.4304,
+      "step": 4593
+    },
+    {
+      "epoch": 0.3378189572762703,
+      "grad_norm": 1.4221373796463013,
+      "learning_rate": 0.00015430701304748235,
+      "loss": 0.4012,
+      "step": 4594
+    },
+    {
+      "epoch": 0.337892492095007,
+      "grad_norm": 1.3530665636062622,
+      "learning_rate": 0.00015428701336037324,
+      "loss": 0.4642,
+      "step": 4595
+    },
+    {
+      "epoch": 0.33796602691374367,
+      "grad_norm": 0.9944523572921753,
+      "learning_rate": 0.00015426701059404745,
+      "loss": 0.4316,
+      "step": 4596
+    },
+    {
+      "epoch": 0.33803956173248034,
+      "grad_norm": 1.600617527961731,
+      "learning_rate": 0.00015424700474963955,
+      "loss": 0.4374,
+      "step": 4597
+    },
+    {
+      "epoch": 0.338113096551217,
+      "grad_norm": 1.2132861614227295,
+      "learning_rate": 0.0001542269958282843,
+      "loss": 0.4327,
+      "step": 4598
+    },
+    {
+      "epoch": 0.3381866313699537,
+      "grad_norm": 1.8754390478134155,
+      "learning_rate": 0.0001542069838311166,
+      "loss": 0.442,
+      "step": 4599
+    },
+    {
+      "epoch": 0.33826016618869037,
+      "grad_norm": 1.4064357280731201,
+      "learning_rate": 0.00015418696875927156,
+      "loss": 0.4427,
+      "step": 4600
+    },
+    {
+      "epoch": 0.338333701007427,
+      "grad_norm": 1.6216259002685547,
+      "learning_rate": 0.0001541669506138845,
+      "loss": 0.4436,
+      "step": 4601
+    },
+    {
+      "epoch": 0.33840723582616367,
+      "grad_norm": 1.7833870649337769,
+      "learning_rate": 0.00015414692939609085,
+      "loss": 0.453,
+      "step": 4602
+    },
+    {
+      "epoch": 0.33848077064490034,
+      "grad_norm": 1.4119951725006104,
+      "learning_rate": 0.00015412690510702617,
+      "loss": 0.4383,
+      "step": 4603
+    },
+    {
+      "epoch": 0.338554305463637,
+      "grad_norm": 1.232721209526062,
+      "learning_rate": 0.00015410687774782632,
+      "loss": 0.4393,
+      "step": 4604
+    },
+    {
+      "epoch": 0.3386278402823737,
+      "grad_norm": 1.6271326541900635,
+      "learning_rate": 0.00015408684731962728,
+      "loss": 0.4672,
+      "step": 4605
+    },
+    {
+      "epoch": 0.33870137510111037,
+      "grad_norm": 1.3842113018035889,
+      "learning_rate": 0.00015406681382356516,
+      "loss": 0.4353,
+      "step": 4606
+    },
+    {
+      "epoch": 0.33877490991984704,
+      "grad_norm": 1.5162009000778198,
+      "learning_rate": 0.00015404677726077635,
+      "loss": 0.419,
+      "step": 4607
+    },
+    {
+      "epoch": 0.3388484447385837,
+      "grad_norm": 1.2850217819213867,
+      "learning_rate": 0.00015402673763239725,
+      "loss": 0.4135,
+      "step": 4608
+    },
+    {
+      "epoch": 0.3389219795573204,
+      "grad_norm": 1.5078673362731934,
+      "learning_rate": 0.0001540066949395646,
+      "loss": 0.4484,
+      "step": 4609
+    },
+    {
+      "epoch": 0.33899551437605707,
+      "grad_norm": 1.4104818105697632,
+      "learning_rate": 0.00015398664918341522,
+      "loss": 0.4209,
+      "step": 4610
+    },
+    {
+      "epoch": 0.33906904919479375,
+      "grad_norm": 1.7364747524261475,
+      "learning_rate": 0.00015396660036508608,
+      "loss": 0.4376,
+      "step": 4611
+    },
+    {
+      "epoch": 0.3391425840135304,
+      "grad_norm": 1.5962331295013428,
+      "learning_rate": 0.00015394654848571444,
+      "loss": 0.414,
+      "step": 4612
+    },
+    {
+      "epoch": 0.3392161188322671,
+      "grad_norm": 1.2211976051330566,
+      "learning_rate": 0.00015392649354643764,
+      "loss": 0.4634,
+      "step": 4613
+    },
+    {
+      "epoch": 0.3392896536510038,
+      "grad_norm": 0.9749736189842224,
+      "learning_rate": 0.00015390643554839324,
+      "loss": 0.3954,
+      "step": 4614
+    },
+    {
+      "epoch": 0.3393631884697404,
+      "grad_norm": 1.6756964921951294,
+      "learning_rate": 0.0001538863744927189,
+      "loss": 0.4404,
+      "step": 4615
+    },
+    {
+      "epoch": 0.33943672328847707,
+      "grad_norm": 1.248443365097046,
+      "learning_rate": 0.00015386631038055254,
+      "loss": 0.4137,
+      "step": 4616
+    },
+    {
+      "epoch": 0.33951025810721375,
+      "grad_norm": 2.0569608211517334,
+      "learning_rate": 0.00015384624321303217,
+      "loss": 0.4317,
+      "step": 4617
+    },
+    {
+      "epoch": 0.3395837929259504,
+      "grad_norm": 1.7874038219451904,
+      "learning_rate": 0.0001538261729912961,
+      "loss": 0.4139,
+      "step": 4618
+    },
+    {
+      "epoch": 0.3396573277446871,
+      "grad_norm": 1.1479073762893677,
+      "learning_rate": 0.00015380609971648267,
+      "loss": 0.4359,
+      "step": 4619
+    },
+    {
+      "epoch": 0.3397308625634238,
+      "grad_norm": 1.4682883024215698,
+      "learning_rate": 0.0001537860233897305,
+      "loss": 0.428,
+      "step": 4620
+    },
+    {
+      "epoch": 0.33980439738216045,
+      "grad_norm": 1.8770085573196411,
+      "learning_rate": 0.0001537659440121783,
+      "loss": 0.4593,
+      "step": 4621
+    },
+    {
+      "epoch": 0.3398779322008971,
+      "grad_norm": 1.436425805091858,
+      "learning_rate": 0.00015374586158496505,
+      "loss": 0.4123,
+      "step": 4622
+    },
+    {
+      "epoch": 0.3399514670196338,
+      "grad_norm": 1.5794891119003296,
+      "learning_rate": 0.0001537257761092298,
+      "loss": 0.4086,
+      "step": 4623
+    },
+    {
+      "epoch": 0.3400250018383705,
+      "grad_norm": 1.4990731477737427,
+      "learning_rate": 0.00015370568758611182,
+      "loss": 0.4392,
+      "step": 4624
+    },
+    {
+      "epoch": 0.34009853665710715,
+      "grad_norm": 1.3896102905273438,
+      "learning_rate": 0.00015368559601675055,
+      "loss": 0.4356,
+      "step": 4625
+    },
+    {
+      "epoch": 0.34017207147584383,
+      "grad_norm": 1.2414023876190186,
+      "learning_rate": 0.0001536655014022856,
+      "loss": 0.4492,
+      "step": 4626
+    },
+    {
+      "epoch": 0.3402456062945805,
+      "grad_norm": 1.8904656171798706,
+      "learning_rate": 0.0001536454037438568,
+      "loss": 0.4293,
+      "step": 4627
+    },
+    {
+      "epoch": 0.3403191411133172,
+      "grad_norm": 1.7423477172851562,
+      "learning_rate": 0.00015362530304260412,
+      "loss": 0.4305,
+      "step": 4628
+    },
+    {
+      "epoch": 0.3403926759320538,
+      "grad_norm": 1.5771070718765259,
+      "learning_rate": 0.00015360519929966758,
+      "loss": 0.4329,
+      "step": 4629
+    },
+    {
+      "epoch": 0.3404662107507905,
+      "grad_norm": 1.7165584564208984,
+      "learning_rate": 0.00015358509251618764,
+      "loss": 0.4345,
+      "step": 4630
+    },
+    {
+      "epoch": 0.34053974556952715,
+      "grad_norm": 1.40392005443573,
+      "learning_rate": 0.00015356498269330465,
+      "loss": 0.4291,
+      "step": 4631
+    },
+    {
+      "epoch": 0.34061328038826383,
+      "grad_norm": 1.3196033239364624,
+      "learning_rate": 0.00015354486983215938,
+      "loss": 0.4467,
+      "step": 4632
+    },
+    {
+      "epoch": 0.3406868152070005,
+      "grad_norm": 1.2274671792984009,
+      "learning_rate": 0.00015352475393389252,
+      "loss": 0.4392,
+      "step": 4633
+    },
+    {
+      "epoch": 0.3407603500257372,
+      "grad_norm": 1.1059997081756592,
+      "learning_rate": 0.00015350463499964512,
+      "loss": 0.4384,
+      "step": 4634
+    },
+    {
+      "epoch": 0.34083388484447386,
+      "grad_norm": 1.6456230878829956,
+      "learning_rate": 0.0001534845130305584,
+      "loss": 0.4368,
+      "step": 4635
+    },
+    {
+      "epoch": 0.34090741966321053,
+      "grad_norm": 1.1482006311416626,
+      "learning_rate": 0.00015346438802777364,
+      "loss": 0.4136,
+      "step": 4636
+    },
+    {
+      "epoch": 0.3409809544819472,
+      "grad_norm": 2.0417561531066895,
+      "learning_rate": 0.0001534442599924324,
+      "loss": 0.4341,
+      "step": 4637
+    },
+    {
+      "epoch": 0.3410544893006839,
+      "grad_norm": 1.8286982774734497,
+      "learning_rate": 0.00015342412892567627,
+      "loss": 0.416,
+      "step": 4638
+    },
+    {
+      "epoch": 0.34112802411942056,
+      "grad_norm": 1.065535306930542,
+      "learning_rate": 0.0001534039948286472,
+      "loss": 0.472,
+      "step": 4639
+    },
+    {
+      "epoch": 0.34120155893815723,
+      "grad_norm": 1.0083614587783813,
+      "learning_rate": 0.00015338385770248718,
+      "loss": 0.4034,
+      "step": 4640
+    },
+    {
+      "epoch": 0.3412750937568939,
+      "grad_norm": 1.6362621784210205,
+      "learning_rate": 0.00015336371754833843,
+      "loss": 0.4271,
+      "step": 4641
+    },
+    {
+      "epoch": 0.3413486285756306,
+      "grad_norm": 1.1891475915908813,
+      "learning_rate": 0.0001533435743673433,
+      "loss": 0.453,
+      "step": 4642
+    },
+    {
+      "epoch": 0.3414221633943672,
+      "grad_norm": 1.8062069416046143,
+      "learning_rate": 0.0001533234281606443,
+      "loss": 0.4164,
+      "step": 4643
+    },
+    {
+      "epoch": 0.3414956982131039,
+      "grad_norm": 2.0340960025787354,
+      "learning_rate": 0.0001533032789293842,
+      "loss": 0.4304,
+      "step": 4644
+    },
+    {
+      "epoch": 0.34156923303184056,
+      "grad_norm": 1.0251134634017944,
+      "learning_rate": 0.00015328312667470587,
+      "loss": 0.4315,
+      "step": 4645
+    },
+    {
+      "epoch": 0.34164276785057723,
+      "grad_norm": 1.1524893045425415,
+      "learning_rate": 0.0001532629713977524,
+      "loss": 0.4163,
+      "step": 4646
+    },
+    {
+      "epoch": 0.3417163026693139,
+      "grad_norm": 1.618685245513916,
+      "learning_rate": 0.0001532428130996669,
+      "loss": 0.4493,
+      "step": 4647
+    },
+    {
+      "epoch": 0.3417898374880506,
+      "grad_norm": 1.2429977655410767,
+      "learning_rate": 0.00015322265178159295,
+      "loss": 0.4244,
+      "step": 4648
+    },
+    {
+      "epoch": 0.34186337230678726,
+      "grad_norm": 1.8888405561447144,
+      "learning_rate": 0.00015320248744467394,
+      "loss": 0.4178,
+      "step": 4649
+    },
+    {
+      "epoch": 0.34193690712552394,
+      "grad_norm": 1.9734463691711426,
+      "learning_rate": 0.00015318232009005374,
+      "loss": 0.4211,
+      "step": 4650
+    },
+    {
+      "epoch": 0.3420104419442606,
+      "grad_norm": 1.3316898345947266,
+      "learning_rate": 0.0001531621497188762,
+      "loss": 0.4213,
+      "step": 4651
+    },
+    {
+      "epoch": 0.3420839767629973,
+      "grad_norm": 1.424626350402832,
+      "learning_rate": 0.0001531419763322854,
+      "loss": 0.413,
+      "step": 4652
+    },
+    {
+      "epoch": 0.34215751158173396,
+      "grad_norm": 1.2160736322402954,
+      "learning_rate": 0.00015312179993142566,
+      "loss": 0.3912,
+      "step": 4653
+    },
+    {
+      "epoch": 0.34223104640047064,
+      "grad_norm": 1.141387701034546,
+      "learning_rate": 0.00015310162051744138,
+      "loss": 0.4288,
+      "step": 4654
+    },
+    {
+      "epoch": 0.3423045812192073,
+      "grad_norm": 1.7109410762786865,
+      "learning_rate": 0.00015308143809147708,
+      "loss": 0.4619,
+      "step": 4655
+    },
+    {
+      "epoch": 0.342378116037944,
+      "grad_norm": 1.3340058326721191,
+      "learning_rate": 0.00015306125265467766,
+      "loss": 0.4044,
+      "step": 4656
+    },
+    {
+      "epoch": 0.3424516508566806,
+      "grad_norm": 1.947885513305664,
+      "learning_rate": 0.00015304106420818792,
+      "loss": 0.4555,
+      "step": 4657
+    },
+    {
+      "epoch": 0.3425251856754173,
+      "grad_norm": 1.920723795890808,
+      "learning_rate": 0.0001530208727531531,
+      "loss": 0.4089,
+      "step": 4658
+    },
+    {
+      "epoch": 0.34259872049415396,
+      "grad_norm": 1.1165904998779297,
+      "learning_rate": 0.0001530006782907184,
+      "loss": 0.4339,
+      "step": 4659
+    },
+    {
+      "epoch": 0.34267225531289064,
+      "grad_norm": 1.0283751487731934,
+      "learning_rate": 0.00015298048082202932,
+      "loss": 0.3979,
+      "step": 4660
+    },
+    {
+      "epoch": 0.3427457901316273,
+      "grad_norm": 1.4789230823516846,
+      "learning_rate": 0.00015296028034823142,
+      "loss": 0.4244,
+      "step": 4661
+    },
+    {
+      "epoch": 0.342819324950364,
+      "grad_norm": 1.0509967803955078,
+      "learning_rate": 0.00015294007687047053,
+      "loss": 0.4454,
+      "step": 4662
+    },
+    {
+      "epoch": 0.34289285976910067,
+      "grad_norm": 2.118563652038574,
+      "learning_rate": 0.00015291987038989265,
+      "loss": 0.4289,
+      "step": 4663
+    },
+    {
+      "epoch": 0.34296639458783734,
+      "grad_norm": 1.489534616470337,
+      "learning_rate": 0.00015289966090764384,
+      "loss": 0.4282,
+      "step": 4664
+    },
+    {
+      "epoch": 0.343039929406574,
+      "grad_norm": 1.2113404273986816,
+      "learning_rate": 0.00015287944842487042,
+      "loss": 0.4343,
+      "step": 4665
+    },
+    {
+      "epoch": 0.3431134642253107,
+      "grad_norm": 1.1259331703186035,
+      "learning_rate": 0.00015285923294271894,
+      "loss": 0.4085,
+      "step": 4666
+    },
+    {
+      "epoch": 0.34318699904404737,
+      "grad_norm": 1.3490327596664429,
+      "learning_rate": 0.00015283901446233592,
+      "loss": 0.4198,
+      "step": 4667
+    },
+    {
+      "epoch": 0.34326053386278405,
+      "grad_norm": 0.8853120803833008,
+      "learning_rate": 0.00015281879298486824,
+      "loss": 0.431,
+      "step": 4668
+    },
+    {
+      "epoch": 0.3433340686815207,
+      "grad_norm": 1.333301067352295,
+      "learning_rate": 0.00015279856851146292,
+      "loss": 0.414,
+      "step": 4669
+    },
+    {
+      "epoch": 0.3434076035002574,
+      "grad_norm": 0.9924440979957581,
+      "learning_rate": 0.00015277834104326704,
+      "loss": 0.4268,
+      "step": 4670
+    },
+    {
+      "epoch": 0.343481138318994,
+      "grad_norm": 1.9207812547683716,
+      "learning_rate": 0.00015275811058142798,
+      "loss": 0.4202,
+      "step": 4671
+    },
+    {
+      "epoch": 0.3435546731377307,
+      "grad_norm": 1.8198167085647583,
+      "learning_rate": 0.0001527378771270932,
+      "loss": 0.4539,
+      "step": 4672
+    },
+    {
+      "epoch": 0.34362820795646737,
+      "grad_norm": 1.2916313409805298,
+      "learning_rate": 0.00015271764068141037,
+      "loss": 0.4619,
+      "step": 4673
+    },
+    {
+      "epoch": 0.34370174277520404,
+      "grad_norm": 1.2862550020217896,
+      "learning_rate": 0.00015269740124552734,
+      "loss": 0.4382,
+      "step": 4674
+    },
+    {
+      "epoch": 0.3437752775939407,
+      "grad_norm": 1.4067109823226929,
+      "learning_rate": 0.00015267715882059206,
+      "loss": 0.4227,
+      "step": 4675
+    },
+    {
+      "epoch": 0.3438488124126774,
+      "grad_norm": 1.0319796800613403,
+      "learning_rate": 0.00015265691340775276,
+      "loss": 0.3934,
+      "step": 4676
+    },
+    {
+      "epoch": 0.34392234723141407,
+      "grad_norm": 1.647052526473999,
+      "learning_rate": 0.00015263666500815777,
+      "loss": 0.4338,
+      "step": 4677
+    },
+    {
+      "epoch": 0.34399588205015075,
+      "grad_norm": 1.3643487691879272,
+      "learning_rate": 0.00015261641362295563,
+      "loss": 0.4316,
+      "step": 4678
+    },
+    {
+      "epoch": 0.3440694168688874,
+      "grad_norm": 1.7323286533355713,
+      "learning_rate": 0.00015259615925329492,
+      "loss": 0.4343,
+      "step": 4679
+    },
+    {
+      "epoch": 0.3441429516876241,
+      "grad_norm": 1.678301453590393,
+      "learning_rate": 0.00015257590190032456,
+      "loss": 0.4409,
+      "step": 4680
+    },
+    {
+      "epoch": 0.3442164865063608,
+      "grad_norm": 1.4466962814331055,
+      "learning_rate": 0.00015255564156519354,
+      "loss": 0.4352,
+      "step": 4681
+    },
+    {
+      "epoch": 0.34429002132509745,
+      "grad_norm": 1.391812801361084,
+      "learning_rate": 0.00015253537824905112,
+      "loss": 0.4293,
+      "step": 4682
+    },
+    {
+      "epoch": 0.3443635561438341,
+      "grad_norm": 1.4270145893096924,
+      "learning_rate": 0.00015251511195304656,
+      "loss": 0.4047,
+      "step": 4683
+    },
+    {
+      "epoch": 0.3444370909625708,
+      "grad_norm": 0.9951561093330383,
+      "learning_rate": 0.00015249484267832947,
+      "loss": 0.3978,
+      "step": 4684
+    },
+    {
+      "epoch": 0.3445106257813074,
+      "grad_norm": 1.432283878326416,
+      "learning_rate": 0.00015247457042604946,
+      "loss": 0.406,
+      "step": 4685
+    },
+    {
+      "epoch": 0.3445841606000441,
+      "grad_norm": 1.3827292919158936,
+      "learning_rate": 0.00015245429519735647,
+      "loss": 0.4169,
+      "step": 4686
+    },
+    {
+      "epoch": 0.3446576954187808,
+      "grad_norm": 1.5543605089187622,
+      "learning_rate": 0.00015243401699340049,
+      "loss": 0.4242,
+      "step": 4687
+    },
+    {
+      "epoch": 0.34473123023751745,
+      "grad_norm": 1.3272640705108643,
+      "learning_rate": 0.00015241373581533169,
+      "loss": 0.4529,
+      "step": 4688
+    },
+    {
+      "epoch": 0.3448047650562541,
+      "grad_norm": 1.351506233215332,
+      "learning_rate": 0.00015239345166430054,
+      "loss": 0.3905,
+      "step": 4689
+    },
+    {
+      "epoch": 0.3448782998749908,
+      "grad_norm": 1.1378276348114014,
+      "learning_rate": 0.0001523731645414575,
+      "loss": 0.4445,
+      "step": 4690
+    },
+    {
+      "epoch": 0.3449518346937275,
+      "grad_norm": 1.5244377851486206,
+      "learning_rate": 0.00015235287444795328,
+      "loss": 0.4322,
+      "step": 4691
+    },
+    {
+      "epoch": 0.34502536951246415,
+      "grad_norm": 1.4648634195327759,
+      "learning_rate": 0.00015233258138493879,
+      "loss": 0.4266,
+      "step": 4692
+    },
+    {
+      "epoch": 0.34509890433120083,
+      "grad_norm": 1.3864156007766724,
+      "learning_rate": 0.000152312285353565,
+      "loss": 0.4332,
+      "step": 4693
+    },
+    {
+      "epoch": 0.3451724391499375,
+      "grad_norm": 1.203322410583496,
+      "learning_rate": 0.00015229198635498324,
+      "loss": 0.4054,
+      "step": 4694
+    },
+    {
+      "epoch": 0.3452459739686742,
+      "grad_norm": 1.3305633068084717,
+      "learning_rate": 0.00015227168439034482,
+      "loss": 0.3922,
+      "step": 4695
+    },
+    {
+      "epoch": 0.34531950878741086,
+      "grad_norm": 1.0819169282913208,
+      "learning_rate": 0.00015225137946080127,
+      "loss": 0.4113,
+      "step": 4696
+    },
+    {
+      "epoch": 0.34539304360614753,
+      "grad_norm": 1.5229835510253906,
+      "learning_rate": 0.00015223107156750438,
+      "loss": 0.4198,
+      "step": 4697
+    },
+    {
+      "epoch": 0.3454665784248842,
+      "grad_norm": 1.3031301498413086,
+      "learning_rate": 0.00015221076071160594,
+      "loss": 0.4364,
+      "step": 4698
+    },
+    {
+      "epoch": 0.34554011324362083,
+      "grad_norm": 1.5201082229614258,
+      "learning_rate": 0.00015219044689425808,
+      "loss": 0.416,
+      "step": 4699
+    },
+    {
+      "epoch": 0.3456136480623575,
+      "grad_norm": 1.4129213094711304,
+      "learning_rate": 0.000152170130116613,
+      "loss": 0.4277,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3456871828810942,
+      "grad_norm": 1.3756800889968872,
+      "learning_rate": 0.00015214981037982306,
+      "loss": 0.4099,
+      "step": 4701
+    },
+    {
+      "epoch": 0.34576071769983086,
+      "grad_norm": 1.3033421039581299,
+      "learning_rate": 0.00015212948768504085,
+      "loss": 0.4445,
+      "step": 4702
+    },
+    {
+      "epoch": 0.34583425251856753,
+      "grad_norm": 1.501060962677002,
+      "learning_rate": 0.00015210916203341906,
+      "loss": 0.4562,
+      "step": 4703
+    },
+    {
+      "epoch": 0.3459077873373042,
+      "grad_norm": 1.3668402433395386,
+      "learning_rate": 0.00015208883342611063,
+      "loss": 0.4451,
+      "step": 4704
+    },
+    {
+      "epoch": 0.3459813221560409,
+      "grad_norm": 1.5078171491622925,
+      "learning_rate": 0.0001520685018642686,
+      "loss": 0.4021,
+      "step": 4705
+    },
+    {
+      "epoch": 0.34605485697477756,
+      "grad_norm": 1.5240628719329834,
+      "learning_rate": 0.0001520481673490462,
+      "loss": 0.4256,
+      "step": 4706
+    },
+    {
+      "epoch": 0.34612839179351423,
+      "grad_norm": 1.3355897665023804,
+      "learning_rate": 0.00015202782988159682,
+      "loss": 0.4194,
+      "step": 4707
+    },
+    {
+      "epoch": 0.3462019266122509,
+      "grad_norm": 1.344521403312683,
+      "learning_rate": 0.000152007489463074,
+      "loss": 0.4081,
+      "step": 4708
+    },
+    {
+      "epoch": 0.3462754614309876,
+      "grad_norm": 1.4616703987121582,
+      "learning_rate": 0.0001519871460946315,
+      "loss": 0.4099,
+      "step": 4709
+    },
+    {
+      "epoch": 0.34634899624972426,
+      "grad_norm": 1.0187079906463623,
+      "learning_rate": 0.0001519667997774232,
+      "loss": 0.4948,
+      "step": 4710
+    },
+    {
+      "epoch": 0.34642253106846094,
+      "grad_norm": 1.647832989692688,
+      "learning_rate": 0.0001519464505126032,
+      "loss": 0.4115,
+      "step": 4711
+    },
+    {
+      "epoch": 0.3464960658871976,
+      "grad_norm": 1.2835144996643066,
+      "learning_rate": 0.0001519260983013257,
+      "loss": 0.4154,
+      "step": 4712
+    },
+    {
+      "epoch": 0.34656960070593423,
+      "grad_norm": 2.152420997619629,
+      "learning_rate": 0.00015190574314474507,
+      "loss": 0.4258,
+      "step": 4713
+    },
+    {
+      "epoch": 0.3466431355246709,
+      "grad_norm": 1.8044108152389526,
+      "learning_rate": 0.00015188538504401592,
+      "loss": 0.4152,
+      "step": 4714
+    },
+    {
+      "epoch": 0.3467166703434076,
+      "grad_norm": 1.20719313621521,
+      "learning_rate": 0.00015186502400029302,
+      "loss": 0.4399,
+      "step": 4715
+    },
+    {
+      "epoch": 0.34679020516214426,
+      "grad_norm": 1.8496259450912476,
+      "learning_rate": 0.00015184466001473117,
+      "loss": 0.4519,
+      "step": 4716
+    },
+    {
+      "epoch": 0.34686373998088094,
+      "grad_norm": 1.1550019979476929,
+      "learning_rate": 0.0001518242930884855,
+      "loss": 0.4031,
+      "step": 4717
+    },
+    {
+      "epoch": 0.3469372747996176,
+      "grad_norm": 1.3461452722549438,
+      "learning_rate": 0.00015180392322271128,
+      "loss": 0.46,
+      "step": 4718
+    },
+    {
+      "epoch": 0.3470108096183543,
+      "grad_norm": 1.2742788791656494,
+      "learning_rate": 0.00015178355041856382,
+      "loss": 0.448,
+      "step": 4719
+    },
+    {
+      "epoch": 0.34708434443709096,
+      "grad_norm": 0.875083863735199,
+      "learning_rate": 0.00015176317467719875,
+      "loss": 0.4092,
+      "step": 4720
+    },
+    {
+      "epoch": 0.34715787925582764,
+      "grad_norm": 1.1893845796585083,
+      "learning_rate": 0.00015174279599977175,
+      "loss": 0.4209,
+      "step": 4721
+    },
+    {
+      "epoch": 0.3472314140745643,
+      "grad_norm": 0.9485806822776794,
+      "learning_rate": 0.0001517224143874388,
+      "loss": 0.4031,
+      "step": 4722
+    },
+    {
+      "epoch": 0.347304948893301,
+      "grad_norm": 1.1670351028442383,
+      "learning_rate": 0.00015170202984135587,
+      "loss": 0.4136,
+      "step": 4723
+    },
+    {
+      "epoch": 0.34737848371203767,
+      "grad_norm": 0.8694897890090942,
+      "learning_rate": 0.00015168164236267932,
+      "loss": 0.4079,
+      "step": 4724
+    },
+    {
+      "epoch": 0.34745201853077434,
+      "grad_norm": 1.0546956062316895,
+      "learning_rate": 0.00015166125195256543,
+      "loss": 0.4062,
+      "step": 4725
+    },
+    {
+      "epoch": 0.347525553349511,
+      "grad_norm": 1.3257261514663696,
+      "learning_rate": 0.0001516408586121708,
+      "loss": 0.4172,
+      "step": 4726
+    },
+    {
+      "epoch": 0.34759908816824764,
+      "grad_norm": 0.7376654148101807,
+      "learning_rate": 0.00015162046234265216,
+      "loss": 0.4419,
+      "step": 4727
+    },
+    {
+      "epoch": 0.3476726229869843,
+      "grad_norm": 1.4170465469360352,
+      "learning_rate": 0.00015160006314516645,
+      "loss": 0.4138,
+      "step": 4728
+    },
+    {
+      "epoch": 0.347746157805721,
+      "grad_norm": 0.820874810218811,
+      "learning_rate": 0.00015157966102087066,
+      "loss": 0.4454,
+      "step": 4729
+    },
+    {
+      "epoch": 0.34781969262445767,
+      "grad_norm": 1.8479360342025757,
+      "learning_rate": 0.00015155925597092214,
+      "loss": 0.4129,
+      "step": 4730
+    },
+    {
+      "epoch": 0.34789322744319434,
+      "grad_norm": 1.1042611598968506,
+      "learning_rate": 0.00015153884799647816,
+      "loss": 0.4085,
+      "step": 4731
+    },
+    {
+      "epoch": 0.347966762261931,
+      "grad_norm": 2.077014207839966,
+      "learning_rate": 0.0001515184370986963,
+      "loss": 0.4092,
+      "step": 4732
+    },
+    {
+      "epoch": 0.3480402970806677,
+      "grad_norm": 1.832639217376709,
+      "learning_rate": 0.00015149802327873437,
+      "loss": 0.4518,
+      "step": 4733
+    },
+    {
+      "epoch": 0.34811383189940437,
+      "grad_norm": 1.348051905632019,
+      "learning_rate": 0.00015147760653775016,
+      "loss": 0.4169,
+      "step": 4734
+    },
+    {
+      "epoch": 0.34818736671814104,
+      "grad_norm": 1.2890231609344482,
+      "learning_rate": 0.00015145718687690184,
+      "loss": 0.4283,
+      "step": 4735
+    },
+    {
+      "epoch": 0.3482609015368777,
+      "grad_norm": 1.3662077188491821,
+      "learning_rate": 0.00015143676429734754,
+      "loss": 0.4539,
+      "step": 4736
+    },
+    {
+      "epoch": 0.3483344363556144,
+      "grad_norm": 0.9312816262245178,
+      "learning_rate": 0.00015141633880024567,
+      "loss": 0.4265,
+      "step": 4737
+    },
+    {
+      "epoch": 0.34840797117435107,
+      "grad_norm": 1.414028525352478,
+      "learning_rate": 0.00015139591038675486,
+      "loss": 0.4067,
+      "step": 4738
+    },
+    {
+      "epoch": 0.34848150599308775,
+      "grad_norm": 0.8339105844497681,
+      "learning_rate": 0.00015137547905803373,
+      "loss": 0.3926,
+      "step": 4739
+    },
+    {
+      "epoch": 0.3485550408118244,
+      "grad_norm": 1.0624901056289673,
+      "learning_rate": 0.00015135504481524122,
+      "loss": 0.4142,
+      "step": 4740
+    },
+    {
+      "epoch": 0.34862857563056104,
+      "grad_norm": 0.8300457000732422,
+      "learning_rate": 0.0001513346076595364,
+      "loss": 0.4021,
+      "step": 4741
+    },
+    {
+      "epoch": 0.3487021104492977,
+      "grad_norm": 1.0543749332427979,
+      "learning_rate": 0.0001513141675920784,
+      "loss": 0.412,
+      "step": 4742
+    },
+    {
+      "epoch": 0.3487756452680344,
+      "grad_norm": 1.1316008567810059,
+      "learning_rate": 0.00015129372461402668,
+      "loss": 0.4318,
+      "step": 4743
+    },
+    {
+      "epoch": 0.34884918008677107,
+      "grad_norm": 1.0393389463424683,
+      "learning_rate": 0.00015127327872654076,
+      "loss": 0.4556,
+      "step": 4744
+    },
+    {
+      "epoch": 0.34892271490550775,
+      "grad_norm": 0.7989851832389832,
+      "learning_rate": 0.00015125282993078037,
+      "loss": 0.4187,
+      "step": 4745
+    },
+    {
+      "epoch": 0.3489962497242444,
+      "grad_norm": 1.0232226848602295,
+      "learning_rate": 0.00015123237822790538,
+      "loss": 0.4435,
+      "step": 4746
+    },
+    {
+      "epoch": 0.3490697845429811,
+      "grad_norm": 0.6160323023796082,
+      "learning_rate": 0.00015121192361907583,
+      "loss": 0.4705,
+      "step": 4747
+    },
+    {
+      "epoch": 0.3491433193617178,
+      "grad_norm": 0.8041204214096069,
+      "learning_rate": 0.00015119146610545193,
+      "loss": 0.4089,
+      "step": 4748
+    },
+    {
+      "epoch": 0.34921685418045445,
+      "grad_norm": 0.5843124389648438,
+      "learning_rate": 0.00015117100568819405,
+      "loss": 0.4386,
+      "step": 4749
+    },
+    {
+      "epoch": 0.3492903889991911,
+      "grad_norm": 0.595270574092865,
+      "learning_rate": 0.0001511505423684627,
+      "loss": 0.3996,
+      "step": 4750
+    },
+    {
+      "epoch": 0.3493639238179278,
+      "grad_norm": 0.6051888465881348,
+      "learning_rate": 0.00015113007614741864,
+      "loss": 0.4182,
+      "step": 4751
+    },
+    {
+      "epoch": 0.3494374586366645,
+      "grad_norm": 0.6529853940010071,
+      "learning_rate": 0.00015110960702622266,
+      "loss": 0.4457,
+      "step": 4752
+    },
+    {
+      "epoch": 0.34951099345540115,
+      "grad_norm": 0.7962600588798523,
+      "learning_rate": 0.0001510891350060359,
+      "loss": 0.4599,
+      "step": 4753
+    },
+    {
+      "epoch": 0.34958452827413783,
+      "grad_norm": 0.9224130511283875,
+      "learning_rate": 0.00015106866008801943,
+      "loss": 0.434,
+      "step": 4754
+    },
+    {
+      "epoch": 0.34965806309287445,
+      "grad_norm": 0.6422299146652222,
+      "learning_rate": 0.00015104818227333467,
+      "loss": 0.4361,
+      "step": 4755
+    },
+    {
+      "epoch": 0.3497315979116111,
+      "grad_norm": 1.051575779914856,
+      "learning_rate": 0.00015102770156314314,
+      "loss": 0.3898,
+      "step": 4756
+    },
+    {
+      "epoch": 0.3498051327303478,
+      "grad_norm": 1.0988717079162598,
+      "learning_rate": 0.00015100721795860653,
+      "loss": 0.4441,
+      "step": 4757
+    },
+    {
+      "epoch": 0.3498786675490845,
+      "grad_norm": 0.6920530796051025,
+      "learning_rate": 0.0001509867314608867,
+      "loss": 0.4177,
+      "step": 4758
+    },
+    {
+      "epoch": 0.34995220236782115,
+      "grad_norm": 0.961667001247406,
+      "learning_rate": 0.00015096624207114562,
+      "loss": 0.4298,
+      "step": 4759
+    },
+    {
+      "epoch": 0.35002573718655783,
+      "grad_norm": 0.9656686186790466,
+      "learning_rate": 0.00015094574979054555,
+      "loss": 0.4506,
+      "step": 4760
+    },
+    {
+      "epoch": 0.3500992720052945,
+      "grad_norm": 0.6353631019592285,
+      "learning_rate": 0.00015092525462024876,
+      "loss": 0.4023,
+      "step": 4761
+    },
+    {
+      "epoch": 0.3501728068240312,
+      "grad_norm": 0.8487417101860046,
+      "learning_rate": 0.00015090475656141775,
+      "loss": 0.4099,
+      "step": 4762
+    },
+    {
+      "epoch": 0.35024634164276786,
+      "grad_norm": 1.0335243940353394,
+      "learning_rate": 0.00015088425561521527,
+      "loss": 0.4095,
+      "step": 4763
+    },
+    {
+      "epoch": 0.35031987646150453,
+      "grad_norm": 0.6253835558891296,
+      "learning_rate": 0.0001508637517828041,
+      "loss": 0.4332,
+      "step": 4764
+    },
+    {
+      "epoch": 0.3503934112802412,
+      "grad_norm": 0.6820023655891418,
+      "learning_rate": 0.00015084324506534728,
+      "loss": 0.4026,
+      "step": 4765
+    },
+    {
+      "epoch": 0.3504669460989779,
+      "grad_norm": 1.0236501693725586,
+      "learning_rate": 0.00015082273546400792,
+      "loss": 0.4198,
+      "step": 4766
+    },
+    {
+      "epoch": 0.35054048091771456,
+      "grad_norm": 0.7241955399513245,
+      "learning_rate": 0.00015080222297994932,
+      "loss": 0.4171,
+      "step": 4767
+    },
+    {
+      "epoch": 0.35061401573645123,
+      "grad_norm": 0.746955156326294,
+      "learning_rate": 0.0001507817076143351,
+      "loss": 0.432,
+      "step": 4768
+    },
+    {
+      "epoch": 0.35068755055518785,
+      "grad_norm": 0.6545815467834473,
+      "learning_rate": 0.00015076118936832876,
+      "loss": 0.4477,
+      "step": 4769
+    },
+    {
+      "epoch": 0.35076108537392453,
+      "grad_norm": 0.6341301798820496,
+      "learning_rate": 0.00015074066824309425,
+      "loss": 0.4773,
+      "step": 4770
+    },
+    {
+      "epoch": 0.3508346201926612,
+      "grad_norm": 0.793976902961731,
+      "learning_rate": 0.00015072014423979545,
+      "loss": 0.397,
+      "step": 4771
+    },
+    {
+      "epoch": 0.3509081550113979,
+      "grad_norm": 0.668289303779602,
+      "learning_rate": 0.00015069961735959656,
+      "loss": 0.4213,
+      "step": 4772
+    },
+    {
+      "epoch": 0.35098168983013456,
+      "grad_norm": 0.8561301827430725,
+      "learning_rate": 0.00015067908760366184,
+      "loss": 0.4226,
+      "step": 4773
+    },
+    {
+      "epoch": 0.35105522464887123,
+      "grad_norm": 0.6423175930976868,
+      "learning_rate": 0.00015065855497315583,
+      "loss": 0.4248,
+      "step": 4774
+    },
+    {
+      "epoch": 0.3511287594676079,
+      "grad_norm": 1.044184684753418,
+      "learning_rate": 0.0001506380194692431,
+      "loss": 0.4214,
+      "step": 4775
+    },
+    {
+      "epoch": 0.3512022942863446,
+      "grad_norm": 0.5943567752838135,
+      "learning_rate": 0.0001506174810930885,
+      "loss": 0.4146,
+      "step": 4776
+    },
+    {
+      "epoch": 0.35127582910508126,
+      "grad_norm": 0.9423234462738037,
+      "learning_rate": 0.00015059693984585687,
+      "loss": 0.4526,
+      "step": 4777
+    },
+    {
+      "epoch": 0.35134936392381794,
+      "grad_norm": 0.5873808264732361,
+      "learning_rate": 0.0001505763957287135,
+      "loss": 0.4224,
+      "step": 4778
+    },
+    {
+      "epoch": 0.3514228987425546,
+      "grad_norm": 1.1008095741271973,
+      "learning_rate": 0.00015055584874282352,
+      "loss": 0.407,
+      "step": 4779
+    },
+    {
+      "epoch": 0.3514964335612913,
+      "grad_norm": 0.9028363823890686,
+      "learning_rate": 0.00015053529888935248,
+      "loss": 0.4268,
+      "step": 4780
+    },
+    {
+      "epoch": 0.35156996838002796,
+      "grad_norm": 0.6663876175880432,
+      "learning_rate": 0.00015051474616946595,
+      "loss": 0.4147,
+      "step": 4781
+    },
+    {
+      "epoch": 0.35164350319876464,
+      "grad_norm": 0.8664133548736572,
+      "learning_rate": 0.00015049419058432972,
+      "loss": 0.4104,
+      "step": 4782
+    },
+    {
+      "epoch": 0.35171703801750126,
+      "grad_norm": 0.8919233083724976,
+      "learning_rate": 0.00015047363213510973,
+      "loss": 0.4564,
+      "step": 4783
+    },
+    {
+      "epoch": 0.35179057283623794,
+      "grad_norm": 1.083823323249817,
+      "learning_rate": 0.00015045307082297202,
+      "loss": 0.4254,
+      "step": 4784
+    },
+    {
+      "epoch": 0.3518641076549746,
+      "grad_norm": 0.7901788353919983,
+      "learning_rate": 0.00015043250664908286,
+      "loss": 0.4033,
+      "step": 4785
+    },
+    {
+      "epoch": 0.3519376424737113,
+      "grad_norm": 0.9697321057319641,
+      "learning_rate": 0.00015041193961460876,
+      "loss": 0.4365,
+      "step": 4786
+    },
+    {
+      "epoch": 0.35201117729244796,
+      "grad_norm": 1.6350971460342407,
+      "learning_rate": 0.00015039136972071622,
+      "loss": 0.4061,
+      "step": 4787
+    },
+    {
+      "epoch": 0.35208471211118464,
+      "grad_norm": 0.8983331918716431,
+      "learning_rate": 0.000150370796968572,
+      "loss": 0.429,
+      "step": 4788
+    },
+    {
+      "epoch": 0.3521582469299213,
+      "grad_norm": 1.8220933675765991,
+      "learning_rate": 0.00015035022135934301,
+      "loss": 0.3923,
+      "step": 4789
+    },
+    {
+      "epoch": 0.352231781748658,
+      "grad_norm": 1.2516180276870728,
+      "learning_rate": 0.00015032964289419636,
+      "loss": 0.3875,
+      "step": 4790
+    },
+    {
+      "epoch": 0.35230531656739467,
+      "grad_norm": 1.821452260017395,
+      "learning_rate": 0.00015030906157429924,
+      "loss": 0.4427,
+      "step": 4791
+    },
+    {
+      "epoch": 0.35237885138613134,
+      "grad_norm": 1.8738727569580078,
+      "learning_rate": 0.00015028847740081904,
+      "loss": 0.4155,
+      "step": 4792
+    },
+    {
+      "epoch": 0.352452386204868,
+      "grad_norm": 1.1947901248931885,
+      "learning_rate": 0.00015026789037492333,
+      "loss": 0.3898,
+      "step": 4793
+    },
+    {
+      "epoch": 0.3525259210236047,
+      "grad_norm": 0.9768665432929993,
+      "learning_rate": 0.00015024730049777987,
+      "loss": 0.4407,
+      "step": 4794
+    },
+    {
+      "epoch": 0.35259945584234137,
+      "grad_norm": 1.3151084184646606,
+      "learning_rate": 0.00015022670777055646,
+      "loss": 0.4313,
+      "step": 4795
+    },
+    {
+      "epoch": 0.35267299066107805,
+      "grad_norm": 0.9788702726364136,
+      "learning_rate": 0.00015020611219442116,
+      "loss": 0.4201,
+      "step": 4796
+    },
+    {
+      "epoch": 0.35274652547981467,
+      "grad_norm": 1.8954737186431885,
+      "learning_rate": 0.00015018551377054222,
+      "loss": 0.4009,
+      "step": 4797
+    },
+    {
+      "epoch": 0.35282006029855134,
+      "grad_norm": 2.0001697540283203,
+      "learning_rate": 0.00015016491250008796,
+      "loss": 0.4107,
+      "step": 4798
+    },
+    {
+      "epoch": 0.352893595117288,
+      "grad_norm": 1.1077319383621216,
+      "learning_rate": 0.00015014430838422693,
+      "loss": 0.4173,
+      "step": 4799
+    },
+    {
+      "epoch": 0.3529671299360247,
+      "grad_norm": 1.6779288053512573,
+      "learning_rate": 0.00015012370142412782,
+      "loss": 0.4028,
+      "step": 4800
+    },
+    {
+      "epoch": 0.35304066475476137,
+      "grad_norm": 1.2636544704437256,
+      "learning_rate": 0.00015010309162095945,
+      "loss": 0.4173,
+      "step": 4801
+    },
+    {
+      "epoch": 0.35311419957349804,
+      "grad_norm": 1.470211148262024,
+      "learning_rate": 0.00015008247897589088,
+      "loss": 0.4112,
+      "step": 4802
+    },
+    {
+      "epoch": 0.3531877343922347,
+      "grad_norm": 1.5277173519134521,
+      "learning_rate": 0.00015006186349009122,
+      "loss": 0.4039,
+      "step": 4803
+    },
+    {
+      "epoch": 0.3532612692109714,
+      "grad_norm": 1.0698801279067993,
+      "learning_rate": 0.0001500412451647298,
+      "loss": 0.4083,
+      "step": 4804
+    },
+    {
+      "epoch": 0.35333480402970807,
+      "grad_norm": 1.1106499433517456,
+      "learning_rate": 0.0001500206240009762,
+      "loss": 0.4024,
+      "step": 4805
+    },
+    {
+      "epoch": 0.35340833884844475,
+      "grad_norm": 0.8613413572311401,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.4308,
+      "step": 4806
+    },
+    {
+      "epoch": 0.3534818736671814,
+      "grad_norm": 1.1003516912460327,
+      "learning_rate": 0.00014997937316297107,
+      "loss": 0.4033,
+      "step": 4807
+    },
+    {
+      "epoch": 0.3535554084859181,
+      "grad_norm": 0.9390559792518616,
+      "learning_rate": 0.00014995874349105928,
+      "loss": 0.4148,
+      "step": 4808
+    },
+    {
+      "epoch": 0.3536289433046548,
+      "grad_norm": 0.9874455332756042,
+      "learning_rate": 0.00014993811098543486,
+      "loss": 0.4088,
+      "step": 4809
+    },
+    {
+      "epoch": 0.35370247812339145,
+      "grad_norm": 0.83491450548172,
+      "learning_rate": 0.0001499174756472681,
+      "loss": 0.3963,
+      "step": 4810
+    },
+    {
+      "epoch": 0.35377601294212807,
+      "grad_norm": 0.9899323582649231,
+      "learning_rate": 0.00014989683747772942,
+      "loss": 0.4051,
+      "step": 4811
+    },
+    {
+      "epoch": 0.35384954776086475,
+      "grad_norm": 1.1824452877044678,
+      "learning_rate": 0.00014987619647798946,
+      "loss": 0.4038,
+      "step": 4812
+    },
+    {
+      "epoch": 0.3539230825796014,
+      "grad_norm": 0.7569835186004639,
+      "learning_rate": 0.000149855552649219,
+      "loss": 0.4111,
+      "step": 4813
+    },
+    {
+      "epoch": 0.3539966173983381,
+      "grad_norm": 1.4468446969985962,
+      "learning_rate": 0.000149834905992589,
+      "loss": 0.4292,
+      "step": 4814
+    },
+    {
+      "epoch": 0.3540701522170748,
+      "grad_norm": 0.8968796730041504,
+      "learning_rate": 0.00014981425650927054,
+      "loss": 0.4368,
+      "step": 4815
+    },
+    {
+      "epoch": 0.35414368703581145,
+      "grad_norm": 1.884129524230957,
+      "learning_rate": 0.00014979360420043484,
+      "loss": 0.4228,
+      "step": 4816
+    },
+    {
+      "epoch": 0.3542172218545481,
+      "grad_norm": 1.6205358505249023,
+      "learning_rate": 0.00014977294906725344,
+      "loss": 0.4017,
+      "step": 4817
+    },
+    {
+      "epoch": 0.3542907566732848,
+      "grad_norm": 1.382553219795227,
+      "learning_rate": 0.00014975229111089774,
+      "loss": 0.4223,
+      "step": 4818
+    },
+    {
+      "epoch": 0.3543642914920215,
+      "grad_norm": 1.5113458633422852,
+      "learning_rate": 0.00014973163033253964,
+      "loss": 0.4577,
+      "step": 4819
+    },
+    {
+      "epoch": 0.35443782631075815,
+      "grad_norm": 1.3263285160064697,
+      "learning_rate": 0.00014971096673335098,
+      "loss": 0.3999,
+      "step": 4820
+    },
+    {
+      "epoch": 0.35451136112949483,
+      "grad_norm": 1.1614477634429932,
+      "learning_rate": 0.00014969030031450383,
+      "loss": 0.4715,
+      "step": 4821
+    },
+    {
+      "epoch": 0.3545848959482315,
+      "grad_norm": 1.7208830118179321,
+      "learning_rate": 0.00014966963107717037,
+      "loss": 0.4055,
+      "step": 4822
+    },
+    {
+      "epoch": 0.3546584307669682,
+      "grad_norm": 1.0833990573883057,
+      "learning_rate": 0.00014964895902252304,
+      "loss": 0.3971,
+      "step": 4823
+    },
+    {
+      "epoch": 0.35473196558570486,
+      "grad_norm": 2.1717123985290527,
+      "learning_rate": 0.00014962828415173438,
+      "loss": 0.4184,
+      "step": 4824
+    },
+    {
+      "epoch": 0.3548055004044415,
+      "grad_norm": 1.7330808639526367,
+      "learning_rate": 0.00014960760646597705,
+      "loss": 0.3984,
+      "step": 4825
+    },
+    {
+      "epoch": 0.35487903522317815,
+      "grad_norm": 1.412564754486084,
+      "learning_rate": 0.00014958692596642393,
+      "loss": 0.3962,
+      "step": 4826
+    },
+    {
+      "epoch": 0.3549525700419148,
+      "grad_norm": 1.2361804246902466,
+      "learning_rate": 0.00014956624265424802,
+      "loss": 0.4145,
+      "step": 4827
+    },
+    {
+      "epoch": 0.3550261048606515,
+      "grad_norm": 1.4870113134384155,
+      "learning_rate": 0.00014954555653062254,
+      "loss": 0.4265,
+      "step": 4828
+    },
+    {
+      "epoch": 0.3550996396793882,
+      "grad_norm": 1.1685899496078491,
+      "learning_rate": 0.00014952486759672078,
+      "loss": 0.4147,
+      "step": 4829
+    },
+    {
+      "epoch": 0.35517317449812486,
+      "grad_norm": 1.6464279890060425,
+      "learning_rate": 0.0001495041758537163,
+      "loss": 0.4071,
+      "step": 4830
+    },
+    {
+      "epoch": 0.35524670931686153,
+      "grad_norm": 1.4520907402038574,
+      "learning_rate": 0.00014948348130278268,
+      "loss": 0.4144,
+      "step": 4831
+    },
+    {
+      "epoch": 0.3553202441355982,
+      "grad_norm": 1.27268648147583,
+      "learning_rate": 0.00014946278394509382,
+      "loss": 0.4123,
+      "step": 4832
+    },
+    {
+      "epoch": 0.3553937789543349,
+      "grad_norm": 1.326197624206543,
+      "learning_rate": 0.00014944208378182364,
+      "loss": 0.4525,
+      "step": 4833
+    },
+    {
+      "epoch": 0.35546731377307156,
+      "grad_norm": 1.6107929944992065,
+      "learning_rate": 0.00014942138081414628,
+      "loss": 0.4094,
+      "step": 4834
+    },
+    {
+      "epoch": 0.35554084859180823,
+      "grad_norm": 1.1316583156585693,
+      "learning_rate": 0.00014940067504323607,
+      "loss": 0.4233,
+      "step": 4835
+    },
+    {
+      "epoch": 0.3556143834105449,
+      "grad_norm": 1.7245081663131714,
+      "learning_rate": 0.0001493799664702674,
+      "loss": 0.4098,
+      "step": 4836
+    },
+    {
+      "epoch": 0.3556879182292816,
+      "grad_norm": 1.643394947052002,
+      "learning_rate": 0.00014935925509641492,
+      "loss": 0.4206,
+      "step": 4837
+    },
+    {
+      "epoch": 0.35576145304801826,
+      "grad_norm": 1.4204246997833252,
+      "learning_rate": 0.00014933854092285344,
+      "loss": 0.3997,
+      "step": 4838
+    },
+    {
+      "epoch": 0.3558349878667549,
+      "grad_norm": 1.4776952266693115,
+      "learning_rate": 0.00014931782395075782,
+      "loss": 0.4195,
+      "step": 4839
+    },
+    {
+      "epoch": 0.35590852268549156,
+      "grad_norm": 1.2723568677902222,
+      "learning_rate": 0.00014929710418130316,
+      "loss": 0.4199,
+      "step": 4840
+    },
+    {
+      "epoch": 0.35598205750422823,
+      "grad_norm": 1.2936501502990723,
+      "learning_rate": 0.00014927638161566477,
+      "loss": 0.4553,
+      "step": 4841
+    },
+    {
+      "epoch": 0.3560555923229649,
+      "grad_norm": 1.6092628240585327,
+      "learning_rate": 0.00014925565625501799,
+      "loss": 0.3975,
+      "step": 4842
+    },
+    {
+      "epoch": 0.3561291271417016,
+      "grad_norm": 1.0771235227584839,
+      "learning_rate": 0.0001492349281005384,
+      "loss": 0.3991,
+      "step": 4843
+    },
+    {
+      "epoch": 0.35620266196043826,
+      "grad_norm": 1.4885989427566528,
+      "learning_rate": 0.00014921419715340173,
+      "loss": 0.4523,
+      "step": 4844
+    },
+    {
+      "epoch": 0.35627619677917494,
+      "grad_norm": 1.4145534038543701,
+      "learning_rate": 0.0001491934634147839,
+      "loss": 0.4048,
+      "step": 4845
+    },
+    {
+      "epoch": 0.3563497315979116,
+      "grad_norm": 1.7540035247802734,
+      "learning_rate": 0.00014917272688586084,
+      "loss": 0.4187,
+      "step": 4846
+    },
+    {
+      "epoch": 0.3564232664166483,
+      "grad_norm": 1.668099284172058,
+      "learning_rate": 0.00014915198756780887,
+      "loss": 0.428,
+      "step": 4847
+    },
+    {
+      "epoch": 0.35649680123538496,
+      "grad_norm": 1.3056317567825317,
+      "learning_rate": 0.00014913124546180425,
+      "loss": 0.4015,
+      "step": 4848
+    },
+    {
+      "epoch": 0.35657033605412164,
+      "grad_norm": 1.4041396379470825,
+      "learning_rate": 0.00014911050056902356,
+      "loss": 0.4316,
+      "step": 4849
+    },
+    {
+      "epoch": 0.3566438708728583,
+      "grad_norm": 1.248016119003296,
+      "learning_rate": 0.00014908975289064347,
+      "loss": 0.4106,
+      "step": 4850
+    },
+    {
+      "epoch": 0.356717405691595,
+      "grad_norm": 1.0657660961151123,
+      "learning_rate": 0.00014906900242784077,
+      "loss": 0.3937,
+      "step": 4851
+    },
+    {
+      "epoch": 0.35679094051033167,
+      "grad_norm": 1.4420576095581055,
+      "learning_rate": 0.00014904824918179248,
+      "loss": 0.397,
+      "step": 4852
+    },
+    {
+      "epoch": 0.3568644753290683,
+      "grad_norm": 1.018426537513733,
+      "learning_rate": 0.0001490274931536757,
+      "loss": 0.4251,
+      "step": 4853
+    },
+    {
+      "epoch": 0.35693801014780496,
+      "grad_norm": 1.6903795003890991,
+      "learning_rate": 0.0001490067343446678,
+      "loss": 0.421,
+      "step": 4854
+    },
+    {
+      "epoch": 0.35701154496654164,
+      "grad_norm": 1.68545401096344,
+      "learning_rate": 0.00014898597275594618,
+      "loss": 0.4161,
+      "step": 4855
+    },
+    {
+      "epoch": 0.3570850797852783,
+      "grad_norm": 1.497138261795044,
+      "learning_rate": 0.00014896520838868853,
+      "loss": 0.4069,
+      "step": 4856
+    },
+    {
+      "epoch": 0.357158614604015,
+      "grad_norm": 1.4905441999435425,
+      "learning_rate": 0.00014894444124407253,
+      "loss": 0.4144,
+      "step": 4857
+    },
+    {
+      "epoch": 0.35723214942275167,
+      "grad_norm": 1.0069483518600464,
+      "learning_rate": 0.00014892367132327624,
+      "loss": 0.3863,
+      "step": 4858
+    },
+    {
+      "epoch": 0.35730568424148834,
+      "grad_norm": 1.1909483671188354,
+      "learning_rate": 0.00014890289862747762,
+      "loss": 0.4219,
+      "step": 4859
+    },
+    {
+      "epoch": 0.357379219060225,
+      "grad_norm": 1.0062235593795776,
+      "learning_rate": 0.00014888212315785498,
+      "loss": 0.423,
+      "step": 4860
+    },
+    {
+      "epoch": 0.3574527538789617,
+      "grad_norm": 0.8276734948158264,
+      "learning_rate": 0.00014886134491558678,
+      "loss": 0.4014,
+      "step": 4861
+    },
+    {
+      "epoch": 0.35752628869769837,
+      "grad_norm": 1.2217323780059814,
+      "learning_rate": 0.00014884056390185147,
+      "loss": 0.4092,
+      "step": 4862
+    },
+    {
+      "epoch": 0.35759982351643504,
+      "grad_norm": 0.6547780632972717,
+      "learning_rate": 0.00014881978011782788,
+      "loss": 0.4344,
+      "step": 4863
+    },
+    {
+      "epoch": 0.3576733583351717,
+      "grad_norm": 1.3956083059310913,
+      "learning_rate": 0.00014879899356469483,
+      "loss": 0.3989,
+      "step": 4864
+    },
+    {
+      "epoch": 0.3577468931539084,
+      "grad_norm": 0.7567898035049438,
+      "learning_rate": 0.00014877820424363136,
+      "loss": 0.3885,
+      "step": 4865
+    },
+    {
+      "epoch": 0.35782042797264507,
+      "grad_norm": 1.6147551536560059,
+      "learning_rate": 0.00014875741215581668,
+      "loss": 0.4378,
+      "step": 4866
+    },
+    {
+      "epoch": 0.3578939627913817,
+      "grad_norm": 1.1165851354599,
+      "learning_rate": 0.0001487366173024301,
+      "loss": 0.4136,
+      "step": 4867
+    },
+    {
+      "epoch": 0.35796749761011837,
+      "grad_norm": 1.7348003387451172,
+      "learning_rate": 0.00014871581968465116,
+      "loss": 0.422,
+      "step": 4868
+    },
+    {
+      "epoch": 0.35804103242885504,
+      "grad_norm": 1.625601053237915,
+      "learning_rate": 0.0001486950193036595,
+      "loss": 0.4168,
+      "step": 4869
+    },
+    {
+      "epoch": 0.3581145672475917,
+      "grad_norm": 1.286847472190857,
+      "learning_rate": 0.00014867421616063503,
+      "loss": 0.439,
+      "step": 4870
+    },
+    {
+      "epoch": 0.3581881020663284,
+      "grad_norm": 1.2481893301010132,
+      "learning_rate": 0.00014865341025675758,
+      "loss": 0.3948,
+      "step": 4871
+    },
+    {
+      "epoch": 0.35826163688506507,
+      "grad_norm": 1.5188153982162476,
+      "learning_rate": 0.00014863260159320735,
+      "loss": 0.388,
+      "step": 4872
+    },
+    {
+      "epoch": 0.35833517170380175,
+      "grad_norm": 0.9535976052284241,
+      "learning_rate": 0.00014861179017116466,
+      "loss": 0.3818,
+      "step": 4873
+    },
+    {
+      "epoch": 0.3584087065225384,
+      "grad_norm": 1.60862135887146,
+      "learning_rate": 0.00014859097599180993,
+      "loss": 0.4216,
+      "step": 4874
+    },
+    {
+      "epoch": 0.3584822413412751,
+      "grad_norm": 1.0827770233154297,
+      "learning_rate": 0.00014857015905632374,
+      "loss": 0.4483,
+      "step": 4875
+    },
+    {
+      "epoch": 0.3585557761600118,
+      "grad_norm": 1.9882383346557617,
+      "learning_rate": 0.00014854933936588692,
+      "loss": 0.4366,
+      "step": 4876
+    },
+    {
+      "epoch": 0.35862931097874845,
+      "grad_norm": 1.5756489038467407,
+      "learning_rate": 0.00014852851692168028,
+      "loss": 0.4373,
+      "step": 4877
+    },
+    {
+      "epoch": 0.3587028457974851,
+      "grad_norm": 1.3528766632080078,
+      "learning_rate": 0.00014850769172488498,
+      "loss": 0.4213,
+      "step": 4878
+    },
+    {
+      "epoch": 0.3587763806162218,
+      "grad_norm": 1.26084303855896,
+      "learning_rate": 0.0001484868637766822,
+      "loss": 0.4353,
+      "step": 4879
+    },
+    {
+      "epoch": 0.3588499154349585,
+      "grad_norm": 1.4754844903945923,
+      "learning_rate": 0.00014846603307825334,
+      "loss": 0.4072,
+      "step": 4880
+    },
+    {
+      "epoch": 0.3589234502536951,
+      "grad_norm": 1.3705557584762573,
+      "learning_rate": 0.00014844519963077993,
+      "loss": 0.4132,
+      "step": 4881
+    },
+    {
+      "epoch": 0.3589969850724318,
+      "grad_norm": 1.3496484756469727,
+      "learning_rate": 0.0001484243634354437,
+      "loss": 0.4424,
+      "step": 4882
+    },
+    {
+      "epoch": 0.35907051989116845,
+      "grad_norm": 1.270382046699524,
+      "learning_rate": 0.00014840352449342641,
+      "loss": 0.4072,
+      "step": 4883
+    },
+    {
+      "epoch": 0.3591440547099051,
+      "grad_norm": 1.345586895942688,
+      "learning_rate": 0.0001483826828059102,
+      "loss": 0.4127,
+      "step": 4884
+    },
+    {
+      "epoch": 0.3592175895286418,
+      "grad_norm": 1.0399683713912964,
+      "learning_rate": 0.0001483618383740771,
+      "loss": 0.4024,
+      "step": 4885
+    },
+    {
+      "epoch": 0.3592911243473785,
+      "grad_norm": 1.641903042793274,
+      "learning_rate": 0.0001483409911991095,
+      "loss": 0.403,
+      "step": 4886
+    },
+    {
+      "epoch": 0.35936465916611515,
+      "grad_norm": 1.270461082458496,
+      "learning_rate": 0.0001483201412821899,
+      "loss": 0.4256,
+      "step": 4887
+    },
+    {
+      "epoch": 0.35943819398485183,
+      "grad_norm": 1.2307161092758179,
+      "learning_rate": 0.00014829928862450086,
+      "loss": 0.4189,
+      "step": 4888
+    },
+    {
+      "epoch": 0.3595117288035885,
+      "grad_norm": 0.9978452324867249,
+      "learning_rate": 0.0001482784332272252,
+      "loss": 0.4073,
+      "step": 4889
+    },
+    {
+      "epoch": 0.3595852636223252,
+      "grad_norm": 1.4097927808761597,
+      "learning_rate": 0.00014825757509154586,
+      "loss": 0.4222,
+      "step": 4890
+    },
+    {
+      "epoch": 0.35965879844106186,
+      "grad_norm": 1.135522723197937,
+      "learning_rate": 0.00014823671421864594,
+      "loss": 0.4351,
+      "step": 4891
+    },
+    {
+      "epoch": 0.35973233325979853,
+      "grad_norm": 1.9509438276290894,
+      "learning_rate": 0.0001482158506097087,
+      "loss": 0.4171,
+      "step": 4892
+    },
+    {
+      "epoch": 0.3598058680785352,
+      "grad_norm": 1.6556981801986694,
+      "learning_rate": 0.00014819498426591748,
+      "loss": 0.4043,
+      "step": 4893
+    },
+    {
+      "epoch": 0.3598794028972719,
+      "grad_norm": 1.0072821378707886,
+      "learning_rate": 0.00014817411518845594,
+      "loss": 0.3956,
+      "step": 4894
+    },
+    {
+      "epoch": 0.3599529377160085,
+      "grad_norm": 1.3222860097885132,
+      "learning_rate": 0.0001481532433785077,
+      "loss": 0.4167,
+      "step": 4895
+    },
+    {
+      "epoch": 0.3600264725347452,
+      "grad_norm": 1.2225377559661865,
+      "learning_rate": 0.00014813236883725673,
+      "loss": 0.415,
+      "step": 4896
+    },
+    {
+      "epoch": 0.36010000735348185,
+      "grad_norm": 0.9053601026535034,
+      "learning_rate": 0.00014811149156588696,
+      "loss": 0.4052,
+      "step": 4897
+    },
+    {
+      "epoch": 0.36017354217221853,
+      "grad_norm": 1.4101037979125977,
+      "learning_rate": 0.00014809061156558263,
+      "loss": 0.4139,
+      "step": 4898
+    },
+    {
+      "epoch": 0.3602470769909552,
+      "grad_norm": 1.028584361076355,
+      "learning_rate": 0.00014806972883752809,
+      "loss": 0.3857,
+      "step": 4899
+    },
+    {
+      "epoch": 0.3603206118096919,
+      "grad_norm": 1.5215423107147217,
+      "learning_rate": 0.00014804884338290774,
+      "loss": 0.4157,
+      "step": 4900
+    },
+    {
+      "epoch": 0.36039414662842856,
+      "grad_norm": 1.233782172203064,
+      "learning_rate": 0.00014802795520290634,
+      "loss": 0.4379,
+      "step": 4901
+    },
+    {
+      "epoch": 0.36046768144716523,
+      "grad_norm": 1.8275291919708252,
+      "learning_rate": 0.00014800706429870858,
+      "loss": 0.421,
+      "step": 4902
+    },
+    {
+      "epoch": 0.3605412162659019,
+      "grad_norm": 1.445388913154602,
+      "learning_rate": 0.0001479861706714995,
+      "loss": 0.4107,
+      "step": 4903
+    },
+    {
+      "epoch": 0.3606147510846386,
+      "grad_norm": 1.3957412242889404,
+      "learning_rate": 0.0001479652743224642,
+      "loss": 0.4235,
+      "step": 4904
+    },
+    {
+      "epoch": 0.36068828590337526,
+      "grad_norm": 1.2421716451644897,
+      "learning_rate": 0.0001479443752527879,
+      "loss": 0.4102,
+      "step": 4905
+    },
+    {
+      "epoch": 0.36076182072211194,
+      "grad_norm": 1.5840650796890259,
+      "learning_rate": 0.00014792347346365603,
+      "loss": 0.4126,
+      "step": 4906
+    },
+    {
+      "epoch": 0.3608353555408486,
+      "grad_norm": 1.1768620014190674,
+      "learning_rate": 0.00014790256895625417,
+      "loss": 0.4273,
+      "step": 4907
+    },
+    {
+      "epoch": 0.3609088903595853,
+      "grad_norm": 1.6453810930252075,
+      "learning_rate": 0.00014788166173176803,
+      "loss": 0.4399,
+      "step": 4908
+    },
+    {
+      "epoch": 0.3609824251783219,
+      "grad_norm": 1.3430944681167603,
+      "learning_rate": 0.00014786075179138355,
+      "loss": 0.4162,
+      "step": 4909
+    },
+    {
+      "epoch": 0.3610559599970586,
+      "grad_norm": 1.5645928382873535,
+      "learning_rate": 0.00014783983913628667,
+      "loss": 0.419,
+      "step": 4910
+    },
+    {
+      "epoch": 0.36112949481579526,
+      "grad_norm": 1.3678542375564575,
+      "learning_rate": 0.00014781892376766365,
+      "loss": 0.3951,
+      "step": 4911
+    },
+    {
+      "epoch": 0.36120302963453194,
+      "grad_norm": 1.324373483657837,
+      "learning_rate": 0.00014779800568670082,
+      "loss": 0.4132,
+      "step": 4912
+    },
+    {
+      "epoch": 0.3612765644532686,
+      "grad_norm": 1.109927773475647,
+      "learning_rate": 0.00014777708489458467,
+      "loss": 0.4104,
+      "step": 4913
+    },
+    {
+      "epoch": 0.3613500992720053,
+      "grad_norm": 1.4798014163970947,
+      "learning_rate": 0.00014775616139250182,
+      "loss": 0.3948,
+      "step": 4914
+    },
+    {
+      "epoch": 0.36142363409074196,
+      "grad_norm": 1.13921058177948,
+      "learning_rate": 0.0001477352351816391,
+      "loss": 0.401,
+      "step": 4915
+    },
+    {
+      "epoch": 0.36149716890947864,
+      "grad_norm": 1.5383341312408447,
+      "learning_rate": 0.00014771430626318345,
+      "loss": 0.4002,
+      "step": 4916
+    },
+    {
+      "epoch": 0.3615707037282153,
+      "grad_norm": 1.364884376525879,
+      "learning_rate": 0.00014769337463832206,
+      "loss": 0.4036,
+      "step": 4917
+    },
+    {
+      "epoch": 0.361644238546952,
+      "grad_norm": 1.451902985572815,
+      "learning_rate": 0.00014767244030824205,
+      "loss": 0.3939,
+      "step": 4918
+    },
+    {
+      "epoch": 0.36171777336568867,
+      "grad_norm": 1.2103501558303833,
+      "learning_rate": 0.00014765150327413095,
+      "loss": 0.4039,
+      "step": 4919
+    },
+    {
+      "epoch": 0.36179130818442534,
+      "grad_norm": 1.509482979774475,
+      "learning_rate": 0.0001476305635371763,
+      "loss": 0.4186,
+      "step": 4920
+    },
+    {
+      "epoch": 0.361864843003162,
+      "grad_norm": 1.0770142078399658,
+      "learning_rate": 0.0001476096210985658,
+      "loss": 0.3941,
+      "step": 4921
+    },
+    {
+      "epoch": 0.3619383778218987,
+      "grad_norm": 1.5505980253219604,
+      "learning_rate": 0.00014758867595948737,
+      "loss": 0.429,
+      "step": 4922
+    },
+    {
+      "epoch": 0.3620119126406353,
+      "grad_norm": 1.1988955736160278,
+      "learning_rate": 0.000147567728121129,
+      "loss": 0.4072,
+      "step": 4923
+    },
+    {
+      "epoch": 0.362085447459372,
+      "grad_norm": 1.7752659320831299,
+      "learning_rate": 0.0001475467775846789,
+      "loss": 0.4236,
+      "step": 4924
+    },
+    {
+      "epoch": 0.36215898227810867,
+      "grad_norm": 1.3991492986679077,
+      "learning_rate": 0.00014752582435132537,
+      "loss": 0.4132,
+      "step": 4925
+    },
+    {
+      "epoch": 0.36223251709684534,
+      "grad_norm": 1.3798446655273438,
+      "learning_rate": 0.00014750486842225695,
+      "loss": 0.3913,
+      "step": 4926
+    },
+    {
+      "epoch": 0.362306051915582,
+      "grad_norm": 1.1781492233276367,
+      "learning_rate": 0.00014748390979866228,
+      "loss": 0.426,
+      "step": 4927
+    },
+    {
+      "epoch": 0.3623795867343187,
+      "grad_norm": 1.4907227754592896,
+      "learning_rate": 0.0001474629484817301,
+      "loss": 0.451,
+      "step": 4928
+    },
+    {
+      "epoch": 0.36245312155305537,
+      "grad_norm": 1.4201298952102661,
+      "learning_rate": 0.0001474419844726494,
+      "loss": 0.4163,
+      "step": 4929
+    },
+    {
+      "epoch": 0.36252665637179204,
+      "grad_norm": 1.3643873929977417,
+      "learning_rate": 0.0001474210177726093,
+      "loss": 0.3944,
+      "step": 4930
+    },
+    {
+      "epoch": 0.3626001911905287,
+      "grad_norm": 1.5550377368927002,
+      "learning_rate": 0.000147400048382799,
+      "loss": 0.43,
+      "step": 4931
+    },
+    {
+      "epoch": 0.3626737260092654,
+      "grad_norm": 1.0708537101745605,
+      "learning_rate": 0.00014737907630440791,
+      "loss": 0.4286,
+      "step": 4932
+    },
+    {
+      "epoch": 0.36274726082800207,
+      "grad_norm": 1.1611113548278809,
+      "learning_rate": 0.00014735810153862565,
+      "loss": 0.4066,
+      "step": 4933
+    },
+    {
+      "epoch": 0.36282079564673875,
+      "grad_norm": 1.2909069061279297,
+      "learning_rate": 0.00014733712408664188,
+      "loss": 0.4172,
+      "step": 4934
+    },
+    {
+      "epoch": 0.3628943304654754,
+      "grad_norm": 0.7753896117210388,
+      "learning_rate": 0.00014731614394964648,
+      "loss": 0.4126,
+      "step": 4935
+    },
+    {
+      "epoch": 0.3629678652842121,
+      "grad_norm": 1.8478682041168213,
+      "learning_rate": 0.00014729516112882944,
+      "loss": 0.4229,
+      "step": 4936
+    },
+    {
+      "epoch": 0.3630414001029487,
+      "grad_norm": 1.0739585161209106,
+      "learning_rate": 0.00014727417562538096,
+      "loss": 0.4145,
+      "step": 4937
+    },
+    {
+      "epoch": 0.3631149349216854,
+      "grad_norm": 2.468766212463379,
+      "learning_rate": 0.00014725318744049137,
+      "loss": 0.4421,
+      "step": 4938
+    },
+    {
+      "epoch": 0.36318846974042207,
+      "grad_norm": 2.4803614616394043,
+      "learning_rate": 0.0001472321965753511,
+      "loss": 0.4312,
+      "step": 4939
+    },
+    {
+      "epoch": 0.36326200455915875,
+      "grad_norm": 0.6571427583694458,
+      "learning_rate": 0.00014721120303115082,
+      "loss": 0.4256,
+      "step": 4940
+    },
+    {
+      "epoch": 0.3633355393778954,
+      "grad_norm": 1.86883544921875,
+      "learning_rate": 0.00014719020680908123,
+      "loss": 0.419,
+      "step": 4941
+    },
+    {
+      "epoch": 0.3634090741966321,
+      "grad_norm": 0.9315677285194397,
+      "learning_rate": 0.00014716920791033332,
+      "loss": 0.4044,
+      "step": 4942
+    },
+    {
+      "epoch": 0.3634826090153688,
+      "grad_norm": 2.0523719787597656,
+      "learning_rate": 0.00014714820633609819,
+      "loss": 0.4097,
+      "step": 4943
+    },
+    {
+      "epoch": 0.36355614383410545,
+      "grad_norm": 2.1419615745544434,
+      "learning_rate": 0.000147127202087567,
+      "loss": 0.399,
+      "step": 4944
+    },
+    {
+      "epoch": 0.3636296786528421,
+      "grad_norm": 1.096569299697876,
+      "learning_rate": 0.0001471061951659312,
+      "loss": 0.4129,
+      "step": 4945
+    },
+    {
+      "epoch": 0.3637032134715788,
+      "grad_norm": 1.0699163675308228,
+      "learning_rate": 0.0001470851855723823,
+      "loss": 0.4017,
+      "step": 4946
+    },
+    {
+      "epoch": 0.3637767482903155,
+      "grad_norm": 1.301699161529541,
+      "learning_rate": 0.00014706417330811198,
+      "loss": 0.4261,
+      "step": 4947
+    },
+    {
+      "epoch": 0.36385028310905215,
+      "grad_norm": 0.9124700427055359,
+      "learning_rate": 0.00014704315837431209,
+      "loss": 0.3917,
+      "step": 4948
+    },
+    {
+      "epoch": 0.36392381792778883,
+      "grad_norm": 1.5361807346343994,
+      "learning_rate": 0.0001470221407721746,
+      "loss": 0.4124,
+      "step": 4949
+    },
+    {
+      "epoch": 0.3639973527465255,
+      "grad_norm": 1.0234425067901611,
+      "learning_rate": 0.00014700112050289164,
+      "loss": 0.4097,
+      "step": 4950
+    },
+    {
+      "epoch": 0.3640708875652621,
+      "grad_norm": 1.867141604423523,
+      "learning_rate": 0.0001469800975676555,
+      "loss": 0.4175,
+      "step": 4951
+    },
+    {
+      "epoch": 0.3641444223839988,
+      "grad_norm": 1.4308217763900757,
+      "learning_rate": 0.00014695907196765872,
+      "loss": 0.4403,
+      "step": 4952
+    },
+    {
+      "epoch": 0.3642179572027355,
+      "grad_norm": 1.5180213451385498,
+      "learning_rate": 0.00014693804370409378,
+      "loss": 0.4196,
+      "step": 4953
+    },
+    {
+      "epoch": 0.36429149202147215,
+      "grad_norm": 1.5127155780792236,
+      "learning_rate": 0.00014691701277815347,
+      "loss": 0.4124,
+      "step": 4954
+    },
+    {
+      "epoch": 0.3643650268402088,
+      "grad_norm": 1.2058600187301636,
+      "learning_rate": 0.0001468959791910307,
+      "loss": 0.4268,
+      "step": 4955
+    },
+    {
+      "epoch": 0.3644385616589455,
+      "grad_norm": 1.3611297607421875,
+      "learning_rate": 0.00014687494294391847,
+      "loss": 0.433,
+      "step": 4956
+    },
+    {
+      "epoch": 0.3645120964776822,
+      "grad_norm": 1.3411983251571655,
+      "learning_rate": 0.00014685390403801002,
+      "loss": 0.4156,
+      "step": 4957
+    },
+    {
+      "epoch": 0.36458563129641886,
+      "grad_norm": 1.2192347049713135,
+      "learning_rate": 0.0001468328624744987,
+      "loss": 0.4074,
+      "step": 4958
+    },
+    {
+      "epoch": 0.36465916611515553,
+      "grad_norm": 1.5124156475067139,
+      "learning_rate": 0.00014681181825457795,
+      "loss": 0.3834,
+      "step": 4959
+    },
+    {
+      "epoch": 0.3647327009338922,
+      "grad_norm": 1.7371666431427002,
+      "learning_rate": 0.0001467907713794415,
+      "loss": 0.4298,
+      "step": 4960
+    },
+    {
+      "epoch": 0.3648062357526289,
+      "grad_norm": 1.162608027458191,
+      "learning_rate": 0.00014676972185028312,
+      "loss": 0.4276,
+      "step": 4961
+    },
+    {
+      "epoch": 0.36487977057136556,
+      "grad_norm": 0.9573419690132141,
+      "learning_rate": 0.00014674866966829674,
+      "loss": 0.4332,
+      "step": 4962
+    },
+    {
+      "epoch": 0.36495330539010223,
+      "grad_norm": 0.9457972645759583,
+      "learning_rate": 0.00014672761483467653,
+      "loss": 0.3789,
+      "step": 4963
+    },
+    {
+      "epoch": 0.3650268402088389,
+      "grad_norm": 0.8380709290504456,
+      "learning_rate": 0.00014670655735061661,
+      "loss": 0.4052,
+      "step": 4964
+    },
+    {
+      "epoch": 0.36510037502757553,
+      "grad_norm": 0.975319504737854,
+      "learning_rate": 0.00014668549721731154,
+      "loss": 0.4021,
+      "step": 4965
+    },
+    {
+      "epoch": 0.3651739098463122,
+      "grad_norm": 0.6358691453933716,
+      "learning_rate": 0.00014666443443595576,
+      "loss": 0.4111,
+      "step": 4966
+    },
+    {
+      "epoch": 0.3652474446650489,
+      "grad_norm": 0.8671440482139587,
+      "learning_rate": 0.000146643369007744,
+      "loss": 0.4334,
+      "step": 4967
+    },
+    {
+      "epoch": 0.36532097948378556,
+      "grad_norm": 0.9098048210144043,
+      "learning_rate": 0.00014662230093387117,
+      "loss": 0.4206,
+      "step": 4968
+    },
+    {
+      "epoch": 0.36539451430252223,
+      "grad_norm": 0.7617952823638916,
+      "learning_rate": 0.0001466012302155322,
+      "loss": 0.4089,
+      "step": 4969
+    },
+    {
+      "epoch": 0.3654680491212589,
+      "grad_norm": 0.7437771558761597,
+      "learning_rate": 0.0001465801568539223,
+      "loss": 0.4146,
+      "step": 4970
+    },
+    {
+      "epoch": 0.3655415839399956,
+      "grad_norm": 0.7982977628707886,
+      "learning_rate": 0.0001465590808502367,
+      "loss": 0.4028,
+      "step": 4971
+    },
+    {
+      "epoch": 0.36561511875873226,
+      "grad_norm": 0.6886052489280701,
+      "learning_rate": 0.0001465380022056709,
+      "loss": 0.4126,
+      "step": 4972
+    },
+    {
+      "epoch": 0.36568865357746894,
+      "grad_norm": 1.5190761089324951,
+      "learning_rate": 0.00014651692092142055,
+      "loss": 0.4286,
+      "step": 4973
+    },
+    {
+      "epoch": 0.3657621883962056,
+      "grad_norm": 0.6544710993766785,
+      "learning_rate": 0.00014649583699868131,
+      "loss": 0.4245,
+      "step": 4974
+    },
+    {
+      "epoch": 0.3658357232149423,
+      "grad_norm": 1.275214433670044,
+      "learning_rate": 0.00014647475043864916,
+      "loss": 0.3842,
+      "step": 4975
+    },
+    {
+      "epoch": 0.36590925803367896,
+      "grad_norm": 0.642104983329773,
+      "learning_rate": 0.0001464536612425201,
+      "loss": 0.4048,
+      "step": 4976
+    },
+    {
+      "epoch": 0.36598279285241564,
+      "grad_norm": 1.2338820695877075,
+      "learning_rate": 0.00014643256941149034,
+      "loss": 0.4294,
+      "step": 4977
+    },
+    {
+      "epoch": 0.3660563276711523,
+      "grad_norm": 0.8585917949676514,
+      "learning_rate": 0.00014641147494675626,
+      "loss": 0.4169,
+      "step": 4978
+    },
+    {
+      "epoch": 0.36612986248988894,
+      "grad_norm": 0.7730868458747864,
+      "learning_rate": 0.00014639037784951433,
+      "loss": 0.4419,
+      "step": 4979
+    },
+    {
+      "epoch": 0.3662033973086256,
+      "grad_norm": 1.2569084167480469,
+      "learning_rate": 0.00014636927812096122,
+      "loss": 0.4219,
+      "step": 4980
+    },
+    {
+      "epoch": 0.3662769321273623,
+      "grad_norm": 0.7322567701339722,
+      "learning_rate": 0.00014634817576229372,
+      "loss": 0.4643,
+      "step": 4981
+    },
+    {
+      "epoch": 0.36635046694609896,
+      "grad_norm": 0.879564106464386,
+      "learning_rate": 0.00014632707077470876,
+      "loss": 0.4307,
+      "step": 4982
+    },
+    {
+      "epoch": 0.36642400176483564,
+      "grad_norm": 1.0324656963348389,
+      "learning_rate": 0.00014630596315940345,
+      "loss": 0.4233,
+      "step": 4983
+    },
+    {
+      "epoch": 0.3664975365835723,
+      "grad_norm": 0.7575100064277649,
+      "learning_rate": 0.00014628485291757508,
+      "loss": 0.4241,
+      "step": 4984
+    },
+    {
+      "epoch": 0.366571071402309,
+      "grad_norm": 0.726399838924408,
+      "learning_rate": 0.00014626374005042097,
+      "loss": 0.4201,
+      "step": 4985
+    },
+    {
+      "epoch": 0.36664460622104567,
+      "grad_norm": 1.047380805015564,
+      "learning_rate": 0.00014624262455913871,
+      "loss": 0.4084,
+      "step": 4986
+    },
+    {
+      "epoch": 0.36671814103978234,
+      "grad_norm": 0.9113968014717102,
+      "learning_rate": 0.000146221506444926,
+      "loss": 0.4293,
+      "step": 4987
+    },
+    {
+      "epoch": 0.366791675858519,
+      "grad_norm": 0.9887247085571289,
+      "learning_rate": 0.00014620038570898065,
+      "loss": 0.4162,
+      "step": 4988
+    },
+    {
+      "epoch": 0.3668652106772557,
+      "grad_norm": 0.5721426010131836,
+      "learning_rate": 0.00014617926235250067,
+      "loss": 0.4187,
+      "step": 4989
+    },
+    {
+      "epoch": 0.36693874549599237,
+      "grad_norm": 1.0208877325057983,
+      "learning_rate": 0.0001461581363766842,
+      "loss": 0.3975,
+      "step": 4990
+    },
+    {
+      "epoch": 0.36701228031472904,
+      "grad_norm": 0.6323220729827881,
+      "learning_rate": 0.0001461370077827295,
+      "loss": 0.4191,
+      "step": 4991
+    },
+    {
+      "epoch": 0.3670858151334657,
+      "grad_norm": 0.8219893574714661,
+      "learning_rate": 0.00014611587657183502,
+      "loss": 0.4457,
+      "step": 4992
+    },
+    {
+      "epoch": 0.36715934995220234,
+      "grad_norm": 0.7720189690589905,
+      "learning_rate": 0.0001460947427451994,
+      "loss": 0.4311,
+      "step": 4993
+    },
+    {
+      "epoch": 0.367232884770939,
+      "grad_norm": 0.5986393094062805,
+      "learning_rate": 0.0001460736063040213,
+      "loss": 0.4336,
+      "step": 4994
+    },
+    {
+      "epoch": 0.3673064195896757,
+      "grad_norm": 0.6692777276039124,
+      "learning_rate": 0.00014605246724949962,
+      "loss": 0.4575,
+      "step": 4995
+    },
+    {
+      "epoch": 0.36737995440841237,
+      "grad_norm": 0.6144702434539795,
+      "learning_rate": 0.0001460313255828334,
+      "loss": 0.4466,
+      "step": 4996
+    },
+    {
+      "epoch": 0.36745348922714904,
+      "grad_norm": 0.7801392078399658,
+      "learning_rate": 0.0001460101813052218,
+      "loss": 0.4186,
+      "step": 4997
+    },
+    {
+      "epoch": 0.3675270240458857,
+      "grad_norm": 0.5934844613075256,
+      "learning_rate": 0.0001459890344178642,
+      "loss": 0.4155,
+      "step": 4998
+    },
+    {
+      "epoch": 0.3676005588646224,
+      "grad_norm": 0.553244948387146,
+      "learning_rate": 0.00014596788492196,
+      "loss": 0.439,
+      "step": 4999
+    },
+    {
+      "epoch": 0.36767409368335907,
+      "grad_norm": 0.7262486219406128,
+      "learning_rate": 0.0001459467328187089,
+      "loss": 0.4378,
+      "step": 5000
+    },
+    {
+      "epoch": 0.36774762850209575,
+      "grad_norm": 0.5319398045539856,
+      "learning_rate": 0.00014592557810931057,
+      "loss": 0.4174,
+      "step": 5001
+    },
+    {
+      "epoch": 0.3678211633208324,
+      "grad_norm": 0.8072534799575806,
+      "learning_rate": 0.000145904420794965,
+      "loss": 0.3854,
+      "step": 5002
+    },
+    {
+      "epoch": 0.3678946981395691,
+      "grad_norm": 1.0186197757720947,
+      "learning_rate": 0.00014588326087687224,
+      "loss": 0.4412,
+      "step": 5003
+    },
+    {
+      "epoch": 0.3679682329583058,
+      "grad_norm": 0.6909255981445312,
+      "learning_rate": 0.00014586209835623252,
+      "loss": 0.4355,
+      "step": 5004
+    },
+    {
+      "epoch": 0.36804176777704245,
+      "grad_norm": 0.7253509163856506,
+      "learning_rate": 0.00014584093323424615,
+      "loss": 0.4189,
+      "step": 5005
+    },
+    {
+      "epoch": 0.3681153025957791,
+      "grad_norm": 1.1539678573608398,
+      "learning_rate": 0.00014581976551211368,
+      "loss": 0.411,
+      "step": 5006
+    },
+    {
+      "epoch": 0.36818883741451575,
+      "grad_norm": 0.5590766072273254,
+      "learning_rate": 0.00014579859519103577,
+      "loss": 0.4277,
+      "step": 5007
+    },
+    {
+      "epoch": 0.3682623722332524,
+      "grad_norm": 0.7623084783554077,
+      "learning_rate": 0.00014577742227221317,
+      "loss": 0.4345,
+      "step": 5008
+    },
+    {
+      "epoch": 0.3683359070519891,
+      "grad_norm": 0.544282078742981,
+      "learning_rate": 0.0001457562467568469,
+      "loss": 0.4082,
+      "step": 5009
+    },
+    {
+      "epoch": 0.3684094418707258,
+      "grad_norm": 0.7086735963821411,
+      "learning_rate": 0.000145735068646138,
+      "loss": 0.4227,
+      "step": 5010
+    },
+    {
+      "epoch": 0.36848297668946245,
+      "grad_norm": 0.7281461954116821,
+      "learning_rate": 0.00014571388794128778,
+      "loss": 0.4207,
+      "step": 5011
+    },
+    {
+      "epoch": 0.3685565115081991,
+      "grad_norm": 0.5370771884918213,
+      "learning_rate": 0.00014569270464349758,
+      "loss": 0.3984,
+      "step": 5012
+    },
+    {
+      "epoch": 0.3686300463269358,
+      "grad_norm": 0.8136059045791626,
+      "learning_rate": 0.00014567151875396892,
+      "loss": 0.439,
+      "step": 5013
+    },
+    {
+      "epoch": 0.3687035811456725,
+      "grad_norm": 0.599554717540741,
+      "learning_rate": 0.00014565033027390354,
+      "loss": 0.423,
+      "step": 5014
+    },
+    {
+      "epoch": 0.36877711596440915,
+      "grad_norm": 0.7067887783050537,
+      "learning_rate": 0.00014562913920450324,
+      "loss": 0.416,
+      "step": 5015
+    },
+    {
+      "epoch": 0.36885065078314583,
+      "grad_norm": 0.6466078162193298,
+      "learning_rate": 0.00014560794554697003,
+      "loss": 0.4186,
+      "step": 5016
+    },
+    {
+      "epoch": 0.3689241856018825,
+      "grad_norm": 0.7073410749435425,
+      "learning_rate": 0.000145586749302506,
+      "loss": 0.4448,
+      "step": 5017
+    },
+    {
+      "epoch": 0.3689977204206192,
+      "grad_norm": 0.702012836933136,
+      "learning_rate": 0.00014556555047231344,
+      "loss": 0.4215,
+      "step": 5018
+    },
+    {
+      "epoch": 0.36907125523935586,
+      "grad_norm": 0.6810956001281738,
+      "learning_rate": 0.0001455443490575948,
+      "loss": 0.4215,
+      "step": 5019
+    },
+    {
+      "epoch": 0.36914479005809253,
+      "grad_norm": 0.9431997537612915,
+      "learning_rate": 0.00014552314505955255,
+      "loss": 0.4153,
+      "step": 5020
+    },
+    {
+      "epoch": 0.36921832487682915,
+      "grad_norm": 0.8495431542396545,
+      "learning_rate": 0.00014550193847938953,
+      "loss": 0.4366,
+      "step": 5021
+    },
+    {
+      "epoch": 0.3692918596955658,
+      "grad_norm": 0.6229543089866638,
+      "learning_rate": 0.00014548072931830854,
+      "loss": 0.4242,
+      "step": 5022
+    },
+    {
+      "epoch": 0.3693653945143025,
+      "grad_norm": 0.5812631249427795,
+      "learning_rate": 0.00014545951757751255,
+      "loss": 0.4418,
+      "step": 5023
+    },
+    {
+      "epoch": 0.3694389293330392,
+      "grad_norm": 0.714062511920929,
+      "learning_rate": 0.00014543830325820477,
+      "loss": 0.4182,
+      "step": 5024
+    },
+    {
+      "epoch": 0.36951246415177585,
+      "grad_norm": 0.7961524128913879,
+      "learning_rate": 0.0001454170863615885,
+      "loss": 0.4049,
+      "step": 5025
+    },
+    {
+      "epoch": 0.36958599897051253,
+      "grad_norm": 0.5368533730506897,
+      "learning_rate": 0.00014539586688886707,
+      "loss": 0.4247,
+      "step": 5026
+    },
+    {
+      "epoch": 0.3696595337892492,
+      "grad_norm": 0.8314617276191711,
+      "learning_rate": 0.00014537464484124427,
+      "loss": 0.4213,
+      "step": 5027
+    },
+    {
+      "epoch": 0.3697330686079859,
+      "grad_norm": 1.0332778692245483,
+      "learning_rate": 0.00014535342021992367,
+      "loss": 0.4013,
+      "step": 5028
+    },
+    {
+      "epoch": 0.36980660342672256,
+      "grad_norm": 0.9249008893966675,
+      "learning_rate": 0.00014533219302610923,
+      "loss": 0.4478,
+      "step": 5029
+    },
+    {
+      "epoch": 0.36988013824545923,
+      "grad_norm": 1.1740952730178833,
+      "learning_rate": 0.00014531096326100495,
+      "loss": 0.4677,
+      "step": 5030
+    },
+    {
+      "epoch": 0.3699536730641959,
+      "grad_norm": 0.6761335730552673,
+      "learning_rate": 0.00014528973092581502,
+      "loss": 0.4186,
+      "step": 5031
+    },
+    {
+      "epoch": 0.3700272078829326,
+      "grad_norm": 1.328001856803894,
+      "learning_rate": 0.00014526849602174374,
+      "loss": 0.4138,
+      "step": 5032
+    },
+    {
+      "epoch": 0.37010074270166926,
+      "grad_norm": 0.7031950354576111,
+      "learning_rate": 0.0001452472585499956,
+      "loss": 0.4246,
+      "step": 5033
+    },
+    {
+      "epoch": 0.37017427752040594,
+      "grad_norm": 1.360453486442566,
+      "learning_rate": 0.00014522601851177524,
+      "loss": 0.4521,
+      "step": 5034
+    },
+    {
+      "epoch": 0.37024781233914256,
+      "grad_norm": 0.6986666321754456,
+      "learning_rate": 0.00014520477590828734,
+      "loss": 0.4182,
+      "step": 5035
+    },
+    {
+      "epoch": 0.37032134715787923,
+      "grad_norm": 1.6479068994522095,
+      "learning_rate": 0.00014518353074073685,
+      "loss": 0.4305,
+      "step": 5036
+    },
+    {
+      "epoch": 0.3703948819766159,
+      "grad_norm": 1.1872689723968506,
+      "learning_rate": 0.00014516228301032883,
+      "loss": 0.4302,
+      "step": 5037
+    },
+    {
+      "epoch": 0.3704684167953526,
+      "grad_norm": 1.4170204401016235,
+      "learning_rate": 0.00014514103271826841,
+      "loss": 0.4302,
+      "step": 5038
+    },
+    {
+      "epoch": 0.37054195161408926,
+      "grad_norm": 1.1306731700897217,
+      "learning_rate": 0.00014511977986576103,
+      "loss": 0.4519,
+      "step": 5039
+    },
+    {
+      "epoch": 0.37061548643282594,
+      "grad_norm": 1.1314525604248047,
+      "learning_rate": 0.00014509852445401212,
+      "loss": 0.4196,
+      "step": 5040
+    },
+    {
+      "epoch": 0.3706890212515626,
+      "grad_norm": 0.8659364581108093,
+      "learning_rate": 0.00014507726648422726,
+      "loss": 0.4601,
+      "step": 5041
+    },
+    {
+      "epoch": 0.3707625560702993,
+      "grad_norm": 1.0478239059448242,
+      "learning_rate": 0.0001450560059576123,
+      "loss": 0.421,
+      "step": 5042
+    },
+    {
+      "epoch": 0.37083609088903596,
+      "grad_norm": 0.7197149395942688,
+      "learning_rate": 0.00014503474287537316,
+      "loss": 0.3973,
+      "step": 5043
+    },
+    {
+      "epoch": 0.37090962570777264,
+      "grad_norm": 1.3308261632919312,
+      "learning_rate": 0.00014501347723871585,
+      "loss": 0.4453,
+      "step": 5044
+    },
+    {
+      "epoch": 0.3709831605265093,
+      "grad_norm": 0.6827239990234375,
+      "learning_rate": 0.0001449922090488466,
+      "loss": 0.4359,
+      "step": 5045
+    },
+    {
+      "epoch": 0.371056695345246,
+      "grad_norm": 1.3383679389953613,
+      "learning_rate": 0.00014497093830697177,
+      "loss": 0.4217,
+      "step": 5046
+    },
+    {
+      "epoch": 0.37113023016398267,
+      "grad_norm": 0.7828275561332703,
+      "learning_rate": 0.00014494966501429787,
+      "loss": 0.4379,
+      "step": 5047
+    },
+    {
+      "epoch": 0.37120376498271934,
+      "grad_norm": 1.35869562625885,
+      "learning_rate": 0.00014492838917203152,
+      "loss": 0.4561,
+      "step": 5048
+    },
+    {
+      "epoch": 0.37127729980145596,
+      "grad_norm": 0.8600945472717285,
+      "learning_rate": 0.00014490711078137954,
+      "loss": 0.4171,
+      "step": 5049
+    },
+    {
+      "epoch": 0.37135083462019264,
+      "grad_norm": 1.323081612586975,
+      "learning_rate": 0.00014488582984354884,
+      "loss": 0.4464,
+      "step": 5050
+    },
+    {
+      "epoch": 0.3714243694389293,
+      "grad_norm": 1.0257236957550049,
+      "learning_rate": 0.00014486454635974648,
+      "loss": 0.4122,
+      "step": 5051
+    },
+    {
+      "epoch": 0.371497904257666,
+      "grad_norm": 1.0739845037460327,
+      "learning_rate": 0.00014484326033117973,
+      "loss": 0.4186,
+      "step": 5052
+    },
+    {
+      "epoch": 0.37157143907640267,
+      "grad_norm": 0.959698498249054,
+      "learning_rate": 0.00014482197175905593,
+      "loss": 0.4307,
+      "step": 5053
+    },
+    {
+      "epoch": 0.37164497389513934,
+      "grad_norm": 1.1760653257369995,
+      "learning_rate": 0.00014480068064458258,
+      "loss": 0.4117,
+      "step": 5054
+    },
+    {
+      "epoch": 0.371718508713876,
+      "grad_norm": 0.9323329925537109,
+      "learning_rate": 0.0001447793869889673,
+      "loss": 0.4052,
+      "step": 5055
+    },
+    {
+      "epoch": 0.3717920435326127,
+      "grad_norm": 1.069496750831604,
+      "learning_rate": 0.000144758090793418,
+      "loss": 0.4195,
+      "step": 5056
+    },
+    {
+      "epoch": 0.37186557835134937,
+      "grad_norm": 0.889728844165802,
+      "learning_rate": 0.00014473679205914252,
+      "loss": 0.4112,
+      "step": 5057
+    },
+    {
+      "epoch": 0.37193911317008604,
+      "grad_norm": 1.334961175918579,
+      "learning_rate": 0.000144715490787349,
+      "loss": 0.4516,
+      "step": 5058
+    },
+    {
+      "epoch": 0.3720126479888227,
+      "grad_norm": 1.1187524795532227,
+      "learning_rate": 0.00014469418697924562,
+      "loss": 0.4187,
+      "step": 5059
+    },
+    {
+      "epoch": 0.3720861828075594,
+      "grad_norm": 1.1671600341796875,
+      "learning_rate": 0.00014467288063604083,
+      "loss": 0.4349,
+      "step": 5060
+    },
+    {
+      "epoch": 0.37215971762629607,
+      "grad_norm": 1.2694249153137207,
+      "learning_rate": 0.0001446515717589431,
+      "loss": 0.4306,
+      "step": 5061
+    },
+    {
+      "epoch": 0.37223325244503275,
+      "grad_norm": 0.7935067415237427,
+      "learning_rate": 0.00014463026034916104,
+      "loss": 0.414,
+      "step": 5062
+    },
+    {
+      "epoch": 0.37230678726376937,
+      "grad_norm": 1.5889583826065063,
+      "learning_rate": 0.00014460894640790358,
+      "loss": 0.4344,
+      "step": 5063
+    },
+    {
+      "epoch": 0.37238032208250604,
+      "grad_norm": 0.646699845790863,
+      "learning_rate": 0.00014458762993637957,
+      "loss": 0.4242,
+      "step": 5064
+    },
+    {
+      "epoch": 0.3724538569012427,
+      "grad_norm": 1.3657160997390747,
+      "learning_rate": 0.00014456631093579816,
+      "loss": 0.4246,
+      "step": 5065
+    },
+    {
+      "epoch": 0.3725273917199794,
+      "grad_norm": 0.7521275877952576,
+      "learning_rate": 0.00014454498940736854,
+      "loss": 0.416,
+      "step": 5066
+    },
+    {
+      "epoch": 0.37260092653871607,
+      "grad_norm": 1.2081831693649292,
+      "learning_rate": 0.0001445236653523001,
+      "loss": 0.4166,
+      "step": 5067
+    },
+    {
+      "epoch": 0.37267446135745275,
+      "grad_norm": 0.7633953094482422,
+      "learning_rate": 0.0001445023387718024,
+      "loss": 0.4097,
+      "step": 5068
+    },
+    {
+      "epoch": 0.3727479961761894,
+      "grad_norm": 1.230869174003601,
+      "learning_rate": 0.00014448100966708507,
+      "loss": 0.3932,
+      "step": 5069
+    },
+    {
+      "epoch": 0.3728215309949261,
+      "grad_norm": 1.143875002861023,
+      "learning_rate": 0.00014445967803935794,
+      "loss": 0.396,
+      "step": 5070
+    },
+    {
+      "epoch": 0.3728950658136628,
+      "grad_norm": 0.6804710626602173,
+      "learning_rate": 0.00014443834388983093,
+      "loss": 0.4129,
+      "step": 5071
+    },
+    {
+      "epoch": 0.37296860063239945,
+      "grad_norm": 1.1520339250564575,
+      "learning_rate": 0.00014441700721971415,
+      "loss": 0.4316,
+      "step": 5072
+    },
+    {
+      "epoch": 0.3730421354511361,
+      "grad_norm": 0.5659427046775818,
+      "learning_rate": 0.00014439566803021786,
+      "loss": 0.4093,
+      "step": 5073
+    },
+    {
+      "epoch": 0.3731156702698728,
+      "grad_norm": 1.5667707920074463,
+      "learning_rate": 0.00014437432632255243,
+      "loss": 0.4265,
+      "step": 5074
+    },
+    {
+      "epoch": 0.3731892050886095,
+      "grad_norm": 0.6817637085914612,
+      "learning_rate": 0.0001443529820979284,
+      "loss": 0.4139,
+      "step": 5075
+    },
+    {
+      "epoch": 0.37326273990734615,
+      "grad_norm": 1.8207740783691406,
+      "learning_rate": 0.00014433163535755637,
+      "loss": 0.4374,
+      "step": 5076
+    },
+    {
+      "epoch": 0.3733362747260828,
+      "grad_norm": 1.0014498233795166,
+      "learning_rate": 0.0001443102861026472,
+      "loss": 0.4169,
+      "step": 5077
+    },
+    {
+      "epoch": 0.37340980954481945,
+      "grad_norm": 1.5481880903244019,
+      "learning_rate": 0.00014428893433441188,
+      "loss": 0.4218,
+      "step": 5078
+    },
+    {
+      "epoch": 0.3734833443635561,
+      "grad_norm": 1.0335261821746826,
+      "learning_rate": 0.00014426758005406143,
+      "loss": 0.4211,
+      "step": 5079
+    },
+    {
+      "epoch": 0.3735568791822928,
+      "grad_norm": 1.077883243560791,
+      "learning_rate": 0.00014424622326280717,
+      "loss": 0.4171,
+      "step": 5080
+    },
+    {
+      "epoch": 0.3736304140010295,
+      "grad_norm": 1.0826306343078613,
+      "learning_rate": 0.00014422486396186042,
+      "loss": 0.4183,
+      "step": 5081
+    },
+    {
+      "epoch": 0.37370394881976615,
+      "grad_norm": 0.6614946126937866,
+      "learning_rate": 0.00014420350215243265,
+      "loss": 0.419,
+      "step": 5082
+    },
+    {
+      "epoch": 0.3737774836385028,
+      "grad_norm": 1.051492691040039,
+      "learning_rate": 0.00014418213783573563,
+      "loss": 0.4299,
+      "step": 5083
+    },
+    {
+      "epoch": 0.3738510184572395,
+      "grad_norm": 0.6907543540000916,
+      "learning_rate": 0.00014416077101298114,
+      "loss": 0.4503,
+      "step": 5084
+    },
+    {
+      "epoch": 0.3739245532759762,
+      "grad_norm": 1.0566291809082031,
+      "learning_rate": 0.00014413940168538108,
+      "loss": 0.4018,
+      "step": 5085
+    },
+    {
+      "epoch": 0.37399808809471285,
+      "grad_norm": 1.0763617753982544,
+      "learning_rate": 0.00014411802985414763,
+      "loss": 0.4354,
+      "step": 5086
+    },
+    {
+      "epoch": 0.37407162291344953,
+      "grad_norm": 0.6930245757102966,
+      "learning_rate": 0.00014409665552049292,
+      "loss": 0.4403,
+      "step": 5087
+    },
+    {
+      "epoch": 0.3741451577321862,
+      "grad_norm": 0.9760132431983948,
+      "learning_rate": 0.0001440752786856294,
+      "loss": 0.4213,
+      "step": 5088
+    },
+    {
+      "epoch": 0.3742186925509229,
+      "grad_norm": 0.8805245161056519,
+      "learning_rate": 0.00014405389935076953,
+      "loss": 0.3885,
+      "step": 5089
+    },
+    {
+      "epoch": 0.37429222736965956,
+      "grad_norm": 0.7334687113761902,
+      "learning_rate": 0.000144032517517126,
+      "loss": 0.4182,
+      "step": 5090
+    },
+    {
+      "epoch": 0.3743657621883962,
+      "grad_norm": 0.5224865674972534,
+      "learning_rate": 0.00014401113318591161,
+      "loss": 0.3972,
+      "step": 5091
+    },
+    {
+      "epoch": 0.37443929700713285,
+      "grad_norm": 1.0553538799285889,
+      "learning_rate": 0.00014398974635833933,
+      "loss": 0.4256,
+      "step": 5092
+    },
+    {
+      "epoch": 0.37451283182586953,
+      "grad_norm": 0.8262841105461121,
+      "learning_rate": 0.0001439683570356222,
+      "loss": 0.3996,
+      "step": 5093
+    },
+    {
+      "epoch": 0.3745863666446062,
+      "grad_norm": 0.640510618686676,
+      "learning_rate": 0.00014394696521897343,
+      "loss": 0.4101,
+      "step": 5094
+    },
+    {
+      "epoch": 0.3746599014633429,
+      "grad_norm": 0.8302705883979797,
+      "learning_rate": 0.00014392557090960642,
+      "loss": 0.3983,
+      "step": 5095
+    },
+    {
+      "epoch": 0.37473343628207956,
+      "grad_norm": 0.589830756187439,
+      "learning_rate": 0.00014390417410873466,
+      "loss": 0.4049,
+      "step": 5096
+    },
+    {
+      "epoch": 0.37480697110081623,
+      "grad_norm": 0.9273269176483154,
+      "learning_rate": 0.00014388277481757183,
+      "loss": 0.4062,
+      "step": 5097
+    },
+    {
+      "epoch": 0.3748805059195529,
+      "grad_norm": 1.2512446641921997,
+      "learning_rate": 0.00014386137303733172,
+      "loss": 0.447,
+      "step": 5098
+    },
+    {
+      "epoch": 0.3749540407382896,
+      "grad_norm": 0.8746530413627625,
+      "learning_rate": 0.00014383996876922824,
+      "loss": 0.4252,
+      "step": 5099
+    },
+    {
+      "epoch": 0.37502757555702626,
+      "grad_norm": 0.7829744219779968,
+      "learning_rate": 0.00014381856201447544,
+      "loss": 0.3892,
+      "step": 5100
+    },
+    {
+      "epoch": 0.37510111037576294,
+      "grad_norm": 0.9945876598358154,
+      "learning_rate": 0.00014379715277428756,
+      "loss": 0.4195,
+      "step": 5101
+    },
+    {
+      "epoch": 0.3751746451944996,
+      "grad_norm": 0.7568415403366089,
+      "learning_rate": 0.00014377574104987894,
+      "loss": 0.3863,
+      "step": 5102
+    },
+    {
+      "epoch": 0.3752481800132363,
+      "grad_norm": 1.143548607826233,
+      "learning_rate": 0.00014375432684246415,
+      "loss": 0.4269,
+      "step": 5103
+    },
+    {
+      "epoch": 0.37532171483197296,
+      "grad_norm": 0.6197658181190491,
+      "learning_rate": 0.0001437329101532577,
+      "loss": 0.4148,
+      "step": 5104
+    },
+    {
+      "epoch": 0.3753952496507096,
+      "grad_norm": 1.1022119522094727,
+      "learning_rate": 0.00014371149098347448,
+      "loss": 0.4144,
+      "step": 5105
+    },
+    {
+      "epoch": 0.37546878446944626,
+      "grad_norm": 0.8405001759529114,
+      "learning_rate": 0.00014369006933432935,
+      "loss": 0.3927,
+      "step": 5106
+    },
+    {
+      "epoch": 0.37554231928818294,
+      "grad_norm": 0.6118815541267395,
+      "learning_rate": 0.00014366864520703735,
+      "loss": 0.4482,
+      "step": 5107
+    },
+    {
+      "epoch": 0.3756158541069196,
+      "grad_norm": 1.1789498329162598,
+      "learning_rate": 0.00014364721860281372,
+      "loss": 0.4209,
+      "step": 5108
+    },
+    {
+      "epoch": 0.3756893889256563,
+      "grad_norm": 0.9167799949645996,
+      "learning_rate": 0.00014362578952287382,
+      "loss": 0.427,
+      "step": 5109
+    },
+    {
+      "epoch": 0.37576292374439296,
+      "grad_norm": 0.7819786071777344,
+      "learning_rate": 0.00014360435796843306,
+      "loss": 0.4306,
+      "step": 5110
+    },
+    {
+      "epoch": 0.37583645856312964,
+      "grad_norm": 1.1003409624099731,
+      "learning_rate": 0.00014358292394070713,
+      "loss": 0.4115,
+      "step": 5111
+    },
+    {
+      "epoch": 0.3759099933818663,
+      "grad_norm": 0.7882094979286194,
+      "learning_rate": 0.00014356148744091175,
+      "loss": 0.4235,
+      "step": 5112
+    },
+    {
+      "epoch": 0.375983528200603,
+      "grad_norm": 0.8280505537986755,
+      "learning_rate": 0.0001435400484702628,
+      "loss": 0.412,
+      "step": 5113
+    },
+    {
+      "epoch": 0.37605706301933967,
+      "grad_norm": 1.3591341972351074,
+      "learning_rate": 0.00014351860702997636,
+      "loss": 0.4055,
+      "step": 5114
+    },
+    {
+      "epoch": 0.37613059783807634,
+      "grad_norm": 0.7187333106994629,
+      "learning_rate": 0.0001434971631212686,
+      "loss": 0.3979,
+      "step": 5115
+    },
+    {
+      "epoch": 0.376204132656813,
+      "grad_norm": 1.196617841720581,
+      "learning_rate": 0.0001434757167453559,
+      "loss": 0.4509,
+      "step": 5116
+    },
+    {
+      "epoch": 0.3762776674755497,
+      "grad_norm": 0.8659620881080627,
+      "learning_rate": 0.00014345426790345463,
+      "loss": 0.3999,
+      "step": 5117
+    },
+    {
+      "epoch": 0.37635120229428637,
+      "grad_norm": 0.9979854822158813,
+      "learning_rate": 0.0001434328165967814,
+      "loss": 0.4312,
+      "step": 5118
+    },
+    {
+      "epoch": 0.376424737113023,
+      "grad_norm": 1.2843397855758667,
+      "learning_rate": 0.00014341136282655298,
+      "loss": 0.4304,
+      "step": 5119
+    },
+    {
+      "epoch": 0.37649827193175966,
+      "grad_norm": 0.704994797706604,
+      "learning_rate": 0.00014338990659398625,
+      "loss": 0.4566,
+      "step": 5120
+    },
+    {
+      "epoch": 0.37657180675049634,
+      "grad_norm": 1.1798436641693115,
+      "learning_rate": 0.00014336844790029826,
+      "loss": 0.4156,
+      "step": 5121
+    },
+    {
+      "epoch": 0.376645341569233,
+      "grad_norm": 0.6163409948348999,
+      "learning_rate": 0.00014334698674670613,
+      "loss": 0.416,
+      "step": 5122
+    },
+    {
+      "epoch": 0.3767188763879697,
+      "grad_norm": 0.838018536567688,
+      "learning_rate": 0.00014332552313442714,
+      "loss": 0.405,
+      "step": 5123
+    },
+    {
+      "epoch": 0.37679241120670637,
+      "grad_norm": 0.9430814385414124,
+      "learning_rate": 0.00014330405706467877,
+      "loss": 0.3966,
+      "step": 5124
+    },
+    {
+      "epoch": 0.37686594602544304,
+      "grad_norm": 0.552925705909729,
+      "learning_rate": 0.00014328258853867854,
+      "loss": 0.4493,
+      "step": 5125
+    },
+    {
+      "epoch": 0.3769394808441797,
+      "grad_norm": 1.133847951889038,
+      "learning_rate": 0.00014326111755764428,
+      "loss": 0.3918,
+      "step": 5126
+    },
+    {
+      "epoch": 0.3770130156629164,
+      "grad_norm": 0.6417760848999023,
+      "learning_rate": 0.00014323964412279376,
+      "loss": 0.4065,
+      "step": 5127
+    },
+    {
+      "epoch": 0.37708655048165307,
+      "grad_norm": 1.0035101175308228,
+      "learning_rate": 0.00014321816823534498,
+      "loss": 0.4199,
+      "step": 5128
+    },
+    {
+      "epoch": 0.37716008530038975,
+      "grad_norm": 1.1370567083358765,
+      "learning_rate": 0.00014319668989651606,
+      "loss": 0.4258,
+      "step": 5129
+    },
+    {
+      "epoch": 0.3772336201191264,
+      "grad_norm": 0.5499017238616943,
+      "learning_rate": 0.00014317520910752535,
+      "loss": 0.394,
+      "step": 5130
+    },
+    {
+      "epoch": 0.3773071549378631,
+      "grad_norm": 1.039860486984253,
+      "learning_rate": 0.00014315372586959116,
+      "loss": 0.3945,
+      "step": 5131
+    },
+    {
+      "epoch": 0.3773806897565998,
+      "grad_norm": 1.1369397640228271,
+      "learning_rate": 0.00014313224018393218,
+      "loss": 0.3984,
+      "step": 5132
+    },
+    {
+      "epoch": 0.3774542245753364,
+      "grad_norm": 0.6970899105072021,
+      "learning_rate": 0.00014311075205176695,
+      "loss": 0.4127,
+      "step": 5133
+    },
+    {
+      "epoch": 0.37752775939407307,
+      "grad_norm": 1.8273850679397583,
+      "learning_rate": 0.0001430892614743144,
+      "loss": 0.4097,
+      "step": 5134
+    },
+    {
+      "epoch": 0.37760129421280975,
+      "grad_norm": 0.9839183688163757,
+      "learning_rate": 0.00014306776845279343,
+      "loss": 0.3922,
+      "step": 5135
+    },
+    {
+      "epoch": 0.3776748290315464,
+      "grad_norm": 2.2881877422332764,
+      "learning_rate": 0.00014304627298842318,
+      "loss": 0.3966,
+      "step": 5136
+    },
+    {
+      "epoch": 0.3777483638502831,
+      "grad_norm": 2.1768269538879395,
+      "learning_rate": 0.00014302477508242294,
+      "loss": 0.4172,
+      "step": 5137
+    },
+    {
+      "epoch": 0.3778218986690198,
+      "grad_norm": 1.2997361421585083,
+      "learning_rate": 0.00014300327473601202,
+      "loss": 0.4329,
+      "step": 5138
+    },
+    {
+      "epoch": 0.37789543348775645,
+      "grad_norm": 1.075760006904602,
+      "learning_rate": 0.00014298177195041,
+      "loss": 0.4093,
+      "step": 5139
+    },
+    {
+      "epoch": 0.3779689683064931,
+      "grad_norm": 1.5534027814865112,
+      "learning_rate": 0.0001429602667268365,
+      "loss": 0.4455,
+      "step": 5140
+    },
+    {
+      "epoch": 0.3780425031252298,
+      "grad_norm": 1.1472992897033691,
+      "learning_rate": 0.0001429387590665113,
+      "loss": 0.3994,
+      "step": 5141
+    },
+    {
+      "epoch": 0.3781160379439665,
+      "grad_norm": 2.172029733657837,
+      "learning_rate": 0.00014291724897065442,
+      "loss": 0.4294,
+      "step": 5142
+    },
+    {
+      "epoch": 0.37818957276270315,
+      "grad_norm": 1.9710450172424316,
+      "learning_rate": 0.00014289573644048586,
+      "loss": 0.4127,
+      "step": 5143
+    },
+    {
+      "epoch": 0.37826310758143983,
+      "grad_norm": 1.194344162940979,
+      "learning_rate": 0.00014287422147722588,
+      "loss": 0.3963,
+      "step": 5144
+    },
+    {
+      "epoch": 0.3783366424001765,
+      "grad_norm": 1.306895136833191,
+      "learning_rate": 0.00014285270408209483,
+      "loss": 0.4407,
+      "step": 5145
+    },
+    {
+      "epoch": 0.3784101772189132,
+      "grad_norm": 1.3173253536224365,
+      "learning_rate": 0.00014283118425631312,
+      "loss": 0.4117,
+      "step": 5146
+    },
+    {
+      "epoch": 0.3784837120376498,
+      "grad_norm": 0.7849929928779602,
+      "learning_rate": 0.00014280966200110146,
+      "loss": 0.415,
+      "step": 5147
+    },
+    {
+      "epoch": 0.3785572468563865,
+      "grad_norm": 1.5001903772354126,
+      "learning_rate": 0.00014278813731768058,
+      "loss": 0.4048,
+      "step": 5148
+    },
+    {
+      "epoch": 0.37863078167512315,
+      "grad_norm": 0.6936997771263123,
+      "learning_rate": 0.0001427666102072714,
+      "loss": 0.3937,
+      "step": 5149
+    },
+    {
+      "epoch": 0.3787043164938598,
+      "grad_norm": 1.5875952243804932,
+      "learning_rate": 0.00014274508067109502,
+      "loss": 0.4288,
+      "step": 5150
+    },
+    {
+      "epoch": 0.3787778513125965,
+      "grad_norm": 0.8814023733139038,
+      "learning_rate": 0.0001427235487103725,
+      "loss": 0.3946,
+      "step": 5151
+    },
+    {
+      "epoch": 0.3788513861313332,
+      "grad_norm": 1.8299263715744019,
+      "learning_rate": 0.0001427020143263252,
+      "loss": 0.4393,
+      "step": 5152
+    },
+    {
+      "epoch": 0.37892492095006985,
+      "grad_norm": 1.4380383491516113,
+      "learning_rate": 0.00014268047752017462,
+      "loss": 0.4166,
+      "step": 5153
+    },
+    {
+      "epoch": 0.37899845576880653,
+      "grad_norm": 1.6749608516693115,
+      "learning_rate": 0.00014265893829314225,
+      "loss": 0.3923,
+      "step": 5154
+    },
+    {
+      "epoch": 0.3790719905875432,
+      "grad_norm": 1.4807780981063843,
+      "learning_rate": 0.00014263739664644993,
+      "loss": 0.3977,
+      "step": 5155
+    },
+    {
+      "epoch": 0.3791455254062799,
+      "grad_norm": 1.4099818468093872,
+      "learning_rate": 0.00014261585258131947,
+      "loss": 0.3951,
+      "step": 5156
+    },
+    {
+      "epoch": 0.37921906022501656,
+      "grad_norm": 1.1192595958709717,
+      "learning_rate": 0.00014259430609897287,
+      "loss": 0.3779,
+      "step": 5157
+    },
+    {
+      "epoch": 0.37929259504375323,
+      "grad_norm": 1.6431769132614136,
+      "learning_rate": 0.0001425727572006323,
+      "loss": 0.3828,
+      "step": 5158
+    },
+    {
+      "epoch": 0.3793661298624899,
+      "grad_norm": 1.665228009223938,
+      "learning_rate": 0.00014255120588751996,
+      "loss": 0.4151,
+      "step": 5159
+    },
+    {
+      "epoch": 0.3794396646812266,
+      "grad_norm": 1.6385974884033203,
+      "learning_rate": 0.00014252965216085835,
+      "loss": 0.4496,
+      "step": 5160
+    },
+    {
+      "epoch": 0.3795131994999632,
+      "grad_norm": 1.5212091207504272,
+      "learning_rate": 0.00014250809602186996,
+      "loss": 0.381,
+      "step": 5161
+    },
+    {
+      "epoch": 0.3795867343186999,
+      "grad_norm": 1.2465786933898926,
+      "learning_rate": 0.00014248653747177755,
+      "loss": 0.3885,
+      "step": 5162
+    },
+    {
+      "epoch": 0.37966026913743656,
+      "grad_norm": 1.153539776802063,
+      "learning_rate": 0.00014246497651180387,
+      "loss": 0.3892,
+      "step": 5163
+    },
+    {
+      "epoch": 0.37973380395617323,
+      "grad_norm": 1.6663804054260254,
+      "learning_rate": 0.0001424434131431719,
+      "loss": 0.4131,
+      "step": 5164
+    },
+    {
+      "epoch": 0.3798073387749099,
+      "grad_norm": 1.4392728805541992,
+      "learning_rate": 0.00014242184736710476,
+      "loss": 0.4022,
+      "step": 5165
+    },
+    {
+      "epoch": 0.3798808735936466,
+      "grad_norm": 1.4244643449783325,
+      "learning_rate": 0.00014240027918482567,
+      "loss": 0.4174,
+      "step": 5166
+    },
+    {
+      "epoch": 0.37995440841238326,
+      "grad_norm": 1.3106597661972046,
+      "learning_rate": 0.000142378708597558,
+      "loss": 0.3986,
+      "step": 5167
+    },
+    {
+      "epoch": 0.38002794323111994,
+      "grad_norm": 1.7027530670166016,
+      "learning_rate": 0.00014235713560652524,
+      "loss": 0.4094,
+      "step": 5168
+    },
+    {
+      "epoch": 0.3801014780498566,
+      "grad_norm": 1.5409117937088013,
+      "learning_rate": 0.00014233556021295103,
+      "loss": 0.3788,
+      "step": 5169
+    },
+    {
+      "epoch": 0.3801750128685933,
+      "grad_norm": 1.4266774654388428,
+      "learning_rate": 0.0001423139824180592,
+      "loss": 0.399,
+      "step": 5170
+    },
+    {
+      "epoch": 0.38024854768732996,
+      "grad_norm": 1.2863729000091553,
+      "learning_rate": 0.0001422924022230736,
+      "loss": 0.4003,
+      "step": 5171
+    },
+    {
+      "epoch": 0.38032208250606664,
+      "grad_norm": 1.3909480571746826,
+      "learning_rate": 0.00014227081962921834,
+      "loss": 0.4003,
+      "step": 5172
+    },
+    {
+      "epoch": 0.3803956173248033,
+      "grad_norm": 1.4347032308578491,
+      "learning_rate": 0.00014224923463771757,
+      "loss": 0.4351,
+      "step": 5173
+    },
+    {
+      "epoch": 0.38046915214354,
+      "grad_norm": 1.4599568843841553,
+      "learning_rate": 0.00014222764724979562,
+      "loss": 0.4223,
+      "step": 5174
+    },
+    {
+      "epoch": 0.3805426869622766,
+      "grad_norm": 1.4259036779403687,
+      "learning_rate": 0.00014220605746667698,
+      "loss": 0.4089,
+      "step": 5175
+    },
+    {
+      "epoch": 0.3806162217810133,
+      "grad_norm": 1.5393867492675781,
+      "learning_rate": 0.0001421844652895862,
+      "loss": 0.4134,
+      "step": 5176
+    },
+    {
+      "epoch": 0.38068975659974996,
+      "grad_norm": 1.1938477754592896,
+      "learning_rate": 0.000142162870719748,
+      "loss": 0.4065,
+      "step": 5177
+    },
+    {
+      "epoch": 0.38076329141848664,
+      "grad_norm": 2.1020429134368896,
+      "learning_rate": 0.0001421412737583873,
+      "loss": 0.4238,
+      "step": 5178
+    },
+    {
+      "epoch": 0.3808368262372233,
+      "grad_norm": 1.8504172563552856,
+      "learning_rate": 0.0001421196744067291,
+      "loss": 0.4188,
+      "step": 5179
+    },
+    {
+      "epoch": 0.38091036105596,
+      "grad_norm": 1.016095519065857,
+      "learning_rate": 0.0001420980726659985,
+      "loss": 0.41,
+      "step": 5180
+    },
+    {
+      "epoch": 0.38098389587469667,
+      "grad_norm": 1.1699246168136597,
+      "learning_rate": 0.0001420764685374208,
+      "loss": 0.396,
+      "step": 5181
+    },
+    {
+      "epoch": 0.38105743069343334,
+      "grad_norm": 1.5215619802474976,
+      "learning_rate": 0.0001420548620222214,
+      "loss": 0.4198,
+      "step": 5182
+    },
+    {
+      "epoch": 0.38113096551217,
+      "grad_norm": 1.2906463146209717,
+      "learning_rate": 0.0001420332531216258,
+      "loss": 0.4047,
+      "step": 5183
+    },
+    {
+      "epoch": 0.3812045003309067,
+      "grad_norm": 1.479929804801941,
+      "learning_rate": 0.00014201164183685972,
+      "loss": 0.4265,
+      "step": 5184
+    },
+    {
+      "epoch": 0.38127803514964337,
+      "grad_norm": 1.3991458415985107,
+      "learning_rate": 0.00014199002816914905,
+      "loss": 0.3878,
+      "step": 5185
+    },
+    {
+      "epoch": 0.38135156996838004,
+      "grad_norm": 1.5103811025619507,
+      "learning_rate": 0.00014196841211971966,
+      "loss": 0.4154,
+      "step": 5186
+    },
+    {
+      "epoch": 0.3814251047871167,
+      "grad_norm": 1.175586223602295,
+      "learning_rate": 0.00014194679368979759,
+      "loss": 0.4267,
+      "step": 5187
+    },
+    {
+      "epoch": 0.3814986396058534,
+      "grad_norm": 1.7926621437072754,
+      "learning_rate": 0.00014192517288060915,
+      "loss": 0.4396,
+      "step": 5188
+    },
+    {
+      "epoch": 0.38157217442459,
+      "grad_norm": 1.6002286672592163,
+      "learning_rate": 0.00014190354969338063,
+      "loss": 0.3995,
+      "step": 5189
+    },
+    {
+      "epoch": 0.3816457092433267,
+      "grad_norm": 1.2763903141021729,
+      "learning_rate": 0.0001418819241293386,
+      "loss": 0.4264,
+      "step": 5190
+    },
+    {
+      "epoch": 0.38171924406206337,
+      "grad_norm": 1.3555335998535156,
+      "learning_rate": 0.0001418602961897096,
+      "loss": 0.4178,
+      "step": 5191
+    },
+    {
+      "epoch": 0.38179277888080004,
+      "grad_norm": 1.3354471921920776,
+      "learning_rate": 0.0001418386658757204,
+      "loss": 0.4014,
+      "step": 5192
+    },
+    {
+      "epoch": 0.3818663136995367,
+      "grad_norm": 1.2262763977050781,
+      "learning_rate": 0.00014181703318859795,
+      "loss": 0.4302,
+      "step": 5193
+    },
+    {
+      "epoch": 0.3819398485182734,
+      "grad_norm": 1.7969937324523926,
+      "learning_rate": 0.00014179539812956925,
+      "loss": 0.4131,
+      "step": 5194
+    },
+    {
+      "epoch": 0.38201338333701007,
+      "grad_norm": 1.2891439199447632,
+      "learning_rate": 0.00014177376069986143,
+      "loss": 0.4361,
+      "step": 5195
+    },
+    {
+      "epoch": 0.38208691815574675,
+      "grad_norm": 1.6413480043411255,
+      "learning_rate": 0.00014175212090070184,
+      "loss": 0.4422,
+      "step": 5196
+    },
+    {
+      "epoch": 0.3821604529744834,
+      "grad_norm": 1.2461574077606201,
+      "learning_rate": 0.0001417304787333179,
+      "loss": 0.4098,
+      "step": 5197
+    },
+    {
+      "epoch": 0.3822339877932201,
+      "grad_norm": 1.5322705507278442,
+      "learning_rate": 0.00014170883419893713,
+      "loss": 0.4524,
+      "step": 5198
+    },
+    {
+      "epoch": 0.3823075226119568,
+      "grad_norm": 1.5310248136520386,
+      "learning_rate": 0.0001416871872987873,
+      "loss": 0.452,
+      "step": 5199
+    },
+    {
+      "epoch": 0.38238105743069345,
+      "grad_norm": 1.767686128616333,
+      "learning_rate": 0.0001416655380340962,
+      "loss": 0.4161,
+      "step": 5200
+    },
+    {
+      "epoch": 0.3824545922494301,
+      "grad_norm": 1.497470498085022,
+      "learning_rate": 0.00014164388640609184,
+      "loss": 0.4364,
+      "step": 5201
+    },
+    {
+      "epoch": 0.3825281270681668,
+      "grad_norm": 1.4436817169189453,
+      "learning_rate": 0.00014162223241600224,
+      "loss": 0.4157,
+      "step": 5202
+    },
+    {
+      "epoch": 0.3826016618869034,
+      "grad_norm": 1.2799876928329468,
+      "learning_rate": 0.00014160057606505574,
+      "loss": 0.4596,
+      "step": 5203
+    },
+    {
+      "epoch": 0.3826751967056401,
+      "grad_norm": 1.7509173154830933,
+      "learning_rate": 0.00014157891735448063,
+      "loss": 0.4027,
+      "step": 5204
+    },
+    {
+      "epoch": 0.3827487315243768,
+      "grad_norm": 2.109372854232788,
+      "learning_rate": 0.00014155725628550546,
+      "loss": 0.4253,
+      "step": 5205
+    },
+    {
+      "epoch": 0.38282226634311345,
+      "grad_norm": 1.2456899881362915,
+      "learning_rate": 0.00014153559285935886,
+      "loss": 0.4228,
+      "step": 5206
+    },
+    {
+      "epoch": 0.3828958011618501,
+      "grad_norm": 1.693075180053711,
+      "learning_rate": 0.0001415139270772696,
+      "loss": 0.3777,
+      "step": 5207
+    },
+    {
+      "epoch": 0.3829693359805868,
+      "grad_norm": 1.3636173009872437,
+      "learning_rate": 0.00014149225894046657,
+      "loss": 0.448,
+      "step": 5208
+    },
+    {
+      "epoch": 0.3830428707993235,
+      "grad_norm": 1.3557746410369873,
+      "learning_rate": 0.00014147058845017884,
+      "loss": 0.4463,
+      "step": 5209
+    },
+    {
+      "epoch": 0.38311640561806015,
+      "grad_norm": 1.630946397781372,
+      "learning_rate": 0.00014144891560763555,
+      "loss": 0.4386,
+      "step": 5210
+    },
+    {
+      "epoch": 0.3831899404367968,
+      "grad_norm": 1.392741084098816,
+      "learning_rate": 0.00014142724041406606,
+      "loss": 0.4334,
+      "step": 5211
+    },
+    {
+      "epoch": 0.3832634752555335,
+      "grad_norm": 1.6724398136138916,
+      "learning_rate": 0.0001414055628706997,
+      "loss": 0.4211,
+      "step": 5212
+    },
+    {
+      "epoch": 0.3833370100742702,
+      "grad_norm": 1.5179258584976196,
+      "learning_rate": 0.0001413838829787662,
+      "loss": 0.4052,
+      "step": 5213
+    },
+    {
+      "epoch": 0.38341054489300685,
+      "grad_norm": 1.3048725128173828,
+      "learning_rate": 0.00014136220073949514,
+      "loss": 0.4003,
+      "step": 5214
+    },
+    {
+      "epoch": 0.38348407971174353,
+      "grad_norm": 1.1388479471206665,
+      "learning_rate": 0.0001413405161541164,
+      "loss": 0.4051,
+      "step": 5215
+    },
+    {
+      "epoch": 0.3835576145304802,
+      "grad_norm": 1.2411487102508545,
+      "learning_rate": 0.00014131882922386,
+      "loss": 0.4092,
+      "step": 5216
+    },
+    {
+      "epoch": 0.3836311493492168,
+      "grad_norm": 1.0915395021438599,
+      "learning_rate": 0.00014129713994995595,
+      "loss": 0.4244,
+      "step": 5217
+    },
+    {
+      "epoch": 0.3837046841679535,
+      "grad_norm": 1.128366231918335,
+      "learning_rate": 0.00014127544833363455,
+      "loss": 0.4148,
+      "step": 5218
+    },
+    {
+      "epoch": 0.3837782189866902,
+      "grad_norm": 1.0866105556488037,
+      "learning_rate": 0.00014125375437612617,
+      "loss": 0.3995,
+      "step": 5219
+    },
+    {
+      "epoch": 0.38385175380542685,
+      "grad_norm": 1.1734848022460938,
+      "learning_rate": 0.0001412320580786613,
+      "loss": 0.4231,
+      "step": 5220
+    },
+    {
+      "epoch": 0.38392528862416353,
+      "grad_norm": 1.257815957069397,
+      "learning_rate": 0.00014121035944247062,
+      "loss": 0.4312,
+      "step": 5221
+    },
+    {
+      "epoch": 0.3839988234429002,
+      "grad_norm": 0.8484708070755005,
+      "learning_rate": 0.0001411886584687848,
+      "loss": 0.4216,
+      "step": 5222
+    },
+    {
+      "epoch": 0.3840723582616369,
+      "grad_norm": 1.385899305343628,
+      "learning_rate": 0.00014116695515883483,
+      "loss": 0.4113,
+      "step": 5223
+    },
+    {
+      "epoch": 0.38414589308037356,
+      "grad_norm": 0.7559594511985779,
+      "learning_rate": 0.0001411452495138517,
+      "loss": 0.4076,
+      "step": 5224
+    },
+    {
+      "epoch": 0.38421942789911023,
+      "grad_norm": 1.4380513429641724,
+      "learning_rate": 0.00014112354153506662,
+      "loss": 0.3865,
+      "step": 5225
+    },
+    {
+      "epoch": 0.3842929627178469,
+      "grad_norm": 0.7764405608177185,
+      "learning_rate": 0.00014110183122371088,
+      "loss": 0.4281,
+      "step": 5226
+    },
+    {
+      "epoch": 0.3843664975365836,
+      "grad_norm": 1.588934302330017,
+      "learning_rate": 0.00014108011858101588,
+      "loss": 0.4236,
+      "step": 5227
+    },
+    {
+      "epoch": 0.38444003235532026,
+      "grad_norm": 0.811995267868042,
+      "learning_rate": 0.00014105840360821318,
+      "loss": 0.4694,
+      "step": 5228
+    },
+    {
+      "epoch": 0.38451356717405694,
+      "grad_norm": 1.7981704473495483,
+      "learning_rate": 0.00014103668630653453,
+      "loss": 0.4027,
+      "step": 5229
+    },
+    {
+      "epoch": 0.3845871019927936,
+      "grad_norm": 1.0378526449203491,
+      "learning_rate": 0.00014101496667721168,
+      "loss": 0.4253,
+      "step": 5230
+    },
+    {
+      "epoch": 0.38466063681153023,
+      "grad_norm": 2.0561628341674805,
+      "learning_rate": 0.00014099324472147665,
+      "loss": 0.4119,
+      "step": 5231
+    },
+    {
+      "epoch": 0.3847341716302669,
+      "grad_norm": 1.9147368669509888,
+      "learning_rate": 0.00014097152044056157,
+      "loss": 0.4434,
+      "step": 5232
+    },
+    {
+      "epoch": 0.3848077064490036,
+      "grad_norm": 0.9833489656448364,
+      "learning_rate": 0.00014094979383569856,
+      "loss": 0.4246,
+      "step": 5233
+    },
+    {
+      "epoch": 0.38488124126774026,
+      "grad_norm": 1.3836669921875,
+      "learning_rate": 0.00014092806490812002,
+      "loss": 0.4003,
+      "step": 5234
+    },
+    {
+      "epoch": 0.38495477608647694,
+      "grad_norm": 1.0445523262023926,
+      "learning_rate": 0.0001409063336590585,
+      "loss": 0.4339,
+      "step": 5235
+    },
+    {
+      "epoch": 0.3850283109052136,
+      "grad_norm": 1.436002492904663,
+      "learning_rate": 0.00014088460008974648,
+      "loss": 0.4031,
+      "step": 5236
+    },
+    {
+      "epoch": 0.3851018457239503,
+      "grad_norm": 1.4181071519851685,
+      "learning_rate": 0.00014086286420141687,
+      "loss": 0.4215,
+      "step": 5237
+    },
+    {
+      "epoch": 0.38517538054268696,
+      "grad_norm": 0.9434112906455994,
+      "learning_rate": 0.00014084112599530246,
+      "loss": 0.428,
+      "step": 5238
+    },
+    {
+      "epoch": 0.38524891536142364,
+      "grad_norm": 0.8451420068740845,
+      "learning_rate": 0.0001408193854726363,
+      "loss": 0.4098,
+      "step": 5239
+    },
+    {
+      "epoch": 0.3853224501801603,
+      "grad_norm": 0.9944307804107666,
+      "learning_rate": 0.00014079764263465148,
+      "loss": 0.3909,
+      "step": 5240
+    },
+    {
+      "epoch": 0.385395984998897,
+      "grad_norm": 1.2261481285095215,
+      "learning_rate": 0.0001407758974825813,
+      "loss": 0.4128,
+      "step": 5241
+    },
+    {
+      "epoch": 0.38546951981763367,
+      "grad_norm": 0.6743568181991577,
+      "learning_rate": 0.00014075415001765923,
+      "loss": 0.4086,
+      "step": 5242
+    },
+    {
+      "epoch": 0.38554305463637034,
+      "grad_norm": 1.5041277408599854,
+      "learning_rate": 0.00014073240024111872,
+      "loss": 0.4167,
+      "step": 5243
+    },
+    {
+      "epoch": 0.385616589455107,
+      "grad_norm": 0.7347652316093445,
+      "learning_rate": 0.00014071064815419354,
+      "loss": 0.3941,
+      "step": 5244
+    },
+    {
+      "epoch": 0.38569012427384364,
+      "grad_norm": 1.5358291864395142,
+      "learning_rate": 0.00014068889375811738,
+      "loss": 0.4139,
+      "step": 5245
+    },
+    {
+      "epoch": 0.3857636590925803,
+      "grad_norm": 0.8688084483146667,
+      "learning_rate": 0.0001406671370541242,
+      "loss": 0.427,
+      "step": 5246
+    },
+    {
+      "epoch": 0.385837193911317,
+      "grad_norm": 1.650594711303711,
+      "learning_rate": 0.00014064537804344814,
+      "loss": 0.4181,
+      "step": 5247
+    },
+    {
+      "epoch": 0.38591072873005366,
+      "grad_norm": 0.9864984750747681,
+      "learning_rate": 0.00014062361672732333,
+      "loss": 0.3977,
+      "step": 5248
+    },
+    {
+      "epoch": 0.38598426354879034,
+      "grad_norm": 1.5723658800125122,
+      "learning_rate": 0.00014060185310698414,
+      "loss": 0.3879,
+      "step": 5249
+    },
+    {
+      "epoch": 0.386057798367527,
+      "grad_norm": 1.1201845407485962,
+      "learning_rate": 0.00014058008718366498,
+      "loss": 0.406,
+      "step": 5250
+    },
+    {
+      "epoch": 0.3861313331862637,
+      "grad_norm": 1.540482521057129,
+      "learning_rate": 0.0001405583189586004,
+      "loss": 0.4108,
+      "step": 5251
+    },
+    {
+      "epoch": 0.38620486800500037,
+      "grad_norm": 1.1461591720581055,
+      "learning_rate": 0.00014053654843302523,
+      "loss": 0.3925,
+      "step": 5252
+    },
+    {
+      "epoch": 0.38627840282373704,
+      "grad_norm": 2.0613198280334473,
+      "learning_rate": 0.00014051477560817415,
+      "loss": 0.3994,
+      "step": 5253
+    },
+    {
+      "epoch": 0.3863519376424737,
+      "grad_norm": 1.387586236000061,
+      "learning_rate": 0.00014049300048528233,
+      "loss": 0.4137,
+      "step": 5254
+    },
+    {
+      "epoch": 0.3864254724612104,
+      "grad_norm": 1.826144814491272,
+      "learning_rate": 0.0001404712230655848,
+      "loss": 0.4065,
+      "step": 5255
+    },
+    {
+      "epoch": 0.38649900727994707,
+      "grad_norm": 1.6160030364990234,
+      "learning_rate": 0.00014044944335031676,
+      "loss": 0.4208,
+      "step": 5256
+    },
+    {
+      "epoch": 0.38657254209868375,
+      "grad_norm": 1.6204158067703247,
+      "learning_rate": 0.0001404276613407136,
+      "loss": 0.4131,
+      "step": 5257
+    },
+    {
+      "epoch": 0.3866460769174204,
+      "grad_norm": 1.5740786790847778,
+      "learning_rate": 0.00014040587703801081,
+      "loss": 0.4189,
+      "step": 5258
+    },
+    {
+      "epoch": 0.38671961173615704,
+      "grad_norm": 1.3930875062942505,
+      "learning_rate": 0.00014038409044344404,
+      "loss": 0.4052,
+      "step": 5259
+    },
+    {
+      "epoch": 0.3867931465548937,
+      "grad_norm": 1.2291408777236938,
+      "learning_rate": 0.00014036230155824904,
+      "loss": 0.4276,
+      "step": 5260
+    },
+    {
+      "epoch": 0.3868666813736304,
+      "grad_norm": 1.5938684940338135,
+      "learning_rate": 0.00014034051038366174,
+      "loss": 0.4182,
+      "step": 5261
+    },
+    {
+      "epoch": 0.38694021619236707,
+      "grad_norm": 1.394155740737915,
+      "learning_rate": 0.0001403187169209181,
+      "loss": 0.4105,
+      "step": 5262
+    },
+    {
+      "epoch": 0.38701375101110375,
+      "grad_norm": 1.7210568189620972,
+      "learning_rate": 0.00014029692117125425,
+      "loss": 0.4389,
+      "step": 5263
+    },
+    {
+      "epoch": 0.3870872858298404,
+      "grad_norm": 1.721894383430481,
+      "learning_rate": 0.0001402751231359065,
+      "loss": 0.4157,
+      "step": 5264
+    },
+    {
+      "epoch": 0.3871608206485771,
+      "grad_norm": 0.9717778563499451,
+      "learning_rate": 0.0001402533228161113,
+      "loss": 0.4553,
+      "step": 5265
+    },
+    {
+      "epoch": 0.3872343554673138,
+      "grad_norm": 1.0442895889282227,
+      "learning_rate": 0.00014023152021310512,
+      "loss": 0.4135,
+      "step": 5266
+    },
+    {
+      "epoch": 0.38730789028605045,
+      "grad_norm": 1.2494888305664062,
+      "learning_rate": 0.0001402097153281247,
+      "loss": 0.405,
+      "step": 5267
+    },
+    {
+      "epoch": 0.3873814251047871,
+      "grad_norm": 0.7298777103424072,
+      "learning_rate": 0.00014018790816240677,
+      "loss": 0.4282,
+      "step": 5268
+    },
+    {
+      "epoch": 0.3874549599235238,
+      "grad_norm": 1.0938760042190552,
+      "learning_rate": 0.00014016609871718826,
+      "loss": 0.4332,
+      "step": 5269
+    },
+    {
+      "epoch": 0.3875284947422605,
+      "grad_norm": 0.5289247632026672,
+      "learning_rate": 0.00014014428699370626,
+      "loss": 0.4251,
+      "step": 5270
+    },
+    {
+      "epoch": 0.38760202956099715,
+      "grad_norm": 0.9741097092628479,
+      "learning_rate": 0.00014012247299319792,
+      "loss": 0.399,
+      "step": 5271
+    },
+    {
+      "epoch": 0.38767556437973383,
+      "grad_norm": 0.9131091237068176,
+      "learning_rate": 0.00014010065671690058,
+      "loss": 0.3984,
+      "step": 5272
+    },
+    {
+      "epoch": 0.38774909919847045,
+      "grad_norm": 0.544917643070221,
+      "learning_rate": 0.0001400788381660517,
+      "loss": 0.4114,
+      "step": 5273
+    },
+    {
+      "epoch": 0.3878226340172071,
+      "grad_norm": 0.9091061949729919,
+      "learning_rate": 0.00014005701734188877,
+      "loss": 0.4332,
+      "step": 5274
+    },
+    {
+      "epoch": 0.3878961688359438,
+      "grad_norm": 0.6890798807144165,
+      "learning_rate": 0.00014003519424564956,
+      "loss": 0.3906,
+      "step": 5275
+    },
+    {
+      "epoch": 0.3879697036546805,
+      "grad_norm": 0.5947189331054688,
+      "learning_rate": 0.00014001336887857187,
+      "loss": 0.4092,
+      "step": 5276
+    },
+    {
+      "epoch": 0.38804323847341715,
+      "grad_norm": 0.8869073390960693,
+      "learning_rate": 0.00013999154124189368,
+      "loss": 0.4107,
+      "step": 5277
+    },
+    {
+      "epoch": 0.3881167732921538,
+      "grad_norm": 0.5621468424797058,
+      "learning_rate": 0.00013996971133685307,
+      "loss": 0.423,
+      "step": 5278
+    },
+    {
+      "epoch": 0.3881903081108905,
+      "grad_norm": 0.7538550496101379,
+      "learning_rate": 0.00013994787916468825,
+      "loss": 0.4145,
+      "step": 5279
+    },
+    {
+      "epoch": 0.3882638429296272,
+      "grad_norm": 0.8903896808624268,
+      "learning_rate": 0.00013992604472663758,
+      "loss": 0.4502,
+      "step": 5280
+    },
+    {
+      "epoch": 0.38833737774836385,
+      "grad_norm": 0.7622554898262024,
+      "learning_rate": 0.0001399042080239395,
+      "loss": 0.4031,
+      "step": 5281
+    },
+    {
+      "epoch": 0.38841091256710053,
+      "grad_norm": 0.5227164030075073,
+      "learning_rate": 0.00013988236905783263,
+      "loss": 0.4107,
+      "step": 5282
+    },
+    {
+      "epoch": 0.3884844473858372,
+      "grad_norm": 1.0409135818481445,
+      "learning_rate": 0.00013986052782955573,
+      "loss": 0.3984,
+      "step": 5283
+    },
+    {
+      "epoch": 0.3885579822045739,
+      "grad_norm": 1.032423734664917,
+      "learning_rate": 0.00013983868434034756,
+      "loss": 0.4213,
+      "step": 5284
+    },
+    {
+      "epoch": 0.38863151702331056,
+      "grad_norm": 0.9382403492927551,
+      "learning_rate": 0.00013981683859144722,
+      "loss": 0.4096,
+      "step": 5285
+    },
+    {
+      "epoch": 0.38870505184204723,
+      "grad_norm": 1.8431988954544067,
+      "learning_rate": 0.00013979499058409378,
+      "loss": 0.3868,
+      "step": 5286
+    },
+    {
+      "epoch": 0.38877858666078385,
+      "grad_norm": 0.8853835463523865,
+      "learning_rate": 0.00013977314031952644,
+      "loss": 0.4241,
+      "step": 5287
+    },
+    {
+      "epoch": 0.38885212147952053,
+      "grad_norm": 1.5960603952407837,
+      "learning_rate": 0.00013975128779898462,
+      "loss": 0.4228,
+      "step": 5288
+    },
+    {
+      "epoch": 0.3889256562982572,
+      "grad_norm": 1.0412615537643433,
+      "learning_rate": 0.0001397294330237078,
+      "loss": 0.4051,
+      "step": 5289
+    },
+    {
+      "epoch": 0.3889991911169939,
+      "grad_norm": 2.095820426940918,
+      "learning_rate": 0.00013970757599493565,
+      "loss": 0.4124,
+      "step": 5290
+    },
+    {
+      "epoch": 0.38907272593573056,
+      "grad_norm": 2.0038368701934814,
+      "learning_rate": 0.00013968571671390786,
+      "loss": 0.4211,
+      "step": 5291
+    },
+    {
+      "epoch": 0.38914626075446723,
+      "grad_norm": 1.3713821172714233,
+      "learning_rate": 0.0001396638551818643,
+      "loss": 0.4104,
+      "step": 5292
+    },
+    {
+      "epoch": 0.3892197955732039,
+      "grad_norm": 1.5358986854553223,
+      "learning_rate": 0.00013964199140004507,
+      "loss": 0.4036,
+      "step": 5293
+    },
+    {
+      "epoch": 0.3892933303919406,
+      "grad_norm": 1.3150219917297363,
+      "learning_rate": 0.0001396201253696902,
+      "loss": 0.4063,
+      "step": 5294
+    },
+    {
+      "epoch": 0.38936686521067726,
+      "grad_norm": 1.3374303579330444,
+      "learning_rate": 0.00013959825709204004,
+      "loss": 0.4036,
+      "step": 5295
+    },
+    {
+      "epoch": 0.38944040002941394,
+      "grad_norm": 1.7029755115509033,
+      "learning_rate": 0.00013957638656833497,
+      "loss": 0.3908,
+      "step": 5296
+    },
+    {
+      "epoch": 0.3895139348481506,
+      "grad_norm": 1.2478910684585571,
+      "learning_rate": 0.00013955451379981542,
+      "loss": 0.3913,
+      "step": 5297
+    },
+    {
+      "epoch": 0.3895874696668873,
+      "grad_norm": 1.6946371793746948,
+      "learning_rate": 0.00013953263878772215,
+      "loss": 0.3966,
+      "step": 5298
+    },
+    {
+      "epoch": 0.38966100448562396,
+      "grad_norm": 1.3652795553207397,
+      "learning_rate": 0.00013951076153329587,
+      "loss": 0.4267,
+      "step": 5299
+    },
+    {
+      "epoch": 0.38973453930436064,
+      "grad_norm": 1.9471856355667114,
+      "learning_rate": 0.00013948888203777754,
+      "loss": 0.3976,
+      "step": 5300
+    },
+    {
+      "epoch": 0.38980807412309726,
+      "grad_norm": 1.9655771255493164,
+      "learning_rate": 0.00013946700030240808,
+      "loss": 0.4168,
+      "step": 5301
+    },
+    {
+      "epoch": 0.38988160894183393,
+      "grad_norm": 1.3890243768692017,
+      "learning_rate": 0.00013944511632842877,
+      "loss": 0.4282,
+      "step": 5302
+    },
+    {
+      "epoch": 0.3899551437605706,
+      "grad_norm": 1.3163052797317505,
+      "learning_rate": 0.0001394232301170808,
+      "loss": 0.4076,
+      "step": 5303
+    },
+    {
+      "epoch": 0.3900286785793073,
+      "grad_norm": 1.2908464670181274,
+      "learning_rate": 0.00013940134166960563,
+      "loss": 0.4388,
+      "step": 5304
+    },
+    {
+      "epoch": 0.39010221339804396,
+      "grad_norm": 1.2095389366149902,
+      "learning_rate": 0.00013937945098724478,
+      "loss": 0.3837,
+      "step": 5305
+    },
+    {
+      "epoch": 0.39017574821678064,
+      "grad_norm": 1.4935866594314575,
+      "learning_rate": 0.00013935755807123988,
+      "loss": 0.3881,
+      "step": 5306
+    },
+    {
+      "epoch": 0.3902492830355173,
+      "grad_norm": 1.3453258275985718,
+      "learning_rate": 0.00013933566292283276,
+      "loss": 0.4063,
+      "step": 5307
+    },
+    {
+      "epoch": 0.390322817854254,
+      "grad_norm": 1.5680605173110962,
+      "learning_rate": 0.00013931376554326538,
+      "loss": 0.4087,
+      "step": 5308
+    },
+    {
+      "epoch": 0.39039635267299067,
+      "grad_norm": 1.1878281831741333,
+      "learning_rate": 0.00013929186593377967,
+      "loss": 0.39,
+      "step": 5309
+    },
+    {
+      "epoch": 0.39046988749172734,
+      "grad_norm": 1.789314866065979,
+      "learning_rate": 0.00013926996409561787,
+      "loss": 0.3958,
+      "step": 5310
+    },
+    {
+      "epoch": 0.390543422310464,
+      "grad_norm": 1.5538617372512817,
+      "learning_rate": 0.00013924806003002229,
+      "loss": 0.4162,
+      "step": 5311
+    },
+    {
+      "epoch": 0.3906169571292007,
+      "grad_norm": 1.5750975608825684,
+      "learning_rate": 0.00013922615373823528,
+      "loss": 0.3997,
+      "step": 5312
+    },
+    {
+      "epoch": 0.39069049194793737,
+      "grad_norm": 1.3477016687393188,
+      "learning_rate": 0.00013920424522149947,
+      "loss": 0.3936,
+      "step": 5313
+    },
+    {
+      "epoch": 0.39076402676667404,
+      "grad_norm": 1.5624631643295288,
+      "learning_rate": 0.0001391823344810575,
+      "loss": 0.4064,
+      "step": 5314
+    },
+    {
+      "epoch": 0.39083756158541066,
+      "grad_norm": 1.3331360816955566,
+      "learning_rate": 0.00013916042151815213,
+      "loss": 0.3896,
+      "step": 5315
+    },
+    {
+      "epoch": 0.39091109640414734,
+      "grad_norm": 1.6263724565505981,
+      "learning_rate": 0.00013913850633402635,
+      "loss": 0.4048,
+      "step": 5316
+    },
+    {
+      "epoch": 0.390984631222884,
+      "grad_norm": 1.2256895303726196,
+      "learning_rate": 0.0001391165889299232,
+      "loss": 0.3821,
+      "step": 5317
+    },
+    {
+      "epoch": 0.3910581660416207,
+      "grad_norm": 1.7958414554595947,
+      "learning_rate": 0.0001390946693070858,
+      "loss": 0.4131,
+      "step": 5318
+    },
+    {
+      "epoch": 0.39113170086035737,
+      "grad_norm": 1.3486003875732422,
+      "learning_rate": 0.00013907274746675753,
+      "loss": 0.408,
+      "step": 5319
+    },
+    {
+      "epoch": 0.39120523567909404,
+      "grad_norm": 1.4244935512542725,
+      "learning_rate": 0.0001390508234101818,
+      "loss": 0.4155,
+      "step": 5320
+    },
+    {
+      "epoch": 0.3912787704978307,
+      "grad_norm": 1.0717090368270874,
+      "learning_rate": 0.00013902889713860215,
+      "loss": 0.3872,
+      "step": 5321
+    },
+    {
+      "epoch": 0.3913523053165674,
+      "grad_norm": 1.8384491205215454,
+      "learning_rate": 0.00013900696865326226,
+      "loss": 0.3935,
+      "step": 5322
+    },
+    {
+      "epoch": 0.39142584013530407,
+      "grad_norm": 1.6049431562423706,
+      "learning_rate": 0.0001389850379554059,
+      "loss": 0.4231,
+      "step": 5323
+    },
+    {
+      "epoch": 0.39149937495404075,
+      "grad_norm": 1.641896367073059,
+      "learning_rate": 0.0001389631050462771,
+      "loss": 0.3882,
+      "step": 5324
+    },
+    {
+      "epoch": 0.3915729097727774,
+      "grad_norm": 1.4674493074417114,
+      "learning_rate": 0.00013894116992711986,
+      "loss": 0.4099,
+      "step": 5325
+    },
+    {
+      "epoch": 0.3916464445915141,
+      "grad_norm": 1.4952462911605835,
+      "learning_rate": 0.00013891923259917835,
+      "loss": 0.3957,
+      "step": 5326
+    },
+    {
+      "epoch": 0.3917199794102508,
+      "grad_norm": 1.3734177350997925,
+      "learning_rate": 0.00013889729306369687,
+      "loss": 0.4009,
+      "step": 5327
+    },
+    {
+      "epoch": 0.39179351422898745,
+      "grad_norm": 1.6993376016616821,
+      "learning_rate": 0.0001388753513219199,
+      "loss": 0.4037,
+      "step": 5328
+    },
+    {
+      "epoch": 0.39186704904772407,
+      "grad_norm": 1.5454388856887817,
+      "learning_rate": 0.00013885340737509197,
+      "loss": 0.3862,
+      "step": 5329
+    },
+    {
+      "epoch": 0.39194058386646075,
+      "grad_norm": 1.2768537998199463,
+      "learning_rate": 0.00013883146122445775,
+      "loss": 0.3826,
+      "step": 5330
+    },
+    {
+      "epoch": 0.3920141186851974,
+      "grad_norm": 1.308760404586792,
+      "learning_rate": 0.00013880951287126213,
+      "loss": 0.4235,
+      "step": 5331
+    },
+    {
+      "epoch": 0.3920876535039341,
+      "grad_norm": 1.5896652936935425,
+      "learning_rate": 0.00013878756231674996,
+      "loss": 0.4002,
+      "step": 5332
+    },
+    {
+      "epoch": 0.3921611883226708,
+      "grad_norm": 1.4374995231628418,
+      "learning_rate": 0.0001387656095621663,
+      "loss": 0.4235,
+      "step": 5333
+    },
+    {
+      "epoch": 0.39223472314140745,
+      "grad_norm": 1.6060630083084106,
+      "learning_rate": 0.00013874365460875637,
+      "loss": 0.4214,
+      "step": 5334
+    },
+    {
+      "epoch": 0.3923082579601441,
+      "grad_norm": 1.212438941001892,
+      "learning_rate": 0.00013872169745776544,
+      "loss": 0.3795,
+      "step": 5335
+    },
+    {
+      "epoch": 0.3923817927788808,
+      "grad_norm": 1.4265576601028442,
+      "learning_rate": 0.000138699738110439,
+      "loss": 0.382,
+      "step": 5336
+    },
+    {
+      "epoch": 0.3924553275976175,
+      "grad_norm": 1.110398769378662,
+      "learning_rate": 0.00013867777656802256,
+      "loss": 0.3897,
+      "step": 5337
+    },
+    {
+      "epoch": 0.39252886241635415,
+      "grad_norm": 1.7398182153701782,
+      "learning_rate": 0.00013865581283176178,
+      "loss": 0.3854,
+      "step": 5338
+    },
+    {
+      "epoch": 0.3926023972350908,
+      "grad_norm": 1.7041441202163696,
+      "learning_rate": 0.00013863384690290256,
+      "loss": 0.4007,
+      "step": 5339
+    },
+    {
+      "epoch": 0.3926759320538275,
+      "grad_norm": 1.3538641929626465,
+      "learning_rate": 0.00013861187878269073,
+      "loss": 0.3959,
+      "step": 5340
+    },
+    {
+      "epoch": 0.3927494668725642,
+      "grad_norm": 1.1713792085647583,
+      "learning_rate": 0.0001385899084723724,
+      "loss": 0.4059,
+      "step": 5341
+    },
+    {
+      "epoch": 0.39282300169130085,
+      "grad_norm": 1.52528977394104,
+      "learning_rate": 0.00013856793597319373,
+      "loss": 0.4035,
+      "step": 5342
+    },
+    {
+      "epoch": 0.3928965365100375,
+      "grad_norm": 1.0591518878936768,
+      "learning_rate": 0.00013854596128640105,
+      "loss": 0.3882,
+      "step": 5343
+    },
+    {
+      "epoch": 0.39297007132877415,
+      "grad_norm": 1.6884559392929077,
+      "learning_rate": 0.00013852398441324074,
+      "loss": 0.3753,
+      "step": 5344
+    },
+    {
+      "epoch": 0.3930436061475108,
+      "grad_norm": 1.7901287078857422,
+      "learning_rate": 0.0001385020053549594,
+      "loss": 0.3786,
+      "step": 5345
+    },
+    {
+      "epoch": 0.3931171409662475,
+      "grad_norm": 1.0582258701324463,
+      "learning_rate": 0.00013848002411280365,
+      "loss": 0.3917,
+      "step": 5346
+    },
+    {
+      "epoch": 0.3931906757849842,
+      "grad_norm": 1.2320116758346558,
+      "learning_rate": 0.00013845804068802036,
+      "loss": 0.3928,
+      "step": 5347
+    },
+    {
+      "epoch": 0.39326421060372085,
+      "grad_norm": 1.1326704025268555,
+      "learning_rate": 0.00013843605508185637,
+      "loss": 0.3673,
+      "step": 5348
+    },
+    {
+      "epoch": 0.39333774542245753,
+      "grad_norm": 0.805597722530365,
+      "learning_rate": 0.00013841406729555886,
+      "loss": 0.4327,
+      "step": 5349
+    },
+    {
+      "epoch": 0.3934112802411942,
+      "grad_norm": 1.3479875326156616,
+      "learning_rate": 0.00013839207733037484,
+      "loss": 0.359,
+      "step": 5350
+    },
+    {
+      "epoch": 0.3934848150599309,
+      "grad_norm": 0.7537283301353455,
+      "learning_rate": 0.00013837008518755167,
+      "loss": 0.3885,
+      "step": 5351
+    },
+    {
+      "epoch": 0.39355834987866756,
+      "grad_norm": 1.5129220485687256,
+      "learning_rate": 0.00013834809086833683,
+      "loss": 0.3878,
+      "step": 5352
+    },
+    {
+      "epoch": 0.39363188469740423,
+      "grad_norm": 0.9027190208435059,
+      "learning_rate": 0.00013832609437397774,
+      "loss": 0.3688,
+      "step": 5353
+    },
+    {
+      "epoch": 0.3937054195161409,
+      "grad_norm": 1.531198501586914,
+      "learning_rate": 0.0001383040957057222,
+      "loss": 0.4124,
+      "step": 5354
+    },
+    {
+      "epoch": 0.3937789543348776,
+      "grad_norm": 1.2161294221878052,
+      "learning_rate": 0.0001382820948648179,
+      "loss": 0.3897,
+      "step": 5355
+    },
+    {
+      "epoch": 0.39385248915361426,
+      "grad_norm": 1.7406623363494873,
+      "learning_rate": 0.00013826009185251273,
+      "loss": 0.3824,
+      "step": 5356
+    },
+    {
+      "epoch": 0.3939260239723509,
+      "grad_norm": 1.7775930166244507,
+      "learning_rate": 0.00013823808667005481,
+      "loss": 0.3789,
+      "step": 5357
+    },
+    {
+      "epoch": 0.39399955879108756,
+      "grad_norm": 0.7175175547599792,
+      "learning_rate": 0.00013821607931869227,
+      "loss": 0.3773,
+      "step": 5358
+    },
+    {
+      "epoch": 0.39407309360982423,
+      "grad_norm": 1.5715382099151611,
+      "learning_rate": 0.00013819406979967338,
+      "loss": 0.3617,
+      "step": 5359
+    },
+    {
+      "epoch": 0.3941466284285609,
+      "grad_norm": 0.6926132440567017,
+      "learning_rate": 0.00013817205811424653,
+      "loss": 0.3849,
+      "step": 5360
+    },
+    {
+      "epoch": 0.3942201632472976,
+      "grad_norm": 1.5308811664581299,
+      "learning_rate": 0.0001381500442636603,
+      "loss": 0.3765,
+      "step": 5361
+    },
+    {
+      "epoch": 0.39429369806603426,
+      "grad_norm": 1.0440236330032349,
+      "learning_rate": 0.00013812802824916324,
+      "loss": 0.4142,
+      "step": 5362
+    },
+    {
+      "epoch": 0.39436723288477094,
+      "grad_norm": 1.4601480960845947,
+      "learning_rate": 0.0001381060100720042,
+      "loss": 0.4063,
+      "step": 5363
+    },
+    {
+      "epoch": 0.3944407677035076,
+      "grad_norm": 1.601181983947754,
+      "learning_rate": 0.00013808398973343206,
+      "loss": 0.3869,
+      "step": 5364
+    },
+    {
+      "epoch": 0.3945143025222443,
+      "grad_norm": 0.6037935018539429,
+      "learning_rate": 0.00013806196723469582,
+      "loss": 0.3954,
+      "step": 5365
+    },
+    {
+      "epoch": 0.39458783734098096,
+      "grad_norm": 1.6026017665863037,
+      "learning_rate": 0.00013803994257704464,
+      "loss": 0.3894,
+      "step": 5366
+    },
+    {
+      "epoch": 0.39466137215971764,
+      "grad_norm": 0.6957495808601379,
+      "learning_rate": 0.00013801791576172778,
+      "loss": 0.3918,
+      "step": 5367
+    },
+    {
+      "epoch": 0.3947349069784543,
+      "grad_norm": 1.3699085712432861,
+      "learning_rate": 0.00013799588678999462,
+      "loss": 0.3841,
+      "step": 5368
+    },
+    {
+      "epoch": 0.394808441797191,
+      "grad_norm": 1.0491230487823486,
+      "learning_rate": 0.0001379738556630946,
+      "loss": 0.3771,
+      "step": 5369
+    },
+    {
+      "epoch": 0.39488197661592767,
+      "grad_norm": 1.406994342803955,
+      "learning_rate": 0.0001379518223822775,
+      "loss": 0.3728,
+      "step": 5370
+    },
+    {
+      "epoch": 0.3949555114346643,
+      "grad_norm": 1.3562266826629639,
+      "learning_rate": 0.0001379297869487929,
+      "loss": 0.4077,
+      "step": 5371
+    },
+    {
+      "epoch": 0.39502904625340096,
+      "grad_norm": 0.9112411737442017,
+      "learning_rate": 0.00013790774936389084,
+      "loss": 0.3756,
+      "step": 5372
+    },
+    {
+      "epoch": 0.39510258107213764,
+      "grad_norm": 0.8936380743980408,
+      "learning_rate": 0.0001378857096288212,
+      "loss": 0.4051,
+      "step": 5373
+    },
+    {
+      "epoch": 0.3951761158908743,
+      "grad_norm": 0.7747612595558167,
+      "learning_rate": 0.00013786366774483407,
+      "loss": 0.3918,
+      "step": 5374
+    },
+    {
+      "epoch": 0.395249650709611,
+      "grad_norm": 0.8968454599380493,
+      "learning_rate": 0.0001378416237131798,
+      "loss": 0.3994,
+      "step": 5375
+    },
+    {
+      "epoch": 0.39532318552834766,
+      "grad_norm": 0.9498363137245178,
+      "learning_rate": 0.00013781957753510867,
+      "loss": 0.3835,
+      "step": 5376
+    },
+    {
+      "epoch": 0.39539672034708434,
+      "grad_norm": 0.6368041634559631,
+      "learning_rate": 0.00013779752921187123,
+      "loss": 0.3534,
+      "step": 5377
+    },
+    {
+      "epoch": 0.395470255165821,
+      "grad_norm": 0.8061744570732117,
+      "learning_rate": 0.00013777547874471805,
+      "loss": 0.3904,
+      "step": 5378
+    },
+    {
+      "epoch": 0.3955437899845577,
+      "grad_norm": 0.7620262503623962,
+      "learning_rate": 0.0001377534261348998,
+      "loss": 0.3881,
+      "step": 5379
+    },
+    {
+      "epoch": 0.39561732480329437,
+      "grad_norm": 0.7316908836364746,
+      "learning_rate": 0.0001377313713836674,
+      "loss": 0.3873,
+      "step": 5380
+    },
+    {
+      "epoch": 0.39569085962203104,
+      "grad_norm": 0.7250636219978333,
+      "learning_rate": 0.0001377093144922718,
+      "loss": 0.388,
+      "step": 5381
+    },
+    {
+      "epoch": 0.3957643944407677,
+      "grad_norm": 1.0403892993927002,
+      "learning_rate": 0.00013768725546196405,
+      "loss": 0.3998,
+      "step": 5382
+    },
+    {
+      "epoch": 0.3958379292595044,
+      "grad_norm": 0.49597427248954773,
+      "learning_rate": 0.00013766519429399545,
+      "loss": 0.369,
+      "step": 5383
+    },
+    {
+      "epoch": 0.39591146407824107,
+      "grad_norm": 0.8841545581817627,
+      "learning_rate": 0.00013764313098961725,
+      "loss": 0.3876,
+      "step": 5384
+    },
+    {
+      "epoch": 0.3959849988969777,
+      "grad_norm": 0.6609196066856384,
+      "learning_rate": 0.00013762106555008095,
+      "loss": 0.3934,
+      "step": 5385
+    },
+    {
+      "epoch": 0.39605853371571437,
+      "grad_norm": 0.5624990463256836,
+      "learning_rate": 0.0001375989979766381,
+      "loss": 0.406,
+      "step": 5386
+    },
+    {
+      "epoch": 0.39613206853445104,
+      "grad_norm": 0.802539050579071,
+      "learning_rate": 0.00013757692827054042,
+      "loss": 0.3656,
+      "step": 5387
+    },
+    {
+      "epoch": 0.3962056033531877,
+      "grad_norm": 0.803851842880249,
+      "learning_rate": 0.00013755485643303968,
+      "loss": 0.3954,
+      "step": 5388
+    },
+    {
+      "epoch": 0.3962791381719244,
+      "grad_norm": 0.484950989484787,
+      "learning_rate": 0.00013753278246538786,
+      "loss": 0.382,
+      "step": 5389
+    },
+    {
+      "epoch": 0.39635267299066107,
+      "grad_norm": 0.8793902397155762,
+      "learning_rate": 0.00013751070636883701,
+      "loss": 0.3892,
+      "step": 5390
+    },
+    {
+      "epoch": 0.39642620780939775,
+      "grad_norm": 0.6134843230247498,
+      "learning_rate": 0.00013748862814463934,
+      "loss": 0.392,
+      "step": 5391
+    },
+    {
+      "epoch": 0.3964997426281344,
+      "grad_norm": 0.7753165364265442,
+      "learning_rate": 0.0001374665477940471,
+      "loss": 0.3986,
+      "step": 5392
+    },
+    {
+      "epoch": 0.3965732774468711,
+      "grad_norm": 0.9749565124511719,
+      "learning_rate": 0.0001374444653183127,
+      "loss": 0.3876,
+      "step": 5393
+    },
+    {
+      "epoch": 0.3966468122656078,
+      "grad_norm": 0.9048434495925903,
+      "learning_rate": 0.00013742238071868873,
+      "loss": 0.3728,
+      "step": 5394
+    },
+    {
+      "epoch": 0.39672034708434445,
+      "grad_norm": 0.6371535658836365,
+      "learning_rate": 0.00013740029399642782,
+      "loss": 0.3827,
+      "step": 5395
+    },
+    {
+      "epoch": 0.3967938819030811,
+      "grad_norm": 0.7106906175613403,
+      "learning_rate": 0.00013737820515278282,
+      "loss": 0.4036,
+      "step": 5396
+    },
+    {
+      "epoch": 0.3968674167218178,
+      "grad_norm": 0.6347586512565613,
+      "learning_rate": 0.00013735611418900652,
+      "loss": 0.4103,
+      "step": 5397
+    },
+    {
+      "epoch": 0.3969409515405545,
+      "grad_norm": 0.5792220234870911,
+      "learning_rate": 0.00013733402110635201,
+      "loss": 0.3953,
+      "step": 5398
+    },
+    {
+      "epoch": 0.3970144863592911,
+      "grad_norm": 0.778342068195343,
+      "learning_rate": 0.0001373119259060724,
+      "loss": 0.4033,
+      "step": 5399
+    },
+    {
+      "epoch": 0.3970880211780278,
+      "grad_norm": 0.6617233157157898,
+      "learning_rate": 0.000137289828589421,
+      "loss": 0.4124,
+      "step": 5400
+    },
+    {
+      "epoch": 0.39716155599676445,
+      "grad_norm": 0.564399778842926,
+      "learning_rate": 0.00013726772915765122,
+      "loss": 0.4053,
+      "step": 5401
+    },
+    {
+      "epoch": 0.3972350908155011,
+      "grad_norm": 0.9219571352005005,
+      "learning_rate": 0.00013724562761201645,
+      "loss": 0.3957,
+      "step": 5402
+    },
+    {
+      "epoch": 0.3973086256342378,
+      "grad_norm": 0.647930383682251,
+      "learning_rate": 0.00013722352395377038,
+      "loss": 0.4259,
+      "step": 5403
+    },
+    {
+      "epoch": 0.3973821604529745,
+      "grad_norm": 0.8791479468345642,
+      "learning_rate": 0.00013720141818416673,
+      "loss": 0.4263,
+      "step": 5404
+    },
+    {
+      "epoch": 0.39745569527171115,
+      "grad_norm": 1.5688081979751587,
+      "learning_rate": 0.00013717931030445942,
+      "loss": 0.4188,
+      "step": 5405
+    },
+    {
+      "epoch": 0.3975292300904478,
+      "grad_norm": 0.6810703277587891,
+      "learning_rate": 0.0001371572003159024,
+      "loss": 0.3869,
+      "step": 5406
+    },
+    {
+      "epoch": 0.3976027649091845,
+      "grad_norm": 1.4565480947494507,
+      "learning_rate": 0.00013713508821974974,
+      "loss": 0.3975,
+      "step": 5407
+    },
+    {
+      "epoch": 0.3976762997279212,
+      "grad_norm": 0.997207522392273,
+      "learning_rate": 0.00013711297401725568,
+      "loss": 0.3954,
+      "step": 5408
+    },
+    {
+      "epoch": 0.39774983454665785,
+      "grad_norm": 1.292260766029358,
+      "learning_rate": 0.0001370908577096746,
+      "loss": 0.4269,
+      "step": 5409
+    },
+    {
+      "epoch": 0.39782336936539453,
+      "grad_norm": 1.091915249824524,
+      "learning_rate": 0.00013706873929826088,
+      "loss": 0.4045,
+      "step": 5410
+    },
+    {
+      "epoch": 0.3978969041841312,
+      "grad_norm": 1.258020281791687,
+      "learning_rate": 0.00013704661878426917,
+      "loss": 0.4016,
+      "step": 5411
+    },
+    {
+      "epoch": 0.3979704390028679,
+      "grad_norm": 1.271705985069275,
+      "learning_rate": 0.00013702449616895414,
+      "loss": 0.4028,
+      "step": 5412
+    },
+    {
+      "epoch": 0.3980439738216045,
+      "grad_norm": 1.0584553480148315,
+      "learning_rate": 0.00013700237145357065,
+      "loss": 0.4162,
+      "step": 5413
+    },
+    {
+      "epoch": 0.3981175086403412,
+      "grad_norm": 1.1587964296340942,
+      "learning_rate": 0.00013698024463937357,
+      "loss": 0.4183,
+      "step": 5414
+    },
+    {
+      "epoch": 0.39819104345907785,
+      "grad_norm": 1.065958023071289,
+      "learning_rate": 0.00013695811572761799,
+      "loss": 0.4147,
+      "step": 5415
+    },
+    {
+      "epoch": 0.39826457827781453,
+      "grad_norm": 0.8832803964614868,
+      "learning_rate": 0.0001369359847195591,
+      "loss": 0.4092,
+      "step": 5416
+    },
+    {
+      "epoch": 0.3983381130965512,
+      "grad_norm": 0.8435798287391663,
+      "learning_rate": 0.00013691385161645212,
+      "loss": 0.4046,
+      "step": 5417
+    },
+    {
+      "epoch": 0.3984116479152879,
+      "grad_norm": 0.7472260594367981,
+      "learning_rate": 0.00013689171641955258,
+      "loss": 0.4197,
+      "step": 5418
+    },
+    {
+      "epoch": 0.39848518273402456,
+      "grad_norm": 0.9601767063140869,
+      "learning_rate": 0.00013686957913011593,
+      "loss": 0.4461,
+      "step": 5419
+    },
+    {
+      "epoch": 0.39855871755276123,
+      "grad_norm": 0.6410945653915405,
+      "learning_rate": 0.0001368474397493978,
+      "loss": 0.4251,
+      "step": 5420
+    },
+    {
+      "epoch": 0.3986322523714979,
+      "grad_norm": 0.7576264142990112,
+      "learning_rate": 0.00013682529827865406,
+      "loss": 0.3892,
+      "step": 5421
+    },
+    {
+      "epoch": 0.3987057871902346,
+      "grad_norm": 0.5743565559387207,
+      "learning_rate": 0.0001368031547191405,
+      "loss": 0.4434,
+      "step": 5422
+    },
+    {
+      "epoch": 0.39877932200897126,
+      "grad_norm": 0.5982747077941895,
+      "learning_rate": 0.0001367810090721132,
+      "loss": 0.3961,
+      "step": 5423
+    },
+    {
+      "epoch": 0.39885285682770794,
+      "grad_norm": 0.7063373327255249,
+      "learning_rate": 0.00013675886133882826,
+      "loss": 0.4058,
+      "step": 5424
+    },
+    {
+      "epoch": 0.3989263916464446,
+      "grad_norm": 0.7009197473526001,
+      "learning_rate": 0.00013673671152054188,
+      "loss": 0.4162,
+      "step": 5425
+    },
+    {
+      "epoch": 0.3989999264651813,
+      "grad_norm": 0.6795799732208252,
+      "learning_rate": 0.00013671455961851046,
+      "loss": 0.3926,
+      "step": 5426
+    },
+    {
+      "epoch": 0.3990734612839179,
+      "grad_norm": 0.5957397222518921,
+      "learning_rate": 0.00013669240563399045,
+      "loss": 0.3861,
+      "step": 5427
+    },
+    {
+      "epoch": 0.3991469961026546,
+      "grad_norm": 0.734565794467926,
+      "learning_rate": 0.00013667024956823848,
+      "loss": 0.404,
+      "step": 5428
+    },
+    {
+      "epoch": 0.39922053092139126,
+      "grad_norm": 0.7476994395256042,
+      "learning_rate": 0.00013664809142251128,
+      "loss": 0.4212,
+      "step": 5429
+    },
+    {
+      "epoch": 0.39929406574012793,
+      "grad_norm": 0.6418420076370239,
+      "learning_rate": 0.00013662593119806562,
+      "loss": 0.4044,
+      "step": 5430
+    },
+    {
+      "epoch": 0.3993676005588646,
+      "grad_norm": 0.5788286924362183,
+      "learning_rate": 0.00013660376889615853,
+      "loss": 0.39,
+      "step": 5431
+    },
+    {
+      "epoch": 0.3994411353776013,
+      "grad_norm": 0.602287769317627,
+      "learning_rate": 0.00013658160451804702,
+      "loss": 0.3706,
+      "step": 5432
+    },
+    {
+      "epoch": 0.39951467019633796,
+      "grad_norm": 0.6339549422264099,
+      "learning_rate": 0.0001365594380649883,
+      "loss": 0.4179,
+      "step": 5433
+    },
+    {
+      "epoch": 0.39958820501507464,
+      "grad_norm": 0.5719407200813293,
+      "learning_rate": 0.00013653726953823965,
+      "loss": 0.3858,
+      "step": 5434
+    },
+    {
+      "epoch": 0.3996617398338113,
+      "grad_norm": 0.6520179510116577,
+      "learning_rate": 0.00013651509893905853,
+      "loss": 0.414,
+      "step": 5435
+    },
+    {
+      "epoch": 0.399735274652548,
+      "grad_norm": 0.6968705058097839,
+      "learning_rate": 0.0001364929262687025,
+      "loss": 0.3822,
+      "step": 5436
+    },
+    {
+      "epoch": 0.39980880947128467,
+      "grad_norm": 0.7189019322395325,
+      "learning_rate": 0.00013647075152842914,
+      "loss": 0.3843,
+      "step": 5437
+    },
+    {
+      "epoch": 0.39988234429002134,
+      "grad_norm": 0.7830271124839783,
+      "learning_rate": 0.00013644857471949623,
+      "loss": 0.3942,
+      "step": 5438
+    },
+    {
+      "epoch": 0.399955879108758,
+      "grad_norm": 0.9444780349731445,
+      "learning_rate": 0.00013642639584316175,
+      "loss": 0.3934,
+      "step": 5439
+    },
+    {
+      "epoch": 0.4000294139274947,
+      "grad_norm": 0.6030577421188354,
+      "learning_rate": 0.00013640421490068364,
+      "loss": 0.3953,
+      "step": 5440
+    },
+    {
+      "epoch": 0.40010294874623137,
+      "grad_norm": 0.9044516086578369,
+      "learning_rate": 0.00013638203189332,
+      "loss": 0.3997,
+      "step": 5441
+    },
+    {
+      "epoch": 0.400176483564968,
+      "grad_norm": 0.6214029788970947,
+      "learning_rate": 0.00013635984682232918,
+      "loss": 0.3984,
+      "step": 5442
+    },
+    {
+      "epoch": 0.40025001838370466,
+      "grad_norm": 0.6807915568351746,
+      "learning_rate": 0.00013633765968896946,
+      "loss": 0.3974,
+      "step": 5443
+    },
+    {
+      "epoch": 0.40032355320244134,
+      "grad_norm": 0.6825141906738281,
+      "learning_rate": 0.00013631547049449934,
+      "loss": 0.3886,
+      "step": 5444
+    },
+    {
+      "epoch": 0.400397088021178,
+      "grad_norm": 0.8937591910362244,
+      "learning_rate": 0.00013629327924017735,
+      "loss": 0.4161,
+      "step": 5445
+    },
+    {
+      "epoch": 0.4004706228399147,
+      "grad_norm": 0.47570785880088806,
+      "learning_rate": 0.00013627108592726233,
+      "loss": 0.3981,
+      "step": 5446
+    },
+    {
+      "epoch": 0.40054415765865137,
+      "grad_norm": 0.9240851402282715,
+      "learning_rate": 0.000136248890557013,
+      "loss": 0.3777,
+      "step": 5447
+    },
+    {
+      "epoch": 0.40061769247738804,
+      "grad_norm": 0.8913319706916809,
+      "learning_rate": 0.00013622669313068835,
+      "loss": 0.378,
+      "step": 5448
+    },
+    {
+      "epoch": 0.4006912272961247,
+      "grad_norm": 0.8909361362457275,
+      "learning_rate": 0.00013620449364954744,
+      "loss": 0.4269,
+      "step": 5449
+    },
+    {
+      "epoch": 0.4007647621148614,
+      "grad_norm": 0.7733377814292908,
+      "learning_rate": 0.00013618229211484944,
+      "loss": 0.417,
+      "step": 5450
+    },
+    {
+      "epoch": 0.40083829693359807,
+      "grad_norm": 1.0607014894485474,
+      "learning_rate": 0.00013616008852785364,
+      "loss": 0.4225,
+      "step": 5451
+    },
+    {
+      "epoch": 0.40091183175233475,
+      "grad_norm": 0.9044339060783386,
+      "learning_rate": 0.00013613788288981946,
+      "loss": 0.399,
+      "step": 5452
+    },
+    {
+      "epoch": 0.4009853665710714,
+      "grad_norm": 0.6224743127822876,
+      "learning_rate": 0.00013611567520200641,
+      "loss": 0.3701,
+      "step": 5453
+    },
+    {
+      "epoch": 0.4010589013898081,
+      "grad_norm": 1.0704073905944824,
+      "learning_rate": 0.0001360934654656742,
+      "loss": 0.3854,
+      "step": 5454
+    },
+    {
+      "epoch": 0.4011324362085448,
+      "grad_norm": 0.8742216229438782,
+      "learning_rate": 0.0001360712536820825,
+      "loss": 0.4031,
+      "step": 5455
+    },
+    {
+      "epoch": 0.4012059710272814,
+      "grad_norm": 1.0123467445373535,
+      "learning_rate": 0.00013604903985249122,
+      "loss": 0.4248,
+      "step": 5456
+    },
+    {
+      "epoch": 0.40127950584601807,
+      "grad_norm": 0.8110641241073608,
+      "learning_rate": 0.00013602682397816036,
+      "loss": 0.4,
+      "step": 5457
+    },
+    {
+      "epoch": 0.40135304066475475,
+      "grad_norm": 1.1058768033981323,
+      "learning_rate": 0.00013600460606035003,
+      "loss": 0.3902,
+      "step": 5458
+    },
+    {
+      "epoch": 0.4014265754834914,
+      "grad_norm": 0.7010652422904968,
+      "learning_rate": 0.00013598238610032043,
+      "loss": 0.4212,
+      "step": 5459
+    },
+    {
+      "epoch": 0.4015001103022281,
+      "grad_norm": 0.8146635890007019,
+      "learning_rate": 0.00013596016409933193,
+      "loss": 0.402,
+      "step": 5460
+    },
+    {
+      "epoch": 0.4015736451209648,
+      "grad_norm": 0.8571239709854126,
+      "learning_rate": 0.00013593794005864498,
+      "loss": 0.3897,
+      "step": 5461
+    },
+    {
+      "epoch": 0.40164717993970145,
+      "grad_norm": 0.5545797944068909,
+      "learning_rate": 0.00013591571397952013,
+      "loss": 0.3952,
+      "step": 5462
+    },
+    {
+      "epoch": 0.4017207147584381,
+      "grad_norm": 0.9537542462348938,
+      "learning_rate": 0.00013589348586321806,
+      "loss": 0.384,
+      "step": 5463
+    },
+    {
+      "epoch": 0.4017942495771748,
+      "grad_norm": 0.5953315496444702,
+      "learning_rate": 0.00013587125571099963,
+      "loss": 0.3757,
+      "step": 5464
+    },
+    {
+      "epoch": 0.4018677843959115,
+      "grad_norm": 1.0505051612854004,
+      "learning_rate": 0.0001358490235241257,
+      "loss": 0.3819,
+      "step": 5465
+    },
+    {
+      "epoch": 0.40194131921464815,
+      "grad_norm": 0.9418569207191467,
+      "learning_rate": 0.00013582678930385735,
+      "loss": 0.3841,
+      "step": 5466
+    },
+    {
+      "epoch": 0.4020148540333848,
+      "grad_norm": 1.3048912286758423,
+      "learning_rate": 0.00013580455305145568,
+      "loss": 0.4079,
+      "step": 5467
+    },
+    {
+      "epoch": 0.4020883888521215,
+      "grad_norm": 0.6403235197067261,
+      "learning_rate": 0.00013578231476818196,
+      "loss": 0.3802,
+      "step": 5468
+    },
+    {
+      "epoch": 0.4021619236708582,
+      "grad_norm": 1.1453293561935425,
+      "learning_rate": 0.0001357600744552976,
+      "loss": 0.3789,
+      "step": 5469
+    },
+    {
+      "epoch": 0.4022354584895948,
+      "grad_norm": 1.2594571113586426,
+      "learning_rate": 0.00013573783211406407,
+      "loss": 0.3897,
+      "step": 5470
+    },
+    {
+      "epoch": 0.4023089933083315,
+      "grad_norm": 0.936592161655426,
+      "learning_rate": 0.00013571558774574298,
+      "loss": 0.3906,
+      "step": 5471
+    },
+    {
+      "epoch": 0.40238252812706815,
+      "grad_norm": 1.2038110494613647,
+      "learning_rate": 0.0001356933413515961,
+      "loss": 0.3613,
+      "step": 5472
+    },
+    {
+      "epoch": 0.4024560629458048,
+      "grad_norm": 1.3452447652816772,
+      "learning_rate": 0.00013567109293288523,
+      "loss": 0.3862,
+      "step": 5473
+    },
+    {
+      "epoch": 0.4025295977645415,
+      "grad_norm": 0.8438129425048828,
+      "learning_rate": 0.00013564884249087227,
+      "loss": 0.4012,
+      "step": 5474
+    },
+    {
+      "epoch": 0.4026031325832782,
+      "grad_norm": 1.4160667657852173,
+      "learning_rate": 0.00013562659002681936,
+      "loss": 0.3804,
+      "step": 5475
+    },
+    {
+      "epoch": 0.40267666740201485,
+      "grad_norm": 0.69748854637146,
+      "learning_rate": 0.00013560433554198867,
+      "loss": 0.3752,
+      "step": 5476
+    },
+    {
+      "epoch": 0.40275020222075153,
+      "grad_norm": 1.5956225395202637,
+      "learning_rate": 0.00013558207903764254,
+      "loss": 0.3926,
+      "step": 5477
+    },
+    {
+      "epoch": 0.4028237370394882,
+      "grad_norm": 0.6061131954193115,
+      "learning_rate": 0.00013555982051504327,
+      "loss": 0.3733,
+      "step": 5478
+    },
+    {
+      "epoch": 0.4028972718582249,
+      "grad_norm": 1.7973487377166748,
+      "learning_rate": 0.00013553755997545346,
+      "loss": 0.3861,
+      "step": 5479
+    },
+    {
+      "epoch": 0.40297080667696156,
+      "grad_norm": 0.8704853057861328,
+      "learning_rate": 0.00013551529742013578,
+      "loss": 0.3938,
+      "step": 5480
+    },
+    {
+      "epoch": 0.40304434149569823,
+      "grad_norm": 2.7593562602996826,
+      "learning_rate": 0.0001354930328503529,
+      "loss": 0.4161,
+      "step": 5481
+    },
+    {
+      "epoch": 0.4031178763144349,
+      "grad_norm": 2.5529003143310547,
+      "learning_rate": 0.00013547076626736776,
+      "loss": 0.3834,
+      "step": 5482
+    },
+    {
+      "epoch": 0.4031914111331716,
+      "grad_norm": 0.9778338670730591,
+      "learning_rate": 0.00013544849767244336,
+      "loss": 0.393,
+      "step": 5483
+    },
+    {
+      "epoch": 0.4032649459519082,
+      "grad_norm": 1.3713244199752808,
+      "learning_rate": 0.00013542622706684275,
+      "loss": 0.3846,
+      "step": 5484
+    },
+    {
+      "epoch": 0.4033384807706449,
+      "grad_norm": 1.2620426416397095,
+      "learning_rate": 0.00013540395445182913,
+      "loss": 0.3822,
+      "step": 5485
+    },
+    {
+      "epoch": 0.40341201558938156,
+      "grad_norm": 0.7238827347755432,
+      "learning_rate": 0.00013538167982866584,
+      "loss": 0.4474,
+      "step": 5486
+    },
+    {
+      "epoch": 0.40348555040811823,
+      "grad_norm": 1.360980749130249,
+      "learning_rate": 0.00013535940319861636,
+      "loss": 0.3817,
+      "step": 5487
+    },
+    {
+      "epoch": 0.4035590852268549,
+      "grad_norm": 0.5389792323112488,
+      "learning_rate": 0.0001353371245629442,
+      "loss": 0.3991,
+      "step": 5488
+    },
+    {
+      "epoch": 0.4036326200455916,
+      "grad_norm": 1.1982396841049194,
+      "learning_rate": 0.00013531484392291304,
+      "loss": 0.3766,
+      "step": 5489
+    },
+    {
+      "epoch": 0.40370615486432826,
+      "grad_norm": 1.074958086013794,
+      "learning_rate": 0.0001352925612797867,
+      "loss": 0.4067,
+      "step": 5490
+    },
+    {
+      "epoch": 0.40377968968306494,
+      "grad_norm": 0.5041555166244507,
+      "learning_rate": 0.00013527027663482897,
+      "loss": 0.4271,
+      "step": 5491
+    },
+    {
+      "epoch": 0.4038532245018016,
+      "grad_norm": 1.3280338048934937,
+      "learning_rate": 0.00013524798998930398,
+      "loss": 0.3777,
+      "step": 5492
+    },
+    {
+      "epoch": 0.4039267593205383,
+      "grad_norm": 0.7686206698417664,
+      "learning_rate": 0.0001352257013444758,
+      "loss": 0.3948,
+      "step": 5493
+    },
+    {
+      "epoch": 0.40400029413927496,
+      "grad_norm": 0.7341105937957764,
+      "learning_rate": 0.00013520341070160863,
+      "loss": 0.3967,
+      "step": 5494
+    },
+    {
+      "epoch": 0.40407382895801164,
+      "grad_norm": 0.9117696285247803,
+      "learning_rate": 0.00013518111806196687,
+      "loss": 0.3752,
+      "step": 5495
+    },
+    {
+      "epoch": 0.4041473637767483,
+      "grad_norm": 0.7535858750343323,
+      "learning_rate": 0.00013515882342681498,
+      "loss": 0.391,
+      "step": 5496
+    },
+    {
+      "epoch": 0.404220898595485,
+      "grad_norm": 0.7512775659561157,
+      "learning_rate": 0.00013513652679741748,
+      "loss": 0.3704,
+      "step": 5497
+    },
+    {
+      "epoch": 0.4042944334142216,
+      "grad_norm": 0.5163897275924683,
+      "learning_rate": 0.00013511422817503915,
+      "loss": 0.3663,
+      "step": 5498
+    },
+    {
+      "epoch": 0.4043679682329583,
+      "grad_norm": 0.6260879039764404,
+      "learning_rate": 0.00013509192756094472,
+      "loss": 0.3859,
+      "step": 5499
+    },
+    {
+      "epoch": 0.40444150305169496,
+      "grad_norm": 0.8389372229576111,
+      "learning_rate": 0.0001350696249563991,
+      "loss": 0.4037,
+      "step": 5500
+    },
+    {
+      "epoch": 0.40451503787043164,
+      "grad_norm": 0.9246091842651367,
+      "learning_rate": 0.0001350473203626674,
+      "loss": 0.3834,
+      "step": 5501
+    },
+    {
+      "epoch": 0.4045885726891683,
+      "grad_norm": 0.7200527191162109,
+      "learning_rate": 0.00013502501378101465,
+      "loss": 0.4168,
+      "step": 5502
+    },
+    {
+      "epoch": 0.404662107507905,
+      "grad_norm": 0.5296318531036377,
+      "learning_rate": 0.00013500270521270616,
+      "loss": 0.3753,
+      "step": 5503
+    },
+    {
+      "epoch": 0.40473564232664166,
+      "grad_norm": 0.9297668933868408,
+      "learning_rate": 0.00013498039465900728,
+      "loss": 0.407,
+      "step": 5504
+    },
+    {
+      "epoch": 0.40480917714537834,
+      "grad_norm": 1.3737620115280151,
+      "learning_rate": 0.00013495808212118354,
+      "loss": 0.3811,
+      "step": 5505
+    },
+    {
+      "epoch": 0.404882711964115,
+      "grad_norm": 0.5532109141349792,
+      "learning_rate": 0.00013493576760050048,
+      "loss": 0.3752,
+      "step": 5506
+    },
+    {
+      "epoch": 0.4049562467828517,
+      "grad_norm": 1.3633466958999634,
+      "learning_rate": 0.00013491345109822378,
+      "loss": 0.3829,
+      "step": 5507
+    },
+    {
+      "epoch": 0.40502978160158837,
+      "grad_norm": 0.6413983702659607,
+      "learning_rate": 0.00013489113261561935,
+      "loss": 0.3838,
+      "step": 5508
+    },
+    {
+      "epoch": 0.40510331642032504,
+      "grad_norm": 0.7873442769050598,
+      "learning_rate": 0.00013486881215395296,
+      "loss": 0.3897,
+      "step": 5509
+    },
+    {
+      "epoch": 0.4051768512390617,
+      "grad_norm": 1.1753053665161133,
+      "learning_rate": 0.00013484648971449082,
+      "loss": 0.3605,
+      "step": 5510
+    },
+    {
+      "epoch": 0.4052503860577984,
+      "grad_norm": 0.5503600835800171,
+      "learning_rate": 0.000134824165298499,
+      "loss": 0.4141,
+      "step": 5511
+    },
+    {
+      "epoch": 0.405323920876535,
+      "grad_norm": 1.4003998041152954,
+      "learning_rate": 0.0001348018389072438,
+      "loss": 0.3768,
+      "step": 5512
+    },
+    {
+      "epoch": 0.4053974556952717,
+      "grad_norm": 0.7370434999465942,
+      "learning_rate": 0.00013477951054199151,
+      "loss": 0.3818,
+      "step": 5513
+    },
+    {
+      "epoch": 0.40547099051400837,
+      "grad_norm": 0.8108879923820496,
+      "learning_rate": 0.00013475718020400872,
+      "loss": 0.3844,
+      "step": 5514
+    },
+    {
+      "epoch": 0.40554452533274504,
+      "grad_norm": 1.4062200784683228,
+      "learning_rate": 0.00013473484789456195,
+      "loss": 0.3902,
+      "step": 5515
+    },
+    {
+      "epoch": 0.4056180601514817,
+      "grad_norm": 0.5734351277351379,
+      "learning_rate": 0.00013471251361491798,
+      "loss": 0.3814,
+      "step": 5516
+    },
+    {
+      "epoch": 0.4056915949702184,
+      "grad_norm": 1.3216009140014648,
+      "learning_rate": 0.00013469017736634356,
+      "loss": 0.3752,
+      "step": 5517
+    },
+    {
+      "epoch": 0.40576512978895507,
+      "grad_norm": 0.6464775800704956,
+      "learning_rate": 0.00013466783915010573,
+      "loss": 0.3714,
+      "step": 5518
+    },
+    {
+      "epoch": 0.40583866460769175,
+      "grad_norm": 0.9327523112297058,
+      "learning_rate": 0.00013464549896747145,
+      "loss": 0.3624,
+      "step": 5519
+    },
+    {
+      "epoch": 0.4059121994264284,
+      "grad_norm": 1.4917923212051392,
+      "learning_rate": 0.00013462315681970788,
+      "loss": 0.3879,
+      "step": 5520
+    },
+    {
+      "epoch": 0.4059857342451651,
+      "grad_norm": 0.6493255496025085,
+      "learning_rate": 0.00013460081270808236,
+      "loss": 0.3544,
+      "step": 5521
+    },
+    {
+      "epoch": 0.4060592690639018,
+      "grad_norm": 1.2554900646209717,
+      "learning_rate": 0.00013457846663386216,
+      "loss": 0.3696,
+      "step": 5522
+    },
+    {
+      "epoch": 0.40613280388263845,
+      "grad_norm": 0.9180776476860046,
+      "learning_rate": 0.00013455611859831493,
+      "loss": 0.4074,
+      "step": 5523
+    },
+    {
+      "epoch": 0.4062063387013751,
+      "grad_norm": 0.8228355646133423,
+      "learning_rate": 0.00013453376860270817,
+      "loss": 0.3915,
+      "step": 5524
+    },
+    {
+      "epoch": 0.4062798735201118,
+      "grad_norm": 0.6121863722801208,
+      "learning_rate": 0.00013451141664830955,
+      "loss": 0.4026,
+      "step": 5525
+    },
+    {
+      "epoch": 0.4063534083388484,
+      "grad_norm": 0.7415707111358643,
+      "learning_rate": 0.00013448906273638703,
+      "loss": 0.3972,
+      "step": 5526
+    },
+    {
+      "epoch": 0.4064269431575851,
+      "grad_norm": 1.2030870914459229,
+      "learning_rate": 0.00013446670686820839,
+      "loss": 0.3646,
+      "step": 5527
+    },
+    {
+      "epoch": 0.4065004779763218,
+      "grad_norm": 0.5867236256599426,
+      "learning_rate": 0.00013444434904504181,
+      "loss": 0.4,
+      "step": 5528
+    },
+    {
+      "epoch": 0.40657401279505845,
+      "grad_norm": 1.2575464248657227,
+      "learning_rate": 0.00013442198926815542,
+      "loss": 0.3533,
+      "step": 5529
+    },
+    {
+      "epoch": 0.4066475476137951,
+      "grad_norm": 0.8823785781860352,
+      "learning_rate": 0.00013439962753881743,
+      "loss": 0.363,
+      "step": 5530
+    },
+    {
+      "epoch": 0.4067210824325318,
+      "grad_norm": 1.1425691843032837,
+      "learning_rate": 0.0001343772638582963,
+      "loss": 0.3515,
+      "step": 5531
+    },
+    {
+      "epoch": 0.4067946172512685,
+      "grad_norm": 0.6774272322654724,
+      "learning_rate": 0.00013435489822786044,
+      "loss": 0.364,
+      "step": 5532
+    },
+    {
+      "epoch": 0.40686815207000515,
+      "grad_norm": 0.8645538687705994,
+      "learning_rate": 0.00013433253064877853,
+      "loss": 0.3996,
+      "step": 5533
+    },
+    {
+      "epoch": 0.4069416868887418,
+      "grad_norm": 1.0310951471328735,
+      "learning_rate": 0.00013431016112231924,
+      "loss": 0.4042,
+      "step": 5534
+    },
+    {
+      "epoch": 0.4070152217074785,
+      "grad_norm": 0.8999442458152771,
+      "learning_rate": 0.0001342877896497514,
+      "loss": 0.3557,
+      "step": 5535
+    },
+    {
+      "epoch": 0.4070887565262152,
+      "grad_norm": 1.0101821422576904,
+      "learning_rate": 0.00013426541623234398,
+      "loss": 0.3576,
+      "step": 5536
+    },
+    {
+      "epoch": 0.40716229134495185,
+      "grad_norm": 0.6814073324203491,
+      "learning_rate": 0.00013424304087136598,
+      "loss": 0.378,
+      "step": 5537
+    },
+    {
+      "epoch": 0.40723582616368853,
+      "grad_norm": 0.8907278776168823,
+      "learning_rate": 0.00013422066356808652,
+      "loss": 0.3861,
+      "step": 5538
+    },
+    {
+      "epoch": 0.4073093609824252,
+      "grad_norm": 0.861304521560669,
+      "learning_rate": 0.00013419828432377492,
+      "loss": 0.3535,
+      "step": 5539
+    },
+    {
+      "epoch": 0.4073828958011618,
+      "grad_norm": 0.6980915665626526,
+      "learning_rate": 0.00013417590313970055,
+      "loss": 0.3677,
+      "step": 5540
+    },
+    {
+      "epoch": 0.4074564306198985,
+      "grad_norm": 1.2023663520812988,
+      "learning_rate": 0.00013415352001713287,
+      "loss": 0.4187,
+      "step": 5541
+    },
+    {
+      "epoch": 0.4075299654386352,
+      "grad_norm": 0.6089658141136169,
+      "learning_rate": 0.00013413113495734156,
+      "loss": 0.3825,
+      "step": 5542
+    },
+    {
+      "epoch": 0.40760350025737185,
+      "grad_norm": 1.2945150136947632,
+      "learning_rate": 0.0001341087479615962,
+      "loss": 0.3754,
+      "step": 5543
+    },
+    {
+      "epoch": 0.40767703507610853,
+      "grad_norm": 0.4846764802932739,
+      "learning_rate": 0.00013408635903116664,
+      "loss": 0.3739,
+      "step": 5544
+    },
+    {
+      "epoch": 0.4077505698948452,
+      "grad_norm": 1.0813872814178467,
+      "learning_rate": 0.00013406396816732286,
+      "loss": 0.3799,
+      "step": 5545
+    },
+    {
+      "epoch": 0.4078241047135819,
+      "grad_norm": 0.6827343702316284,
+      "learning_rate": 0.00013404157537133483,
+      "loss": 0.3629,
+      "step": 5546
+    },
+    {
+      "epoch": 0.40789763953231856,
+      "grad_norm": 0.7742222547531128,
+      "learning_rate": 0.00013401918064447277,
+      "loss": 0.3551,
+      "step": 5547
+    },
+    {
+      "epoch": 0.40797117435105523,
+      "grad_norm": 0.6611517071723938,
+      "learning_rate": 0.00013399678398800683,
+      "loss": 0.3934,
+      "step": 5548
+    },
+    {
+      "epoch": 0.4080447091697919,
+      "grad_norm": 0.7371267676353455,
+      "learning_rate": 0.00013397438540320742,
+      "loss": 0.3651,
+      "step": 5549
+    },
+    {
+      "epoch": 0.4081182439885286,
+      "grad_norm": 0.8696319460868835,
+      "learning_rate": 0.00013395198489134504,
+      "loss": 0.3577,
+      "step": 5550
+    },
+    {
+      "epoch": 0.40819177880726526,
+      "grad_norm": 0.6646525263786316,
+      "learning_rate": 0.00013392958245369026,
+      "loss": 0.3815,
+      "step": 5551
+    },
+    {
+      "epoch": 0.40826531362600194,
+      "grad_norm": 0.593494713306427,
+      "learning_rate": 0.0001339071780915137,
+      "loss": 0.4039,
+      "step": 5552
+    },
+    {
+      "epoch": 0.4083388484447386,
+      "grad_norm": 0.5768930912017822,
+      "learning_rate": 0.00013388477180608627,
+      "loss": 0.3981,
+      "step": 5553
+    },
+    {
+      "epoch": 0.40841238326347523,
+      "grad_norm": 0.7730190753936768,
+      "learning_rate": 0.00013386236359867883,
+      "loss": 0.399,
+      "step": 5554
+    },
+    {
+      "epoch": 0.4084859180822119,
+      "grad_norm": 0.6707442402839661,
+      "learning_rate": 0.00013383995347056234,
+      "loss": 0.3421,
+      "step": 5555
+    },
+    {
+      "epoch": 0.4085594529009486,
+      "grad_norm": 0.650966465473175,
+      "learning_rate": 0.000133817541423008,
+      "loss": 0.3845,
+      "step": 5556
+    },
+    {
+      "epoch": 0.40863298771968526,
+      "grad_norm": 0.6010043621063232,
+      "learning_rate": 0.00013379512745728705,
+      "loss": 0.3781,
+      "step": 5557
+    },
+    {
+      "epoch": 0.40870652253842193,
+      "grad_norm": 0.523661732673645,
+      "learning_rate": 0.00013377271157467076,
+      "loss": 0.3941,
+      "step": 5558
+    },
+    {
+      "epoch": 0.4087800573571586,
+      "grad_norm": 0.630453884601593,
+      "learning_rate": 0.00013375029377643067,
+      "loss": 0.358,
+      "step": 5559
+    },
+    {
+      "epoch": 0.4088535921758953,
+      "grad_norm": 0.5994718074798584,
+      "learning_rate": 0.00013372787406383828,
+      "loss": 0.3886,
+      "step": 5560
+    },
+    {
+      "epoch": 0.40892712699463196,
+      "grad_norm": 0.49538010358810425,
+      "learning_rate": 0.00013370545243816527,
+      "loss": 0.4031,
+      "step": 5561
+    },
+    {
+      "epoch": 0.40900066181336864,
+      "grad_norm": 0.695470929145813,
+      "learning_rate": 0.00013368302890068341,
+      "loss": 0.3863,
+      "step": 5562
+    },
+    {
+      "epoch": 0.4090741966321053,
+      "grad_norm": 0.880057156085968,
+      "learning_rate": 0.00013366060345266462,
+      "loss": 0.3946,
+      "step": 5563
+    },
+    {
+      "epoch": 0.409147731450842,
+      "grad_norm": 0.6406196355819702,
+      "learning_rate": 0.00013363817609538087,
+      "loss": 0.3927,
+      "step": 5564
+    },
+    {
+      "epoch": 0.40922126626957867,
+      "grad_norm": 0.52534419298172,
+      "learning_rate": 0.0001336157468301043,
+      "loss": 0.3738,
+      "step": 5565
+    },
+    {
+      "epoch": 0.40929480108831534,
+      "grad_norm": 0.7290903329849243,
+      "learning_rate": 0.0001335933156581071,
+      "loss": 0.385,
+      "step": 5566
+    },
+    {
+      "epoch": 0.409368335907052,
+      "grad_norm": 0.8880733251571655,
+      "learning_rate": 0.00013357088258066156,
+      "loss": 0.3754,
+      "step": 5567
+    },
+    {
+      "epoch": 0.40944187072578864,
+      "grad_norm": 0.8885848522186279,
+      "learning_rate": 0.0001335484475990401,
+      "loss": 0.3904,
+      "step": 5568
+    },
+    {
+      "epoch": 0.4095154055445253,
+      "grad_norm": 0.6851291656494141,
+      "learning_rate": 0.00013352601071451534,
+      "loss": 0.3483,
+      "step": 5569
+    },
+    {
+      "epoch": 0.409588940363262,
+      "grad_norm": 0.5234619975090027,
+      "learning_rate": 0.00013350357192835985,
+      "loss": 0.384,
+      "step": 5570
+    },
+    {
+      "epoch": 0.40966247518199866,
+      "grad_norm": 0.7325437068939209,
+      "learning_rate": 0.00013348113124184643,
+      "loss": 0.3784,
+      "step": 5571
+    },
+    {
+      "epoch": 0.40973601000073534,
+      "grad_norm": 1.2724902629852295,
+      "learning_rate": 0.00013345868865624787,
+      "loss": 0.4053,
+      "step": 5572
+    },
+    {
+      "epoch": 0.409809544819472,
+      "grad_norm": 0.8627475500106812,
+      "learning_rate": 0.00013343624417283718,
+      "loss": 0.413,
+      "step": 5573
+    },
+    {
+      "epoch": 0.4098830796382087,
+      "grad_norm": 0.7606956362724304,
+      "learning_rate": 0.00013341379779288745,
+      "loss": 0.3798,
+      "step": 5574
+    },
+    {
+      "epoch": 0.40995661445694537,
+      "grad_norm": 1.4024122953414917,
+      "learning_rate": 0.00013339134951767184,
+      "loss": 0.3845,
+      "step": 5575
+    },
+    {
+      "epoch": 0.41003014927568204,
+      "grad_norm": 0.8991518616676331,
+      "learning_rate": 0.00013336889934846365,
+      "loss": 0.3725,
+      "step": 5576
+    },
+    {
+      "epoch": 0.4101036840944187,
+      "grad_norm": 0.5967211127281189,
+      "learning_rate": 0.00013334644728653632,
+      "loss": 0.3803,
+      "step": 5577
+    },
+    {
+      "epoch": 0.4101772189131554,
+      "grad_norm": 1.0525503158569336,
+      "learning_rate": 0.00013332399333316323,
+      "loss": 0.3749,
+      "step": 5578
+    },
+    {
+      "epoch": 0.41025075373189207,
+      "grad_norm": 0.56478351354599,
+      "learning_rate": 0.0001333015374896181,
+      "loss": 0.3729,
+      "step": 5579
+    },
+    {
+      "epoch": 0.41032428855062875,
+      "grad_norm": 0.6359385251998901,
+      "learning_rate": 0.00013327907975717462,
+      "loss": 0.3954,
+      "step": 5580
+    },
+    {
+      "epoch": 0.4103978233693654,
+      "grad_norm": 0.6374857425689697,
+      "learning_rate": 0.0001332566201371066,
+      "loss": 0.3879,
+      "step": 5581
+    },
+    {
+      "epoch": 0.41047135818810204,
+      "grad_norm": 0.5369979739189148,
+      "learning_rate": 0.00013323415863068804,
+      "loss": 0.406,
+      "step": 5582
+    },
+    {
+      "epoch": 0.4105448930068387,
+      "grad_norm": 0.8913755416870117,
+      "learning_rate": 0.0001332116952391929,
+      "loss": 0.3807,
+      "step": 5583
+    },
+    {
+      "epoch": 0.4106184278255754,
+      "grad_norm": 1.212007761001587,
+      "learning_rate": 0.00013318922996389534,
+      "loss": 0.3671,
+      "step": 5584
+    },
+    {
+      "epoch": 0.41069196264431207,
+      "grad_norm": 0.5890079140663147,
+      "learning_rate": 0.00013316676280606963,
+      "loss": 0.3709,
+      "step": 5585
+    },
+    {
+      "epoch": 0.41076549746304875,
+      "grad_norm": 1.5305659770965576,
+      "learning_rate": 0.00013314429376699014,
+      "loss": 0.3985,
+      "step": 5586
+    },
+    {
+      "epoch": 0.4108390322817854,
+      "grad_norm": 0.6786748170852661,
+      "learning_rate": 0.00013312182284793133,
+      "loss": 0.3822,
+      "step": 5587
+    },
+    {
+      "epoch": 0.4109125671005221,
+      "grad_norm": 0.597059965133667,
+      "learning_rate": 0.0001330993500501678,
+      "loss": 0.386,
+      "step": 5588
+    },
+    {
+      "epoch": 0.4109861019192588,
+      "grad_norm": 0.928300678730011,
+      "learning_rate": 0.00013307687537497418,
+      "loss": 0.3936,
+      "step": 5589
+    },
+    {
+      "epoch": 0.41105963673799545,
+      "grad_norm": 1.1289492845535278,
+      "learning_rate": 0.00013305439882362527,
+      "loss": 0.3897,
+      "step": 5590
+    },
+    {
+      "epoch": 0.4111331715567321,
+      "grad_norm": 0.788494884967804,
+      "learning_rate": 0.00013303192039739597,
+      "loss": 0.3793,
+      "step": 5591
+    },
+    {
+      "epoch": 0.4112067063754688,
+      "grad_norm": 0.6005072593688965,
+      "learning_rate": 0.0001330094400975613,
+      "loss": 0.4018,
+      "step": 5592
+    },
+    {
+      "epoch": 0.4112802411942055,
+      "grad_norm": 0.5200372934341431,
+      "learning_rate": 0.00013298695792539637,
+      "loss": 0.4158,
+      "step": 5593
+    },
+    {
+      "epoch": 0.41135377601294215,
+      "grad_norm": 0.6208307147026062,
+      "learning_rate": 0.00013296447388217637,
+      "loss": 0.38,
+      "step": 5594
+    },
+    {
+      "epoch": 0.4114273108316788,
+      "grad_norm": 0.5547485947608948,
+      "learning_rate": 0.0001329419879691766,
+      "loss": 0.3832,
+      "step": 5595
+    },
+    {
+      "epoch": 0.41150084565041545,
+      "grad_norm": 0.5336465835571289,
+      "learning_rate": 0.0001329195001876725,
+      "loss": 0.3638,
+      "step": 5596
+    },
+    {
+      "epoch": 0.4115743804691521,
+      "grad_norm": 0.9332775473594666,
+      "learning_rate": 0.00013289701053893963,
+      "loss": 0.3871,
+      "step": 5597
+    },
+    {
+      "epoch": 0.4116479152878888,
+      "grad_norm": 1.0393710136413574,
+      "learning_rate": 0.00013287451902425362,
+      "loss": 0.3648,
+      "step": 5598
+    },
+    {
+      "epoch": 0.4117214501066255,
+      "grad_norm": 0.7740146517753601,
+      "learning_rate": 0.00013285202564489014,
+      "loss": 0.3881,
+      "step": 5599
+    },
+    {
+      "epoch": 0.41179498492536215,
+      "grad_norm": 0.5062613487243652,
+      "learning_rate": 0.00013282953040212516,
+      "loss": 0.3597,
+      "step": 5600
+    },
+    {
+      "epoch": 0.4118685197440988,
+      "grad_norm": 1.1534883975982666,
+      "learning_rate": 0.00013280703329723455,
+      "loss": 0.3884,
+      "step": 5601
+    },
+    {
+      "epoch": 0.4119420545628355,
+      "grad_norm": 1.280849814414978,
+      "learning_rate": 0.00013278453433149436,
+      "loss": 0.3602,
+      "step": 5602
+    },
+    {
+      "epoch": 0.4120155893815722,
+      "grad_norm": 0.8674570322036743,
+      "learning_rate": 0.00013276203350618082,
+      "loss": 0.3945,
+      "step": 5603
+    },
+    {
+      "epoch": 0.41208912420030885,
+      "grad_norm": 1.5452485084533691,
+      "learning_rate": 0.0001327395308225701,
+      "loss": 0.3906,
+      "step": 5604
+    },
+    {
+      "epoch": 0.41216265901904553,
+      "grad_norm": 0.922274112701416,
+      "learning_rate": 0.00013271702628193873,
+      "loss": 0.3893,
+      "step": 5605
+    },
+    {
+      "epoch": 0.4122361938377822,
+      "grad_norm": 0.8610065579414368,
+      "learning_rate": 0.00013269451988556308,
+      "loss": 0.3854,
+      "step": 5606
+    },
+    {
+      "epoch": 0.4123097286565189,
+      "grad_norm": 0.5877528190612793,
+      "learning_rate": 0.0001326720116347197,
+      "loss": 0.3873,
+      "step": 5607
+    },
+    {
+      "epoch": 0.41238326347525556,
+      "grad_norm": 0.8982342481613159,
+      "learning_rate": 0.00013264950153068537,
+      "loss": 0.3755,
+      "step": 5608
+    },
+    {
+      "epoch": 0.41245679829399223,
+      "grad_norm": 0.8647814393043518,
+      "learning_rate": 0.00013262698957473688,
+      "loss": 0.3694,
+      "step": 5609
+    },
+    {
+      "epoch": 0.41253033311272885,
+      "grad_norm": 0.7044786810874939,
+      "learning_rate": 0.0001326044757681511,
+      "loss": 0.3814,
+      "step": 5610
+    },
+    {
+      "epoch": 0.41260386793146553,
+      "grad_norm": 1.289236307144165,
+      "learning_rate": 0.00013258196011220504,
+      "loss": 0.3619,
+      "step": 5611
+    },
+    {
+      "epoch": 0.4126774027502022,
+      "grad_norm": 0.6586730480194092,
+      "learning_rate": 0.0001325594426081758,
+      "loss": 0.4115,
+      "step": 5612
+    },
+    {
+      "epoch": 0.4127509375689389,
+      "grad_norm": 1.151220440864563,
+      "learning_rate": 0.00013253692325734063,
+      "loss": 0.3599,
+      "step": 5613
+    },
+    {
+      "epoch": 0.41282447238767556,
+      "grad_norm": 1.4508426189422607,
+      "learning_rate": 0.00013251440206097682,
+      "loss": 0.3816,
+      "step": 5614
+    },
+    {
+      "epoch": 0.41289800720641223,
+      "grad_norm": 0.885415256023407,
+      "learning_rate": 0.00013249187902036183,
+      "loss": 0.3669,
+      "step": 5615
+    },
+    {
+      "epoch": 0.4129715420251489,
+      "grad_norm": 1.5464134216308594,
+      "learning_rate": 0.0001324693541367732,
+      "loss": 0.4031,
+      "step": 5616
+    },
+    {
+      "epoch": 0.4130450768438856,
+      "grad_norm": 0.8070924282073975,
+      "learning_rate": 0.00013244682741148846,
+      "loss": 0.3642,
+      "step": 5617
+    },
+    {
+      "epoch": 0.41311861166262226,
+      "grad_norm": 1.1138643026351929,
+      "learning_rate": 0.0001324242988457855,
+      "loss": 0.3731,
+      "step": 5618
+    },
+    {
+      "epoch": 0.41319214648135894,
+      "grad_norm": 1.08867609500885,
+      "learning_rate": 0.00013240176844094204,
+      "loss": 0.3793,
+      "step": 5619
+    },
+    {
+      "epoch": 0.4132656813000956,
+      "grad_norm": 0.658970057964325,
+      "learning_rate": 0.00013237923619823611,
+      "loss": 0.3846,
+      "step": 5620
+    },
+    {
+      "epoch": 0.4133392161188323,
+      "grad_norm": 0.9886105060577393,
+      "learning_rate": 0.00013235670211894575,
+      "loss": 0.3726,
+      "step": 5621
+    },
+    {
+      "epoch": 0.41341275093756896,
+      "grad_norm": 1.370779275894165,
+      "learning_rate": 0.00013233416620434904,
+      "loss": 0.3862,
+      "step": 5622
+    },
+    {
+      "epoch": 0.41348628575630564,
+      "grad_norm": 0.7936215996742249,
+      "learning_rate": 0.00013231162845572436,
+      "loss": 0.369,
+      "step": 5623
+    },
+    {
+      "epoch": 0.41355982057504226,
+      "grad_norm": 2.496429443359375,
+      "learning_rate": 0.00013228908887435,
+      "loss": 0.3782,
+      "step": 5624
+    },
+    {
+      "epoch": 0.41363335539377893,
+      "grad_norm": 1.8493728637695312,
+      "learning_rate": 0.0001322665474615044,
+      "loss": 0.3977,
+      "step": 5625
+    },
+    {
+      "epoch": 0.4137068902125156,
+      "grad_norm": 2.2860379219055176,
+      "learning_rate": 0.0001322440042184662,
+      "loss": 0.3826,
+      "step": 5626
+    },
+    {
+      "epoch": 0.4137804250312523,
+      "grad_norm": 2.4228787422180176,
+      "learning_rate": 0.00013222145914651404,
+      "loss": 0.3797,
+      "step": 5627
+    },
+    {
+      "epoch": 0.41385395984998896,
+      "grad_norm": 0.9125015139579773,
+      "learning_rate": 0.00013219891224692675,
+      "loss": 0.3717,
+      "step": 5628
+    },
+    {
+      "epoch": 0.41392749466872564,
+      "grad_norm": 1.6952743530273438,
+      "learning_rate": 0.00013217636352098316,
+      "loss": 0.3812,
+      "step": 5629
+    },
+    {
+      "epoch": 0.4140010294874623,
+      "grad_norm": 1.3378502130508423,
+      "learning_rate": 0.00013215381296996227,
+      "loss": 0.4116,
+      "step": 5630
+    },
+    {
+      "epoch": 0.414074564306199,
+      "grad_norm": 1.1687004566192627,
+      "learning_rate": 0.0001321312605951432,
+      "loss": 0.3855,
+      "step": 5631
+    },
+    {
+      "epoch": 0.41414809912493566,
+      "grad_norm": 1.369852066040039,
+      "learning_rate": 0.00013210870639780505,
+      "loss": 0.3992,
+      "step": 5632
+    },
+    {
+      "epoch": 0.41422163394367234,
+      "grad_norm": 0.9414073824882507,
+      "learning_rate": 0.00013208615037922725,
+      "loss": 0.3661,
+      "step": 5633
+    },
+    {
+      "epoch": 0.414295168762409,
+      "grad_norm": 1.8468393087387085,
+      "learning_rate": 0.00013206359254068912,
+      "loss": 0.381,
+      "step": 5634
+    },
+    {
+      "epoch": 0.4143687035811457,
+      "grad_norm": 0.6148869395256042,
+      "learning_rate": 0.00013204103288347018,
+      "loss": 0.388,
+      "step": 5635
+    },
+    {
+      "epoch": 0.41444223839988237,
+      "grad_norm": 1.7158936262130737,
+      "learning_rate": 0.00013201847140885007,
+      "loss": 0.3772,
+      "step": 5636
+    },
+    {
+      "epoch": 0.41451577321861904,
+      "grad_norm": 0.634723424911499,
+      "learning_rate": 0.00013199590811810842,
+      "loss": 0.4186,
+      "step": 5637
+    },
+    {
+      "epoch": 0.41458930803735566,
+      "grad_norm": 2.060837507247925,
+      "learning_rate": 0.00013197334301252514,
+      "loss": 0.3955,
+      "step": 5638
+    },
+    {
+      "epoch": 0.41466284285609234,
+      "grad_norm": 1.0538891553878784,
+      "learning_rate": 0.00013195077609338008,
+      "loss": 0.3915,
+      "step": 5639
+    },
+    {
+      "epoch": 0.414736377674829,
+      "grad_norm": 2.4635746479034424,
+      "learning_rate": 0.00013192820736195325,
+      "loss": 0.3812,
+      "step": 5640
+    },
+    {
+      "epoch": 0.4148099124935657,
+      "grad_norm": 2.620873212814331,
+      "learning_rate": 0.00013190563681952485,
+      "loss": 0.3861,
+      "step": 5641
+    },
+    {
+      "epoch": 0.41488344731230237,
+      "grad_norm": 0.7492923736572266,
+      "learning_rate": 0.00013188306446737503,
+      "loss": 0.375,
+      "step": 5642
+    },
+    {
+      "epoch": 0.41495698213103904,
+      "grad_norm": 1.408700942993164,
+      "learning_rate": 0.0001318604903067842,
+      "loss": 0.3399,
+      "step": 5643
+    },
+    {
+      "epoch": 0.4150305169497757,
+      "grad_norm": 1.006116509437561,
+      "learning_rate": 0.00013183791433903267,
+      "loss": 0.3989,
+      "step": 5644
+    },
+    {
+      "epoch": 0.4151040517685124,
+      "grad_norm": 0.5924782156944275,
+      "learning_rate": 0.00013181533656540106,
+      "loss": 0.3798,
+      "step": 5645
+    },
+    {
+      "epoch": 0.41517758658724907,
+      "grad_norm": 0.6952186226844788,
+      "learning_rate": 0.00013179275698716998,
+      "loss": 0.3654,
+      "step": 5646
+    },
+    {
+      "epoch": 0.41525112140598575,
+      "grad_norm": 0.9415836930274963,
+      "learning_rate": 0.0001317701756056202,
+      "loss": 0.3736,
+      "step": 5647
+    },
+    {
+      "epoch": 0.4153246562247224,
+      "grad_norm": 0.7312179803848267,
+      "learning_rate": 0.00013174759242203253,
+      "loss": 0.3868,
+      "step": 5648
+    },
+    {
+      "epoch": 0.4153981910434591,
+      "grad_norm": 0.5605160593986511,
+      "learning_rate": 0.00013172500743768788,
+      "loss": 0.3827,
+      "step": 5649
+    },
+    {
+      "epoch": 0.4154717258621958,
+      "grad_norm": 1.004045009613037,
+      "learning_rate": 0.00013170242065386732,
+      "loss": 0.4024,
+      "step": 5650
+    },
+    {
+      "epoch": 0.41554526068093245,
+      "grad_norm": 1.2914613485336304,
+      "learning_rate": 0.00013167983207185204,
+      "loss": 0.3553,
+      "step": 5651
+    },
+    {
+      "epoch": 0.41561879549966907,
+      "grad_norm": 0.6821404695510864,
+      "learning_rate": 0.00013165724169292325,
+      "loss": 0.401,
+      "step": 5652
+    },
+    {
+      "epoch": 0.41569233031840574,
+      "grad_norm": 1.3807564973831177,
+      "learning_rate": 0.0001316346495183623,
+      "loss": 0.3947,
+      "step": 5653
+    },
+    {
+      "epoch": 0.4157658651371424,
+      "grad_norm": 1.0215864181518555,
+      "learning_rate": 0.00013161205554945062,
+      "loss": 0.3984,
+      "step": 5654
+    },
+    {
+      "epoch": 0.4158393999558791,
+      "grad_norm": 0.7034276127815247,
+      "learning_rate": 0.0001315894597874698,
+      "loss": 0.3862,
+      "step": 5655
+    },
+    {
+      "epoch": 0.4159129347746158,
+      "grad_norm": 0.9996098875999451,
+      "learning_rate": 0.00013156686223370146,
+      "loss": 0.4121,
+      "step": 5656
+    },
+    {
+      "epoch": 0.41598646959335245,
+      "grad_norm": 0.8833569288253784,
+      "learning_rate": 0.00013154426288942742,
+      "loss": 0.3731,
+      "step": 5657
+    },
+    {
+      "epoch": 0.4160600044120891,
+      "grad_norm": 0.5697283148765564,
+      "learning_rate": 0.0001315216617559295,
+      "loss": 0.3877,
+      "step": 5658
+    },
+    {
+      "epoch": 0.4161335392308258,
+      "grad_norm": 1.2310130596160889,
+      "learning_rate": 0.00013149905883448962,
+      "loss": 0.3846,
+      "step": 5659
+    },
+    {
+      "epoch": 0.4162070740495625,
+      "grad_norm": 0.9913083910942078,
+      "learning_rate": 0.00013147645412638992,
+      "loss": 0.3766,
+      "step": 5660
+    },
+    {
+      "epoch": 0.41628060886829915,
+      "grad_norm": 0.779834508895874,
+      "learning_rate": 0.0001314538476329125,
+      "loss": 0.3646,
+      "step": 5661
+    },
+    {
+      "epoch": 0.4163541436870358,
+      "grad_norm": 0.6116977334022522,
+      "learning_rate": 0.00013143123935533966,
+      "loss": 0.3698,
+      "step": 5662
+    },
+    {
+      "epoch": 0.4164276785057725,
+      "grad_norm": 1.10954749584198,
+      "learning_rate": 0.00013140862929495372,
+      "loss": 0.4081,
+      "step": 5663
+    },
+    {
+      "epoch": 0.4165012133245092,
+      "grad_norm": 0.8836270570755005,
+      "learning_rate": 0.00013138601745303724,
+      "loss": 0.3703,
+      "step": 5664
+    },
+    {
+      "epoch": 0.41657474814324585,
+      "grad_norm": 0.6261916756629944,
+      "learning_rate": 0.0001313634038308727,
+      "loss": 0.3755,
+      "step": 5665
+    },
+    {
+      "epoch": 0.4166482829619825,
+      "grad_norm": 0.6503130197525024,
+      "learning_rate": 0.00013134078842974277,
+      "loss": 0.373,
+      "step": 5666
+    },
+    {
+      "epoch": 0.41672181778071915,
+      "grad_norm": 0.7114446759223938,
+      "learning_rate": 0.00013131817125093027,
+      "loss": 0.3905,
+      "step": 5667
+    },
+    {
+      "epoch": 0.4167953525994558,
+      "grad_norm": 0.8573969006538391,
+      "learning_rate": 0.00013129555229571803,
+      "loss": 0.3797,
+      "step": 5668
+    },
+    {
+      "epoch": 0.4168688874181925,
+      "grad_norm": 0.5771171450614929,
+      "learning_rate": 0.00013127293156538905,
+      "loss": 0.3704,
+      "step": 5669
+    },
+    {
+      "epoch": 0.4169424222369292,
+      "grad_norm": 0.5313810110092163,
+      "learning_rate": 0.0001312503090612264,
+      "loss": 0.36,
+      "step": 5670
+    },
+    {
+      "epoch": 0.41701595705566585,
+      "grad_norm": 0.6099579334259033,
+      "learning_rate": 0.00013122768478451323,
+      "loss": 0.3676,
+      "step": 5671
+    },
+    {
+      "epoch": 0.41708949187440253,
+      "grad_norm": 0.6785843968391418,
+      "learning_rate": 0.00013120505873653282,
+      "loss": 0.3732,
+      "step": 5672
+    },
+    {
+      "epoch": 0.4171630266931392,
+      "grad_norm": 0.530133068561554,
+      "learning_rate": 0.00013118243091856852,
+      "loss": 0.3709,
+      "step": 5673
+    },
+    {
+      "epoch": 0.4172365615118759,
+      "grad_norm": 0.9941043853759766,
+      "learning_rate": 0.00013115980133190385,
+      "loss": 0.3983,
+      "step": 5674
+    },
+    {
+      "epoch": 0.41731009633061256,
+      "grad_norm": 1.019351840019226,
+      "learning_rate": 0.00013113716997782237,
+      "loss": 0.4054,
+      "step": 5675
+    },
+    {
+      "epoch": 0.41738363114934923,
+      "grad_norm": 0.5160693526268005,
+      "learning_rate": 0.00013111453685760776,
+      "loss": 0.3849,
+      "step": 5676
+    },
+    {
+      "epoch": 0.4174571659680859,
+      "grad_norm": 1.3396799564361572,
+      "learning_rate": 0.00013109190197254378,
+      "loss": 0.3628,
+      "step": 5677
+    },
+    {
+      "epoch": 0.4175307007868226,
+      "grad_norm": 1.159285545349121,
+      "learning_rate": 0.00013106926532391428,
+      "loss": 0.4249,
+      "step": 5678
+    },
+    {
+      "epoch": 0.41760423560555926,
+      "grad_norm": 0.5795080065727234,
+      "learning_rate": 0.00013104662691300329,
+      "loss": 0.3768,
+      "step": 5679
+    },
+    {
+      "epoch": 0.4176777704242959,
+      "grad_norm": 1.4270164966583252,
+      "learning_rate": 0.00013102398674109483,
+      "loss": 0.3945,
+      "step": 5680
+    },
+    {
+      "epoch": 0.41775130524303256,
+      "grad_norm": 0.6865320205688477,
+      "learning_rate": 0.00013100134480947313,
+      "loss": 0.3619,
+      "step": 5681
+    },
+    {
+      "epoch": 0.41782484006176923,
+      "grad_norm": 0.7482091188430786,
+      "learning_rate": 0.00013097870111942244,
+      "loss": 0.3732,
+      "step": 5682
+    },
+    {
+      "epoch": 0.4178983748805059,
+      "grad_norm": 1.3896028995513916,
+      "learning_rate": 0.00013095605567222708,
+      "loss": 0.3894,
+      "step": 5683
+    },
+    {
+      "epoch": 0.4179719096992426,
+      "grad_norm": 0.49496397376060486,
+      "learning_rate": 0.00013093340846917161,
+      "loss": 0.3716,
+      "step": 5684
+    },
+    {
+      "epoch": 0.41804544451797926,
+      "grad_norm": 1.2402071952819824,
+      "learning_rate": 0.00013091075951154058,
+      "loss": 0.371,
+      "step": 5685
+    },
+    {
+      "epoch": 0.41811897933671593,
+      "grad_norm": 1.0053436756134033,
+      "learning_rate": 0.00013088810880061862,
+      "loss": 0.3659,
+      "step": 5686
+    },
+    {
+      "epoch": 0.4181925141554526,
+      "grad_norm": 0.6414778828620911,
+      "learning_rate": 0.00013086545633769054,
+      "loss": 0.3705,
+      "step": 5687
+    },
+    {
+      "epoch": 0.4182660489741893,
+      "grad_norm": 1.602319359779358,
+      "learning_rate": 0.00013084280212404123,
+      "loss": 0.3662,
+      "step": 5688
+    },
+    {
+      "epoch": 0.41833958379292596,
+      "grad_norm": 1.1224435567855835,
+      "learning_rate": 0.0001308201461609556,
+      "loss": 0.3886,
+      "step": 5689
+    },
+    {
+      "epoch": 0.41841311861166264,
+      "grad_norm": 1.1135684251785278,
+      "learning_rate": 0.00013079748844971876,
+      "loss": 0.3765,
+      "step": 5690
+    },
+    {
+      "epoch": 0.4184866534303993,
+      "grad_norm": 1.049216389656067,
+      "learning_rate": 0.00013077482899161586,
+      "loss": 0.3519,
+      "step": 5691
+    },
+    {
+      "epoch": 0.418560188249136,
+      "grad_norm": 1.135920524597168,
+      "learning_rate": 0.00013075216778793221,
+      "loss": 0.3657,
+      "step": 5692
+    },
+    {
+      "epoch": 0.41863372306787267,
+      "grad_norm": 0.7514716982841492,
+      "learning_rate": 0.00013072950483995315,
+      "loss": 0.3683,
+      "step": 5693
+    },
+    {
+      "epoch": 0.4187072578866093,
+      "grad_norm": 1.0777075290679932,
+      "learning_rate": 0.00013070684014896412,
+      "loss": 0.3649,
+      "step": 5694
+    },
+    {
+      "epoch": 0.41878079270534596,
+      "grad_norm": 0.617273211479187,
+      "learning_rate": 0.00013068417371625072,
+      "loss": 0.3675,
+      "step": 5695
+    },
+    {
+      "epoch": 0.41885432752408264,
+      "grad_norm": 0.8675044178962708,
+      "learning_rate": 0.00013066150554309858,
+      "loss": 0.4098,
+      "step": 5696
+    },
+    {
+      "epoch": 0.4189278623428193,
+      "grad_norm": 0.8890429139137268,
+      "learning_rate": 0.00013063883563079354,
+      "loss": 0.3938,
+      "step": 5697
+    },
+    {
+      "epoch": 0.419001397161556,
+      "grad_norm": 1.034532070159912,
+      "learning_rate": 0.00013061616398062138,
+      "loss": 0.372,
+      "step": 5698
+    },
+    {
+      "epoch": 0.41907493198029266,
+      "grad_norm": 0.6683034896850586,
+      "learning_rate": 0.0001305934905938681,
+      "loss": 0.4062,
+      "step": 5699
+    },
+    {
+      "epoch": 0.41914846679902934,
+      "grad_norm": 0.7009810209274292,
+      "learning_rate": 0.00013057081547181973,
+      "loss": 0.3857,
+      "step": 5700
+    },
+    {
+      "epoch": 0.419222001617766,
+      "grad_norm": 0.8212664723396301,
+      "learning_rate": 0.00013054813861576245,
+      "loss": 0.3853,
+      "step": 5701
+    },
+    {
+      "epoch": 0.4192955364365027,
+      "grad_norm": 0.9663318395614624,
+      "learning_rate": 0.0001305254600269825,
+      "loss": 0.4294,
+      "step": 5702
+    },
+    {
+      "epoch": 0.41936907125523937,
+      "grad_norm": 0.9116215109825134,
+      "learning_rate": 0.00013050277970676625,
+      "loss": 0.378,
+      "step": 5703
+    },
+    {
+      "epoch": 0.41944260607397604,
+      "grad_norm": 0.9498606324195862,
+      "learning_rate": 0.00013048009765640015,
+      "loss": 0.387,
+      "step": 5704
+    },
+    {
+      "epoch": 0.4195161408927127,
+      "grad_norm": 0.9058437347412109,
+      "learning_rate": 0.00013045741387717076,
+      "loss": 0.4155,
+      "step": 5705
+    },
+    {
+      "epoch": 0.4195896757114494,
+      "grad_norm": 0.604651689529419,
+      "learning_rate": 0.0001304347283703647,
+      "loss": 0.4031,
+      "step": 5706
+    },
+    {
+      "epoch": 0.41966321053018607,
+      "grad_norm": 0.8440231084823608,
+      "learning_rate": 0.0001304120411372687,
+      "loss": 0.3907,
+      "step": 5707
+    },
+    {
+      "epoch": 0.4197367453489227,
+      "grad_norm": 1.271134376525879,
+      "learning_rate": 0.00013038935217916966,
+      "loss": 0.4023,
+      "step": 5708
+    },
+    {
+      "epoch": 0.41981028016765937,
+      "grad_norm": 0.6118009686470032,
+      "learning_rate": 0.00013036666149735448,
+      "loss": 0.392,
+      "step": 5709
+    },
+    {
+      "epoch": 0.41988381498639604,
+      "grad_norm": 0.5815375447273254,
+      "learning_rate": 0.00013034396909311022,
+      "loss": 0.3819,
+      "step": 5710
+    },
+    {
+      "epoch": 0.4199573498051327,
+      "grad_norm": 0.6046547889709473,
+      "learning_rate": 0.00013032127496772405,
+      "loss": 0.3734,
+      "step": 5711
+    },
+    {
+      "epoch": 0.4200308846238694,
+      "grad_norm": 0.8129457235336304,
+      "learning_rate": 0.0001302985791224831,
+      "loss": 0.3693,
+      "step": 5712
+    },
+    {
+      "epoch": 0.42010441944260607,
+      "grad_norm": 0.6200842261314392,
+      "learning_rate": 0.00013027588155867478,
+      "loss": 0.3725,
+      "step": 5713
+    },
+    {
+      "epoch": 0.42017795426134275,
+      "grad_norm": 0.8321853876113892,
+      "learning_rate": 0.00013025318227758649,
+      "loss": 0.417,
+      "step": 5714
+    },
+    {
+      "epoch": 0.4202514890800794,
+      "grad_norm": 1.034400224685669,
+      "learning_rate": 0.00013023048128050582,
+      "loss": 0.3788,
+      "step": 5715
+    },
+    {
+      "epoch": 0.4203250238988161,
+      "grad_norm": 1.026525616645813,
+      "learning_rate": 0.00013020777856872034,
+      "loss": 0.4048,
+      "step": 5716
+    },
+    {
+      "epoch": 0.4203985587175528,
+      "grad_norm": 0.5964288115501404,
+      "learning_rate": 0.00013018507414351777,
+      "loss": 0.4097,
+      "step": 5717
+    },
+    {
+      "epoch": 0.42047209353628945,
+      "grad_norm": 0.5327749848365784,
+      "learning_rate": 0.00013016236800618593,
+      "loss": 0.3868,
+      "step": 5718
+    },
+    {
+      "epoch": 0.4205456283550261,
+      "grad_norm": 0.5167765617370605,
+      "learning_rate": 0.00013013966015801276,
+      "loss": 0.3902,
+      "step": 5719
+    },
+    {
+      "epoch": 0.4206191631737628,
+      "grad_norm": 0.6252542734146118,
+      "learning_rate": 0.00013011695060028623,
+      "loss": 0.3814,
+      "step": 5720
+    },
+    {
+      "epoch": 0.4206926979924995,
+      "grad_norm": 0.7856331467628479,
+      "learning_rate": 0.0001300942393342945,
+      "loss": 0.368,
+      "step": 5721
+    },
+    {
+      "epoch": 0.4207662328112361,
+      "grad_norm": 0.6552218794822693,
+      "learning_rate": 0.00013007152636132575,
+      "loss": 0.3856,
+      "step": 5722
+    },
+    {
+      "epoch": 0.42083976762997277,
+      "grad_norm": 0.6096869111061096,
+      "learning_rate": 0.0001300488116826683,
+      "loss": 0.3973,
+      "step": 5723
+    },
+    {
+      "epoch": 0.42091330244870945,
+      "grad_norm": 0.5448920726776123,
+      "learning_rate": 0.00013002609529961053,
+      "loss": 0.394,
+      "step": 5724
+    },
+    {
+      "epoch": 0.4209868372674461,
+      "grad_norm": 0.6078454852104187,
+      "learning_rate": 0.00013000337721344097,
+      "loss": 0.3848,
+      "step": 5725
+    },
+    {
+      "epoch": 0.4210603720861828,
+      "grad_norm": 0.6443542242050171,
+      "learning_rate": 0.00012998065742544817,
+      "loss": 0.4091,
+      "step": 5726
+    },
+    {
+      "epoch": 0.4211339069049195,
+      "grad_norm": 0.5145369172096252,
+      "learning_rate": 0.00012995793593692084,
+      "loss": 0.4033,
+      "step": 5727
+    },
+    {
+      "epoch": 0.42120744172365615,
+      "grad_norm": 0.8420312404632568,
+      "learning_rate": 0.00012993521274914775,
+      "loss": 0.3909,
+      "step": 5728
+    },
+    {
+      "epoch": 0.4212809765423928,
+      "grad_norm": 0.9483248591423035,
+      "learning_rate": 0.00012991248786341786,
+      "loss": 0.3915,
+      "step": 5729
+    },
+    {
+      "epoch": 0.4213545113611295,
+      "grad_norm": 1.560941219329834,
+      "learning_rate": 0.00012988976128102002,
+      "loss": 0.4039,
+      "step": 5730
+    },
+    {
+      "epoch": 0.4214280461798662,
+      "grad_norm": 0.6341535449028015,
+      "learning_rate": 0.0001298670330032434,
+      "loss": 0.4053,
+      "step": 5731
+    },
+    {
+      "epoch": 0.42150158099860285,
+      "grad_norm": 0.6830475330352783,
+      "learning_rate": 0.00012984430303137714,
+      "loss": 0.3985,
+      "step": 5732
+    },
+    {
+      "epoch": 0.42157511581733953,
+      "grad_norm": 1.0064412355422974,
+      "learning_rate": 0.00012982157136671053,
+      "loss": 0.3978,
+      "step": 5733
+    },
+    {
+      "epoch": 0.4216486506360762,
+      "grad_norm": 0.748592734336853,
+      "learning_rate": 0.00012979883801053296,
+      "loss": 0.3932,
+      "step": 5734
+    },
+    {
+      "epoch": 0.4217221854548129,
+      "grad_norm": 0.6349329352378845,
+      "learning_rate": 0.0001297761029641338,
+      "loss": 0.3942,
+      "step": 5735
+    },
+    {
+      "epoch": 0.4217957202735495,
+      "grad_norm": 1.2099114656448364,
+      "learning_rate": 0.00012975336622880268,
+      "loss": 0.4363,
+      "step": 5736
+    },
+    {
+      "epoch": 0.4218692550922862,
+      "grad_norm": 0.976640522480011,
+      "learning_rate": 0.00012973062780582918,
+      "loss": 0.3943,
+      "step": 5737
+    },
+    {
+      "epoch": 0.42194278991102285,
+      "grad_norm": 0.7390783429145813,
+      "learning_rate": 0.00012970788769650315,
+      "loss": 0.3956,
+      "step": 5738
+    },
+    {
+      "epoch": 0.42201632472975953,
+      "grad_norm": 0.4710020124912262,
+      "learning_rate": 0.00012968514590211435,
+      "loss": 0.401,
+      "step": 5739
+    },
+    {
+      "epoch": 0.4220898595484962,
+      "grad_norm": 0.6993487477302551,
+      "learning_rate": 0.00012966240242395278,
+      "loss": 0.3958,
+      "step": 5740
+    },
+    {
+      "epoch": 0.4221633943672329,
+      "grad_norm": 0.905491292476654,
+      "learning_rate": 0.0001296396572633084,
+      "loss": 0.3754,
+      "step": 5741
+    },
+    {
+      "epoch": 0.42223692918596956,
+      "grad_norm": 0.8790413737297058,
+      "learning_rate": 0.00012961691042147138,
+      "loss": 0.35,
+      "step": 5742
+    },
+    {
+      "epoch": 0.42231046400470623,
+      "grad_norm": 0.8146606683731079,
+      "learning_rate": 0.00012959416189973198,
+      "loss": 0.4062,
+      "step": 5743
+    },
+    {
+      "epoch": 0.4223839988234429,
+      "grad_norm": 0.8076046109199524,
+      "learning_rate": 0.00012957141169938046,
+      "loss": 0.3873,
+      "step": 5744
+    },
+    {
+      "epoch": 0.4224575336421796,
+      "grad_norm": 1.4110077619552612,
+      "learning_rate": 0.00012954865982170725,
+      "loss": 0.3637,
+      "step": 5745
+    },
+    {
+      "epoch": 0.42253106846091626,
+      "grad_norm": 0.5844987630844116,
+      "learning_rate": 0.0001295259062680029,
+      "loss": 0.371,
+      "step": 5746
+    },
+    {
+      "epoch": 0.42260460327965294,
+      "grad_norm": 1.176565408706665,
+      "learning_rate": 0.00012950315103955797,
+      "loss": 0.3827,
+      "step": 5747
+    },
+    {
+      "epoch": 0.4226781380983896,
+      "grad_norm": 1.0575135946273804,
+      "learning_rate": 0.00012948039413766317,
+      "loss": 0.4146,
+      "step": 5748
+    },
+    {
+      "epoch": 0.4227516729171263,
+      "grad_norm": 0.510694682598114,
+      "learning_rate": 0.00012945763556360931,
+      "loss": 0.4216,
+      "step": 5749
+    },
+    {
+      "epoch": 0.4228252077358629,
+      "grad_norm": 1.080812692642212,
+      "learning_rate": 0.00012943487531868726,
+      "loss": 0.3767,
+      "step": 5750
+    },
+    {
+      "epoch": 0.4228987425545996,
+      "grad_norm": 1.186002254486084,
+      "learning_rate": 0.00012941211340418803,
+      "loss": 0.4171,
+      "step": 5751
+    },
+    {
+      "epoch": 0.42297227737333626,
+      "grad_norm": 0.5007926821708679,
+      "learning_rate": 0.00012938934982140274,
+      "loss": 0.4023,
+      "step": 5752
+    },
+    {
+      "epoch": 0.42304581219207293,
+      "grad_norm": 0.7209011316299438,
+      "learning_rate": 0.00012936658457162245,
+      "loss": 0.41,
+      "step": 5753
+    },
+    {
+      "epoch": 0.4231193470108096,
+      "grad_norm": 0.8972499370574951,
+      "learning_rate": 0.0001293438176561385,
+      "loss": 0.4198,
+      "step": 5754
+    },
+    {
+      "epoch": 0.4231928818295463,
+      "grad_norm": 0.7526555061340332,
+      "learning_rate": 0.00012932104907624226,
+      "loss": 0.3865,
+      "step": 5755
+    },
+    {
+      "epoch": 0.42326641664828296,
+      "grad_norm": 0.5499320030212402,
+      "learning_rate": 0.00012929827883322514,
+      "loss": 0.3686,
+      "step": 5756
+    },
+    {
+      "epoch": 0.42333995146701964,
+      "grad_norm": 0.7455468773841858,
+      "learning_rate": 0.00012927550692837876,
+      "loss": 0.4064,
+      "step": 5757
+    },
+    {
+      "epoch": 0.4234134862857563,
+      "grad_norm": 0.9450761079788208,
+      "learning_rate": 0.00012925273336299474,
+      "loss": 0.4046,
+      "step": 5758
+    },
+    {
+      "epoch": 0.423487021104493,
+      "grad_norm": 0.8606468439102173,
+      "learning_rate": 0.00012922995813836478,
+      "loss": 0.3609,
+      "step": 5759
+    },
+    {
+      "epoch": 0.42356055592322966,
+      "grad_norm": 0.7722803354263306,
+      "learning_rate": 0.00012920718125578073,
+      "loss": 0.3612,
+      "step": 5760
+    },
+    {
+      "epoch": 0.42363409074196634,
+      "grad_norm": 0.6637990474700928,
+      "learning_rate": 0.0001291844027165346,
+      "loss": 0.3831,
+      "step": 5761
+    },
+    {
+      "epoch": 0.423707625560703,
+      "grad_norm": 0.5735304355621338,
+      "learning_rate": 0.00012916162252191827,
+      "loss": 0.3741,
+      "step": 5762
+    },
+    {
+      "epoch": 0.4237811603794397,
+      "grad_norm": 0.6215642094612122,
+      "learning_rate": 0.00012913884067322398,
+      "loss": 0.4078,
+      "step": 5763
+    },
+    {
+      "epoch": 0.4238546951981763,
+      "grad_norm": 0.9986411333084106,
+      "learning_rate": 0.0001291160571717439,
+      "loss": 0.3842,
+      "step": 5764
+    },
+    {
+      "epoch": 0.423928230016913,
+      "grad_norm": 1.0530990362167358,
+      "learning_rate": 0.00012909327201877032,
+      "loss": 0.3678,
+      "step": 5765
+    },
+    {
+      "epoch": 0.42400176483564966,
+      "grad_norm": 0.5584278106689453,
+      "learning_rate": 0.00012907048521559566,
+      "loss": 0.4094,
+      "step": 5766
+    },
+    {
+      "epoch": 0.42407529965438634,
+      "grad_norm": 0.6446212530136108,
+      "learning_rate": 0.0001290476967635124,
+      "loss": 0.3854,
+      "step": 5767
+    },
+    {
+      "epoch": 0.424148834473123,
+      "grad_norm": 0.7615039944648743,
+      "learning_rate": 0.00012902490666381308,
+      "loss": 0.4038,
+      "step": 5768
+    },
+    {
+      "epoch": 0.4242223692918597,
+      "grad_norm": 0.6883451342582703,
+      "learning_rate": 0.00012900211491779048,
+      "loss": 0.39,
+      "step": 5769
+    },
+    {
+      "epoch": 0.42429590411059637,
+      "grad_norm": 0.5260112881660461,
+      "learning_rate": 0.0001289793215267373,
+      "loss": 0.402,
+      "step": 5770
+    },
+    {
+      "epoch": 0.42436943892933304,
+      "grad_norm": 0.5529025197029114,
+      "learning_rate": 0.0001289565264919464,
+      "loss": 0.3975,
+      "step": 5771
+    },
+    {
+      "epoch": 0.4244429737480697,
+      "grad_norm": 0.6275097131729126,
+      "learning_rate": 0.00012893372981471077,
+      "loss": 0.3834,
+      "step": 5772
+    },
+    {
+      "epoch": 0.4245165085668064,
+      "grad_norm": 0.6903419494628906,
+      "learning_rate": 0.00012891093149632342,
+      "loss": 0.4326,
+      "step": 5773
+    },
+    {
+      "epoch": 0.42459004338554307,
+      "grad_norm": 0.5181411504745483,
+      "learning_rate": 0.00012888813153807757,
+      "loss": 0.403,
+      "step": 5774
+    },
+    {
+      "epoch": 0.42466357820427975,
+      "grad_norm": 0.8082814812660217,
+      "learning_rate": 0.0001288653299412664,
+      "loss": 0.412,
+      "step": 5775
+    },
+    {
+      "epoch": 0.4247371130230164,
+      "grad_norm": 1.0764414072036743,
+      "learning_rate": 0.00012884252670718324,
+      "loss": 0.3954,
+      "step": 5776
+    },
+    {
+      "epoch": 0.4248106478417531,
+      "grad_norm": 0.9594189524650574,
+      "learning_rate": 0.00012881972183712156,
+      "loss": 0.402,
+      "step": 5777
+    },
+    {
+      "epoch": 0.4248841826604897,
+      "grad_norm": 0.7062293887138367,
+      "learning_rate": 0.0001287969153323748,
+      "loss": 0.3944,
+      "step": 5778
+    },
+    {
+      "epoch": 0.4249577174792264,
+      "grad_norm": 1.6993083953857422,
+      "learning_rate": 0.00012877410719423663,
+      "loss": 0.3718,
+      "step": 5779
+    },
+    {
+      "epoch": 0.42503125229796307,
+      "grad_norm": 0.7299807667732239,
+      "learning_rate": 0.00012875129742400074,
+      "loss": 0.3783,
+      "step": 5780
+    },
+    {
+      "epoch": 0.42510478711669974,
+      "grad_norm": 1.0449832677841187,
+      "learning_rate": 0.0001287284860229609,
+      "loss": 0.3908,
+      "step": 5781
+    },
+    {
+      "epoch": 0.4251783219354364,
+      "grad_norm": 1.513929843902588,
+      "learning_rate": 0.00012870567299241102,
+      "loss": 0.3761,
+      "step": 5782
+    },
+    {
+      "epoch": 0.4252518567541731,
+      "grad_norm": 0.8666626811027527,
+      "learning_rate": 0.00012868285833364507,
+      "loss": 0.3997,
+      "step": 5783
+    },
+    {
+      "epoch": 0.4253253915729098,
+      "grad_norm": 1.473301887512207,
+      "learning_rate": 0.00012866004204795716,
+      "loss": 0.4068,
+      "step": 5784
+    },
+    {
+      "epoch": 0.42539892639164645,
+      "grad_norm": 1.0973891019821167,
+      "learning_rate": 0.0001286372241366414,
+      "loss": 0.3731,
+      "step": 5785
+    },
+    {
+      "epoch": 0.4254724612103831,
+      "grad_norm": 0.6871383786201477,
+      "learning_rate": 0.00012861440460099205,
+      "loss": 0.3699,
+      "step": 5786
+    },
+    {
+      "epoch": 0.4255459960291198,
+      "grad_norm": 1.4036836624145508,
+      "learning_rate": 0.00012859158344230348,
+      "loss": 0.3762,
+      "step": 5787
+    },
+    {
+      "epoch": 0.4256195308478565,
+      "grad_norm": 0.6041257977485657,
+      "learning_rate": 0.0001285687606618701,
+      "loss": 0.3693,
+      "step": 5788
+    },
+    {
+      "epoch": 0.42569306566659315,
+      "grad_norm": 1.924532413482666,
+      "learning_rate": 0.00012854593626098653,
+      "loss": 0.4116,
+      "step": 5789
+    },
+    {
+      "epoch": 0.4257666004853298,
+      "grad_norm": 0.7644591927528381,
+      "learning_rate": 0.00012852311024094727,
+      "loss": 0.3912,
+      "step": 5790
+    },
+    {
+      "epoch": 0.4258401353040665,
+      "grad_norm": 0.9224348664283752,
+      "learning_rate": 0.0001285002826030471,
+      "loss": 0.3956,
+      "step": 5791
+    },
+    {
+      "epoch": 0.4259136701228031,
+      "grad_norm": 1.050727367401123,
+      "learning_rate": 0.00012847745334858087,
+      "loss": 0.4,
+      "step": 5792
+    },
+    {
+      "epoch": 0.4259872049415398,
+      "grad_norm": 0.5792177319526672,
+      "learning_rate": 0.0001284546224788434,
+      "loss": 0.3789,
+      "step": 5793
+    },
+    {
+      "epoch": 0.4260607397602765,
+      "grad_norm": 1.0291272401809692,
+      "learning_rate": 0.0001284317899951297,
+      "loss": 0.4038,
+      "step": 5794
+    },
+    {
+      "epoch": 0.42613427457901315,
+      "grad_norm": 1.7161177396774292,
+      "learning_rate": 0.00012840895589873486,
+      "loss": 0.4001,
+      "step": 5795
+    },
+    {
+      "epoch": 0.4262078093977498,
+      "grad_norm": 1.1370793581008911,
+      "learning_rate": 0.00012838612019095409,
+      "loss": 0.3977,
+      "step": 5796
+    },
+    {
+      "epoch": 0.4262813442164865,
+      "grad_norm": 1.7436190843582153,
+      "learning_rate": 0.00012836328287308261,
+      "loss": 0.4164,
+      "step": 5797
+    },
+    {
+      "epoch": 0.4263548790352232,
+      "grad_norm": 1.6523699760437012,
+      "learning_rate": 0.0001283404439464158,
+      "loss": 0.3917,
+      "step": 5798
+    },
+    {
+      "epoch": 0.42642841385395985,
+      "grad_norm": 1.3603549003601074,
+      "learning_rate": 0.00012831760341224912,
+      "loss": 0.3965,
+      "step": 5799
+    },
+    {
+      "epoch": 0.42650194867269653,
+      "grad_norm": 2.5799033641815186,
+      "learning_rate": 0.00012829476127187807,
+      "loss": 0.415,
+      "step": 5800
+    },
+    {
+      "epoch": 0.4265754834914332,
+      "grad_norm": 1.9075422286987305,
+      "learning_rate": 0.0001282719175265983,
+      "loss": 0.3942,
+      "step": 5801
+    },
+    {
+      "epoch": 0.4266490183101699,
+      "grad_norm": 1.878692626953125,
+      "learning_rate": 0.00012824907217770553,
+      "loss": 0.3831,
+      "step": 5802
+    },
+    {
+      "epoch": 0.42672255312890656,
+      "grad_norm": 1.4000884294509888,
+      "learning_rate": 0.00012822622522649558,
+      "loss": 0.3876,
+      "step": 5803
+    },
+    {
+      "epoch": 0.42679608794764323,
+      "grad_norm": 1.3346909284591675,
+      "learning_rate": 0.0001282033766742643,
+      "loss": 0.3569,
+      "step": 5804
+    },
+    {
+      "epoch": 0.4268696227663799,
+      "grad_norm": 0.9085993766784668,
+      "learning_rate": 0.0001281805265223078,
+      "loss": 0.3798,
+      "step": 5805
+    },
+    {
+      "epoch": 0.42694315758511653,
+      "grad_norm": 1.7837226390838623,
+      "learning_rate": 0.00012815767477192207,
+      "loss": 0.4203,
+      "step": 5806
+    },
+    {
+      "epoch": 0.4270166924038532,
+      "grad_norm": 1.2665958404541016,
+      "learning_rate": 0.00012813482142440329,
+      "loss": 0.4133,
+      "step": 5807
+    },
+    {
+      "epoch": 0.4270902272225899,
+      "grad_norm": 1.5940535068511963,
+      "learning_rate": 0.00012811196648104776,
+      "loss": 0.3589,
+      "step": 5808
+    },
+    {
+      "epoch": 0.42716376204132656,
+      "grad_norm": 1.6585439443588257,
+      "learning_rate": 0.00012808910994315182,
+      "loss": 0.3951,
+      "step": 5809
+    },
+    {
+      "epoch": 0.42723729686006323,
+      "grad_norm": 1.0982197523117065,
+      "learning_rate": 0.00012806625181201197,
+      "loss": 0.3762,
+      "step": 5810
+    },
+    {
+      "epoch": 0.4273108316787999,
+      "grad_norm": 1.7897385358810425,
+      "learning_rate": 0.0001280433920889246,
+      "loss": 0.3987,
+      "step": 5811
+    },
+    {
+      "epoch": 0.4273843664975366,
+      "grad_norm": 1.483762264251709,
+      "learning_rate": 0.0001280205307751865,
+      "loss": 0.3839,
+      "step": 5812
+    },
+    {
+      "epoch": 0.42745790131627326,
+      "grad_norm": 1.0165494680404663,
+      "learning_rate": 0.00012799766787209435,
+      "loss": 0.4134,
+      "step": 5813
+    },
+    {
+      "epoch": 0.42753143613500993,
+      "grad_norm": 1.1283659934997559,
+      "learning_rate": 0.00012797480338094488,
+      "loss": 0.3991,
+      "step": 5814
+    },
+    {
+      "epoch": 0.4276049709537466,
+      "grad_norm": 0.7512165904045105,
+      "learning_rate": 0.00012795193730303505,
+      "loss": 0.3845,
+      "step": 5815
+    },
+    {
+      "epoch": 0.4276785057724833,
+      "grad_norm": 1.0367255210876465,
+      "learning_rate": 0.00012792906963966186,
+      "loss": 0.4359,
+      "step": 5816
+    },
+    {
+      "epoch": 0.42775204059121996,
+      "grad_norm": 0.9226056933403015,
+      "learning_rate": 0.00012790620039212236,
+      "loss": 0.3924,
+      "step": 5817
+    },
+    {
+      "epoch": 0.42782557540995664,
+      "grad_norm": 1.1500507593154907,
+      "learning_rate": 0.00012788332956171372,
+      "loss": 0.3756,
+      "step": 5818
+    },
+    {
+      "epoch": 0.4278991102286933,
+      "grad_norm": 0.660908043384552,
+      "learning_rate": 0.00012786045714973315,
+      "loss": 0.3973,
+      "step": 5819
+    },
+    {
+      "epoch": 0.42797264504742993,
+      "grad_norm": 1.2839248180389404,
+      "learning_rate": 0.00012783758315747812,
+      "loss": 0.3738,
+      "step": 5820
+    },
+    {
+      "epoch": 0.4280461798661666,
+      "grad_norm": 0.5021310448646545,
+      "learning_rate": 0.00012781470758624596,
+      "loss": 0.3902,
+      "step": 5821
+    },
+    {
+      "epoch": 0.4281197146849033,
+      "grad_norm": 1.2307029962539673,
+      "learning_rate": 0.00012779183043733422,
+      "loss": 0.4219,
+      "step": 5822
+    },
+    {
+      "epoch": 0.42819324950363996,
+      "grad_norm": 0.9203187823295593,
+      "learning_rate": 0.00012776895171204058,
+      "loss": 0.3922,
+      "step": 5823
+    },
+    {
+      "epoch": 0.42826678432237664,
+      "grad_norm": 0.9146305322647095,
+      "learning_rate": 0.00012774607141166264,
+      "loss": 0.3847,
+      "step": 5824
+    },
+    {
+      "epoch": 0.4283403191411133,
+      "grad_norm": 0.5992823243141174,
+      "learning_rate": 0.00012772318953749827,
+      "loss": 0.4041,
+      "step": 5825
+    },
+    {
+      "epoch": 0.42841385395985,
+      "grad_norm": 0.9921274185180664,
+      "learning_rate": 0.0001277003060908453,
+      "loss": 0.3739,
+      "step": 5826
+    },
+    {
+      "epoch": 0.42848738877858666,
+      "grad_norm": 1.590208888053894,
+      "learning_rate": 0.00012767742107300175,
+      "loss": 0.3964,
+      "step": 5827
+    },
+    {
+      "epoch": 0.42856092359732334,
+      "grad_norm": 0.9028909206390381,
+      "learning_rate": 0.00012765453448526572,
+      "loss": 0.3756,
+      "step": 5828
+    },
+    {
+      "epoch": 0.42863445841606,
+      "grad_norm": 1.0419018268585205,
+      "learning_rate": 0.00012763164632893526,
+      "loss": 0.3823,
+      "step": 5829
+    },
+    {
+      "epoch": 0.4287079932347967,
+      "grad_norm": 1.5758228302001953,
+      "learning_rate": 0.00012760875660530865,
+      "loss": 0.3686,
+      "step": 5830
+    },
+    {
+      "epoch": 0.42878152805353337,
+      "grad_norm": 0.6949378252029419,
+      "learning_rate": 0.00012758586531568428,
+      "loss": 0.393,
+      "step": 5831
+    },
+    {
+      "epoch": 0.42885506287227004,
+      "grad_norm": 1.9521554708480835,
+      "learning_rate": 0.00012756297246136047,
+      "loss": 0.3643,
+      "step": 5832
+    },
+    {
+      "epoch": 0.4289285976910067,
+      "grad_norm": 1.259911298751831,
+      "learning_rate": 0.00012754007804363578,
+      "loss": 0.391,
+      "step": 5833
+    },
+    {
+      "epoch": 0.42900213250974334,
+      "grad_norm": 2.0075478553771973,
+      "learning_rate": 0.0001275171820638088,
+      "loss": 0.3966,
+      "step": 5834
+    },
+    {
+      "epoch": 0.42907566732848,
+      "grad_norm": 1.9768753051757812,
+      "learning_rate": 0.00012749428452317823,
+      "loss": 0.3767,
+      "step": 5835
+    },
+    {
+      "epoch": 0.4291492021472167,
+      "grad_norm": 1.2505030632019043,
+      "learning_rate": 0.00012747138542304283,
+      "loss": 0.3979,
+      "step": 5836
+    },
+    {
+      "epoch": 0.42922273696595337,
+      "grad_norm": 1.3394509553909302,
+      "learning_rate": 0.00012744848476470142,
+      "loss": 0.3772,
+      "step": 5837
+    },
+    {
+      "epoch": 0.42929627178469004,
+      "grad_norm": 1.5486911535263062,
+      "learning_rate": 0.00012742558254945303,
+      "loss": 0.3761,
+      "step": 5838
+    },
+    {
+      "epoch": 0.4293698066034267,
+      "grad_norm": 1.0664561986923218,
+      "learning_rate": 0.00012740267877859663,
+      "loss": 0.386,
+      "step": 5839
+    },
+    {
+      "epoch": 0.4294433414221634,
+      "grad_norm": 2.2041237354278564,
+      "learning_rate": 0.00012737977345343138,
+      "loss": 0.3679,
+      "step": 5840
+    },
+    {
+      "epoch": 0.42951687624090007,
+      "grad_norm": 1.7499923706054688,
+      "learning_rate": 0.00012735686657525648,
+      "loss": 0.4121,
+      "step": 5841
+    },
+    {
+      "epoch": 0.42959041105963675,
+      "grad_norm": 1.596841812133789,
+      "learning_rate": 0.00012733395814537124,
+      "loss": 0.399,
+      "step": 5842
+    },
+    {
+      "epoch": 0.4296639458783734,
+      "grad_norm": 1.586442232131958,
+      "learning_rate": 0.00012731104816507505,
+      "loss": 0.3813,
+      "step": 5843
+    },
+    {
+      "epoch": 0.4297374806971101,
+      "grad_norm": 1.561530590057373,
+      "learning_rate": 0.00012728813663566736,
+      "loss": 0.3877,
+      "step": 5844
+    },
+    {
+      "epoch": 0.4298110155158468,
+      "grad_norm": 1.1756069660186768,
+      "learning_rate": 0.0001272652235584478,
+      "loss": 0.376,
+      "step": 5845
+    },
+    {
+      "epoch": 0.42988455033458345,
+      "grad_norm": 1.7247222661972046,
+      "learning_rate": 0.00012724230893471596,
+      "loss": 0.3752,
+      "step": 5846
+    },
+    {
+      "epoch": 0.4299580851533201,
+      "grad_norm": 1.5416760444641113,
+      "learning_rate": 0.00012721939276577162,
+      "loss": 0.3885,
+      "step": 5847
+    },
+    {
+      "epoch": 0.43003161997205674,
+      "grad_norm": 1.8851720094680786,
+      "learning_rate": 0.00012719647505291456,
+      "loss": 0.3852,
+      "step": 5848
+    },
+    {
+      "epoch": 0.4301051547907934,
+      "grad_norm": 1.908400535583496,
+      "learning_rate": 0.00012717355579744478,
+      "loss": 0.3852,
+      "step": 5849
+    },
+    {
+      "epoch": 0.4301786896095301,
+      "grad_norm": 1.3369773626327515,
+      "learning_rate": 0.00012715063500066221,
+      "loss": 0.3879,
+      "step": 5850
+    },
+    {
+      "epoch": 0.43025222442826677,
+      "grad_norm": 1.4149270057678223,
+      "learning_rate": 0.000127127712663867,
+      "loss": 0.3624,
+      "step": 5851
+    },
+    {
+      "epoch": 0.43032575924700345,
+      "grad_norm": 1.4835931062698364,
+      "learning_rate": 0.00012710478878835926,
+      "loss": 0.3792,
+      "step": 5852
+    },
+    {
+      "epoch": 0.4303992940657401,
+      "grad_norm": 1.0476834774017334,
+      "learning_rate": 0.00012708186337543934,
+      "loss": 0.3781,
+      "step": 5853
+    },
+    {
+      "epoch": 0.4304728288844768,
+      "grad_norm": 1.924220323562622,
+      "learning_rate": 0.00012705893642640751,
+      "loss": 0.3721,
+      "step": 5854
+    },
+    {
+      "epoch": 0.4305463637032135,
+      "grad_norm": 1.5818896293640137,
+      "learning_rate": 0.00012703600794256427,
+      "loss": 0.3734,
+      "step": 5855
+    },
+    {
+      "epoch": 0.43061989852195015,
+      "grad_norm": 1.2372429370880127,
+      "learning_rate": 0.0001270130779252101,
+      "loss": 0.3514,
+      "step": 5856
+    },
+    {
+      "epoch": 0.4306934333406868,
+      "grad_norm": 1.1889479160308838,
+      "learning_rate": 0.00012699014637564568,
+      "loss": 0.3981,
+      "step": 5857
+    },
+    {
+      "epoch": 0.4307669681594235,
+      "grad_norm": 1.8194389343261719,
+      "learning_rate": 0.00012696721329517166,
+      "loss": 0.3889,
+      "step": 5858
+    },
+    {
+      "epoch": 0.4308405029781602,
+      "grad_norm": 1.238734245300293,
+      "learning_rate": 0.00012694427868508885,
+      "loss": 0.3774,
+      "step": 5859
+    },
+    {
+      "epoch": 0.43091403779689685,
+      "grad_norm": 1.8524407148361206,
+      "learning_rate": 0.0001269213425466981,
+      "loss": 0.3859,
+      "step": 5860
+    },
+    {
+      "epoch": 0.43098757261563353,
+      "grad_norm": 1.8122177124023438,
+      "learning_rate": 0.0001268984048813004,
+      "loss": 0.3681,
+      "step": 5861
+    },
+    {
+      "epoch": 0.43106110743437015,
+      "grad_norm": 1.3818330764770508,
+      "learning_rate": 0.00012687546569019678,
+      "loss": 0.3822,
+      "step": 5862
+    },
+    {
+      "epoch": 0.4311346422531068,
+      "grad_norm": 1.58583402633667,
+      "learning_rate": 0.0001268525249746884,
+      "loss": 0.3867,
+      "step": 5863
+    },
+    {
+      "epoch": 0.4312081770718435,
+      "grad_norm": 1.18800950050354,
+      "learning_rate": 0.00012682958273607645,
+      "loss": 0.4278,
+      "step": 5864
+    },
+    {
+      "epoch": 0.4312817118905802,
+      "grad_norm": 0.9897989630699158,
+      "learning_rate": 0.00012680663897566224,
+      "loss": 0.3846,
+      "step": 5865
+    },
+    {
+      "epoch": 0.43135524670931685,
+      "grad_norm": 1.7587473392486572,
+      "learning_rate": 0.00012678369369474718,
+      "loss": 0.3965,
+      "step": 5866
+    },
+    {
+      "epoch": 0.43142878152805353,
+      "grad_norm": 1.4206221103668213,
+      "learning_rate": 0.00012676074689463277,
+      "loss": 0.3679,
+      "step": 5867
+    },
+    {
+      "epoch": 0.4315023163467902,
+      "grad_norm": 1.7732003927230835,
+      "learning_rate": 0.00012673779857662052,
+      "loss": 0.3904,
+      "step": 5868
+    },
+    {
+      "epoch": 0.4315758511655269,
+      "grad_norm": 1.9719446897506714,
+      "learning_rate": 0.00012671484874201218,
+      "loss": 0.3921,
+      "step": 5869
+    },
+    {
+      "epoch": 0.43164938598426356,
+      "grad_norm": 0.9708046317100525,
+      "learning_rate": 0.00012669189739210936,
+      "loss": 0.3805,
+      "step": 5870
+    },
+    {
+      "epoch": 0.43172292080300023,
+      "grad_norm": 1.0361801385879517,
+      "learning_rate": 0.000126668944528214,
+      "loss": 0.364,
+      "step": 5871
+    },
+    {
+      "epoch": 0.4317964556217369,
+      "grad_norm": 1.3588337898254395,
+      "learning_rate": 0.00012664599015162794,
+      "loss": 0.3953,
+      "step": 5872
+    },
+    {
+      "epoch": 0.4318699904404736,
+      "grad_norm": 0.8108475208282471,
+      "learning_rate": 0.0001266230342636532,
+      "loss": 0.3846,
+      "step": 5873
+    },
+    {
+      "epoch": 0.43194352525921026,
+      "grad_norm": 1.892851710319519,
+      "learning_rate": 0.00012660007686559184,
+      "loss": 0.3785,
+      "step": 5874
+    },
+    {
+      "epoch": 0.43201706007794693,
+      "grad_norm": 1.3042407035827637,
+      "learning_rate": 0.00012657711795874607,
+      "loss": 0.4031,
+      "step": 5875
+    },
+    {
+      "epoch": 0.43209059489668356,
+      "grad_norm": 2.0653698444366455,
+      "learning_rate": 0.00012655415754441813,
+      "loss": 0.4075,
+      "step": 5876
+    },
+    {
+      "epoch": 0.43216412971542023,
+      "grad_norm": 1.9579707384109497,
+      "learning_rate": 0.00012653119562391034,
+      "loss": 0.3656,
+      "step": 5877
+    },
+    {
+      "epoch": 0.4322376645341569,
+      "grad_norm": 0.9974393248558044,
+      "learning_rate": 0.0001265082321985251,
+      "loss": 0.3808,
+      "step": 5878
+    },
+    {
+      "epoch": 0.4323111993528936,
+      "grad_norm": 1.4509613513946533,
+      "learning_rate": 0.000126485267269565,
+      "loss": 0.398,
+      "step": 5879
+    },
+    {
+      "epoch": 0.43238473417163026,
+      "grad_norm": 1.196758508682251,
+      "learning_rate": 0.0001264623008383326,
+      "loss": 0.3556,
+      "step": 5880
+    },
+    {
+      "epoch": 0.43245826899036693,
+      "grad_norm": 0.7982028126716614,
+      "learning_rate": 0.00012643933290613056,
+      "loss": 0.3815,
+      "step": 5881
+    },
+    {
+      "epoch": 0.4325318038091036,
+      "grad_norm": 1.581884503364563,
+      "learning_rate": 0.0001264163634742616,
+      "loss": 0.3975,
+      "step": 5882
+    },
+    {
+      "epoch": 0.4326053386278403,
+      "grad_norm": 0.7045242786407471,
+      "learning_rate": 0.00012639339254402868,
+      "loss": 0.3527,
+      "step": 5883
+    },
+    {
+      "epoch": 0.43267887344657696,
+      "grad_norm": 2.0847396850585938,
+      "learning_rate": 0.00012637042011673467,
+      "loss": 0.3821,
+      "step": 5884
+    },
+    {
+      "epoch": 0.43275240826531364,
+      "grad_norm": 1.2128289937973022,
+      "learning_rate": 0.0001263474461936826,
+      "loss": 0.3972,
+      "step": 5885
+    },
+    {
+      "epoch": 0.4328259430840503,
+      "grad_norm": 2.2787933349609375,
+      "learning_rate": 0.00012632447077617556,
+      "loss": 0.3757,
+      "step": 5886
+    },
+    {
+      "epoch": 0.432899477902787,
+      "grad_norm": 2.380673408508301,
+      "learning_rate": 0.00012630149386551678,
+      "loss": 0.4053,
+      "step": 5887
+    },
+    {
+      "epoch": 0.43297301272152366,
+      "grad_norm": 0.7981759309768677,
+      "learning_rate": 0.0001262785154630095,
+      "loss": 0.4006,
+      "step": 5888
+    },
+    {
+      "epoch": 0.43304654754026034,
+      "grad_norm": 1.7699722051620483,
+      "learning_rate": 0.0001262555355699571,
+      "loss": 0.4236,
+      "step": 5889
+    },
+    {
+      "epoch": 0.43312008235899696,
+      "grad_norm": 0.7440626621246338,
+      "learning_rate": 0.000126232554187663,
+      "loss": 0.3781,
+      "step": 5890
+    },
+    {
+      "epoch": 0.43319361717773364,
+      "grad_norm": 1.7385975122451782,
+      "learning_rate": 0.00012620957131743074,
+      "loss": 0.397,
+      "step": 5891
+    },
+    {
+      "epoch": 0.4332671519964703,
+      "grad_norm": 0.6405933499336243,
+      "learning_rate": 0.00012618658696056398,
+      "loss": 0.3737,
+      "step": 5892
+    },
+    {
+      "epoch": 0.433340686815207,
+      "grad_norm": 1.2188366651535034,
+      "learning_rate": 0.0001261636011183663,
+      "loss": 0.3881,
+      "step": 5893
+    },
+    {
+      "epoch": 0.43341422163394366,
+      "grad_norm": 0.7590214610099792,
+      "learning_rate": 0.0001261406137921416,
+      "loss": 0.3567,
+      "step": 5894
+    },
+    {
+      "epoch": 0.43348775645268034,
+      "grad_norm": 0.9732641577720642,
+      "learning_rate": 0.0001261176249831937,
+      "loss": 0.3744,
+      "step": 5895
+    },
+    {
+      "epoch": 0.433561291271417,
+      "grad_norm": 1.3910542726516724,
+      "learning_rate": 0.0001260946346928265,
+      "loss": 0.3641,
+      "step": 5896
+    },
+    {
+      "epoch": 0.4336348260901537,
+      "grad_norm": 0.7744175791740417,
+      "learning_rate": 0.0001260716429223441,
+      "loss": 0.3793,
+      "step": 5897
+    },
+    {
+      "epoch": 0.43370836090889037,
+      "grad_norm": 1.548194408416748,
+      "learning_rate": 0.00012604864967305065,
+      "loss": 0.3935,
+      "step": 5898
+    },
+    {
+      "epoch": 0.43378189572762704,
+      "grad_norm": 0.8192543983459473,
+      "learning_rate": 0.00012602565494625027,
+      "loss": 0.3886,
+      "step": 5899
+    },
+    {
+      "epoch": 0.4338554305463637,
+      "grad_norm": 1.777228832244873,
+      "learning_rate": 0.0001260026587432473,
+      "loss": 0.361,
+      "step": 5900
+    },
+    {
+      "epoch": 0.4339289653651004,
+      "grad_norm": 1.3509855270385742,
+      "learning_rate": 0.000125979661065346,
+      "loss": 0.4066,
+      "step": 5901
+    },
+    {
+      "epoch": 0.43400250018383707,
+      "grad_norm": 2.024864673614502,
+      "learning_rate": 0.00012595666191385097,
+      "loss": 0.377,
+      "step": 5902
+    },
+    {
+      "epoch": 0.43407603500257375,
+      "grad_norm": 2.0132036209106445,
+      "learning_rate": 0.0001259336612900667,
+      "loss": 0.3761,
+      "step": 5903
+    },
+    {
+      "epoch": 0.43414956982131037,
+      "grad_norm": 1.3902997970581055,
+      "learning_rate": 0.00012591065919529776,
+      "loss": 0.3865,
+      "step": 5904
+    },
+    {
+      "epoch": 0.43422310464004704,
+      "grad_norm": 1.1483049392700195,
+      "learning_rate": 0.0001258876556308489,
+      "loss": 0.3839,
+      "step": 5905
+    },
+    {
+      "epoch": 0.4342966394587837,
+      "grad_norm": 1.7527952194213867,
+      "learning_rate": 0.00012586465059802486,
+      "loss": 0.3831,
+      "step": 5906
+    },
+    {
+      "epoch": 0.4343701742775204,
+      "grad_norm": 1.5796821117401123,
+      "learning_rate": 0.00012584164409813056,
+      "loss": 0.3842,
+      "step": 5907
+    },
+    {
+      "epoch": 0.43444370909625707,
+      "grad_norm": 1.8428033590316772,
+      "learning_rate": 0.00012581863613247095,
+      "loss": 0.3718,
+      "step": 5908
+    },
+    {
+      "epoch": 0.43451724391499374,
+      "grad_norm": 2.1778392791748047,
+      "learning_rate": 0.00012579562670235106,
+      "loss": 0.3959,
+      "step": 5909
+    },
+    {
+      "epoch": 0.4345907787337304,
+      "grad_norm": 1.4758045673370361,
+      "learning_rate": 0.000125772615809076,
+      "loss": 0.3926,
+      "step": 5910
+    },
+    {
+      "epoch": 0.4346643135524671,
+      "grad_norm": 1.258288860321045,
+      "learning_rate": 0.00012574960345395093,
+      "loss": 0.4132,
+      "step": 5911
+    },
+    {
+      "epoch": 0.43473784837120377,
+      "grad_norm": 1.8735655546188354,
+      "learning_rate": 0.0001257265896382812,
+      "loss": 0.3812,
+      "step": 5912
+    },
+    {
+      "epoch": 0.43481138318994045,
+      "grad_norm": 1.4210708141326904,
+      "learning_rate": 0.0001257035743633722,
+      "loss": 0.3815,
+      "step": 5913
+    },
+    {
+      "epoch": 0.4348849180086771,
+      "grad_norm": 1.4847701787948608,
+      "learning_rate": 0.00012568055763052927,
+      "loss": 0.3934,
+      "step": 5914
+    },
+    {
+      "epoch": 0.4349584528274138,
+      "grad_norm": 1.3773399591445923,
+      "learning_rate": 0.00012565753944105806,
+      "loss": 0.387,
+      "step": 5915
+    },
+    {
+      "epoch": 0.4350319876461505,
+      "grad_norm": 1.2998204231262207,
+      "learning_rate": 0.0001256345197962641,
+      "loss": 0.3801,
+      "step": 5916
+    },
+    {
+      "epoch": 0.43510552246488715,
+      "grad_norm": 1.1145050525665283,
+      "learning_rate": 0.00012561149869745313,
+      "loss": 0.368,
+      "step": 5917
+    },
+    {
+      "epoch": 0.43517905728362377,
+      "grad_norm": 1.703808307647705,
+      "learning_rate": 0.00012558847614593092,
+      "loss": 0.3683,
+      "step": 5918
+    },
+    {
+      "epoch": 0.43525259210236045,
+      "grad_norm": 1.236952543258667,
+      "learning_rate": 0.00012556545214300333,
+      "loss": 0.3742,
+      "step": 5919
+    },
+    {
+      "epoch": 0.4353261269210971,
+      "grad_norm": 2.0550522804260254,
+      "learning_rate": 0.00012554242668997635,
+      "loss": 0.4069,
+      "step": 5920
+    },
+    {
+      "epoch": 0.4353996617398338,
+      "grad_norm": 1.7520710229873657,
+      "learning_rate": 0.00012551939978815594,
+      "loss": 0.3851,
+      "step": 5921
+    },
+    {
+      "epoch": 0.4354731965585705,
+      "grad_norm": 1.4205292463302612,
+      "learning_rate": 0.00012549637143884827,
+      "loss": 0.3565,
+      "step": 5922
+    },
+    {
+      "epoch": 0.43554673137730715,
+      "grad_norm": 1.491978406906128,
+      "learning_rate": 0.0001254733416433595,
+      "loss": 0.3907,
+      "step": 5923
+    },
+    {
+      "epoch": 0.4356202661960438,
+      "grad_norm": 1.2483025789260864,
+      "learning_rate": 0.00012545031040299588,
+      "loss": 0.4045,
+      "step": 5924
+    },
+    {
+      "epoch": 0.4356938010147805,
+      "grad_norm": 1.1470203399658203,
+      "learning_rate": 0.00012542727771906381,
+      "loss": 0.3708,
+      "step": 5925
+    },
+    {
+      "epoch": 0.4357673358335172,
+      "grad_norm": 1.8256465196609497,
+      "learning_rate": 0.00012540424359286972,
+      "loss": 0.3647,
+      "step": 5926
+    },
+    {
+      "epoch": 0.43584087065225385,
+      "grad_norm": 1.7860547304153442,
+      "learning_rate": 0.00012538120802572014,
+      "loss": 0.3726,
+      "step": 5927
+    },
+    {
+      "epoch": 0.43591440547099053,
+      "grad_norm": 1.4075762033462524,
+      "learning_rate": 0.00012535817101892165,
+      "loss": 0.3607,
+      "step": 5928
+    },
+    {
+      "epoch": 0.4359879402897272,
+      "grad_norm": 1.4865667819976807,
+      "learning_rate": 0.00012533513257378088,
+      "loss": 0.3967,
+      "step": 5929
+    },
+    {
+      "epoch": 0.4360614751084639,
+      "grad_norm": 1.3073983192443848,
+      "learning_rate": 0.0001253120926916047,
+      "loss": 0.4029,
+      "step": 5930
+    },
+    {
+      "epoch": 0.43613500992720056,
+      "grad_norm": 1.0129506587982178,
+      "learning_rate": 0.00012528905137369992,
+      "loss": 0.3795,
+      "step": 5931
+    },
+    {
+      "epoch": 0.4362085447459372,
+      "grad_norm": 1.8294744491577148,
+      "learning_rate": 0.00012526600862137343,
+      "loss": 0.3664,
+      "step": 5932
+    },
+    {
+      "epoch": 0.43628207956467385,
+      "grad_norm": 1.5671509504318237,
+      "learning_rate": 0.0001252429644359323,
+      "loss": 0.4049,
+      "step": 5933
+    },
+    {
+      "epoch": 0.43635561438341053,
+      "grad_norm": 1.3951820135116577,
+      "learning_rate": 0.00012521991881868355,
+      "loss": 0.3986,
+      "step": 5934
+    },
+    {
+      "epoch": 0.4364291492021472,
+      "grad_norm": 1.5219182968139648,
+      "learning_rate": 0.0001251968717709344,
+      "loss": 0.3599,
+      "step": 5935
+    },
+    {
+      "epoch": 0.4365026840208839,
+      "grad_norm": 1.4801759719848633,
+      "learning_rate": 0.00012517382329399213,
+      "loss": 0.3693,
+      "step": 5936
+    },
+    {
+      "epoch": 0.43657621883962056,
+      "grad_norm": 1.250544786453247,
+      "learning_rate": 0.000125150773389164,
+      "loss": 0.3684,
+      "step": 5937
+    },
+    {
+      "epoch": 0.43664975365835723,
+      "grad_norm": 1.5800957679748535,
+      "learning_rate": 0.00012512772205775747,
+      "loss": 0.3826,
+      "step": 5938
+    },
+    {
+      "epoch": 0.4367232884770939,
+      "grad_norm": 1.3196001052856445,
+      "learning_rate": 0.00012510466930108,
+      "loss": 0.375,
+      "step": 5939
+    },
+    {
+      "epoch": 0.4367968232958306,
+      "grad_norm": 1.8058947324752808,
+      "learning_rate": 0.00012508161512043926,
+      "loss": 0.358,
+      "step": 5940
+    },
+    {
+      "epoch": 0.43687035811456726,
+      "grad_norm": 1.7516498565673828,
+      "learning_rate": 0.0001250585595171428,
+      "loss": 0.3575,
+      "step": 5941
+    },
+    {
+      "epoch": 0.43694389293330393,
+      "grad_norm": 1.3904705047607422,
+      "learning_rate": 0.00012503550249249836,
+      "loss": 0.3504,
+      "step": 5942
+    },
+    {
+      "epoch": 0.4370174277520406,
+      "grad_norm": 1.2901880741119385,
+      "learning_rate": 0.00012501244404781385,
+      "loss": 0.3958,
+      "step": 5943
+    },
+    {
+      "epoch": 0.4370909625707773,
+      "grad_norm": 1.7471380233764648,
+      "learning_rate": 0.0001249893841843971,
+      "loss": 0.3716,
+      "step": 5944
+    },
+    {
+      "epoch": 0.43716449738951396,
+      "grad_norm": 1.3950375318527222,
+      "learning_rate": 0.0001249663229035561,
+      "loss": 0.3647,
+      "step": 5945
+    },
+    {
+      "epoch": 0.4372380322082506,
+      "grad_norm": 1.4921221733093262,
+      "learning_rate": 0.00012494326020659892,
+      "loss": 0.373,
+      "step": 5946
+    },
+    {
+      "epoch": 0.43731156702698726,
+      "grad_norm": 1.5703705549240112,
+      "learning_rate": 0.0001249201960948337,
+      "loss": 0.3871,
+      "step": 5947
+    },
+    {
+      "epoch": 0.43738510184572393,
+      "grad_norm": 1.3970322608947754,
+      "learning_rate": 0.00012489713056956867,
+      "loss": 0.3631,
+      "step": 5948
+    },
+    {
+      "epoch": 0.4374586366644606,
+      "grad_norm": 1.2769110202789307,
+      "learning_rate": 0.00012487406363211208,
+      "loss": 0.3717,
+      "step": 5949
+    },
+    {
+      "epoch": 0.4375321714831973,
+      "grad_norm": 1.7132784128189087,
+      "learning_rate": 0.0001248509952837724,
+      "loss": 0.3778,
+      "step": 5950
+    },
+    {
+      "epoch": 0.43760570630193396,
+      "grad_norm": 1.5597718954086304,
+      "learning_rate": 0.00012482792552585804,
+      "loss": 0.3721,
+      "step": 5951
+    },
+    {
+      "epoch": 0.43767924112067064,
+      "grad_norm": 1.6801702976226807,
+      "learning_rate": 0.0001248048543596775,
+      "loss": 0.3662,
+      "step": 5952
+    },
+    {
+      "epoch": 0.4377527759394073,
+      "grad_norm": 1.449372410774231,
+      "learning_rate": 0.00012478178178653947,
+      "loss": 0.3475,
+      "step": 5953
+    },
+    {
+      "epoch": 0.437826310758144,
+      "grad_norm": 1.6048544645309448,
+      "learning_rate": 0.00012475870780775264,
+      "loss": 0.3774,
+      "step": 5954
+    },
+    {
+      "epoch": 0.43789984557688066,
+      "grad_norm": 1.5646110773086548,
+      "learning_rate": 0.00012473563242462572,
+      "loss": 0.3538,
+      "step": 5955
+    },
+    {
+      "epoch": 0.43797338039561734,
+      "grad_norm": 1.067352294921875,
+      "learning_rate": 0.00012471255563846769,
+      "loss": 0.3798,
+      "step": 5956
+    },
+    {
+      "epoch": 0.438046915214354,
+      "grad_norm": 1.083954095840454,
+      "learning_rate": 0.00012468947745058738,
+      "loss": 0.3613,
+      "step": 5957
+    },
+    {
+      "epoch": 0.4381204500330907,
+      "grad_norm": 1.6652755737304688,
+      "learning_rate": 0.00012466639786229387,
+      "loss": 0.3696,
+      "step": 5958
+    },
+    {
+      "epoch": 0.43819398485182737,
+      "grad_norm": 1.4268313646316528,
+      "learning_rate": 0.00012464331687489624,
+      "loss": 0.4236,
+      "step": 5959
+    },
+    {
+      "epoch": 0.438267519670564,
+      "grad_norm": 2.026304244995117,
+      "learning_rate": 0.00012462023448970362,
+      "loss": 0.383,
+      "step": 5960
+    },
+    {
+      "epoch": 0.43834105448930066,
+      "grad_norm": 1.8628499507904053,
+      "learning_rate": 0.00012459715070802538,
+      "loss": 0.3781,
+      "step": 5961
+    },
+    {
+      "epoch": 0.43841458930803734,
+      "grad_norm": 1.2094273567199707,
+      "learning_rate": 0.00012457406553117078,
+      "loss": 0.3675,
+      "step": 5962
+    },
+    {
+      "epoch": 0.438488124126774,
+      "grad_norm": 0.8049122095108032,
+      "learning_rate": 0.00012455097896044924,
+      "loss": 0.355,
+      "step": 5963
+    },
+    {
+      "epoch": 0.4385616589455107,
+      "grad_norm": 1.624107003211975,
+      "learning_rate": 0.00012452789099717029,
+      "loss": 0.367,
+      "step": 5964
+    },
+    {
+      "epoch": 0.43863519376424737,
+      "grad_norm": 0.7108777761459351,
+      "learning_rate": 0.00012450480164264343,
+      "loss": 0.3719,
+      "step": 5965
+    },
+    {
+      "epoch": 0.43870872858298404,
+      "grad_norm": 1.8896458148956299,
+      "learning_rate": 0.0001244817108981784,
+      "loss": 0.3901,
+      "step": 5966
+    },
+    {
+      "epoch": 0.4387822634017207,
+      "grad_norm": 1.337319016456604,
+      "learning_rate": 0.00012445861876508489,
+      "loss": 0.3677,
+      "step": 5967
+    },
+    {
+      "epoch": 0.4388557982204574,
+      "grad_norm": 2.333338499069214,
+      "learning_rate": 0.0001244355252446727,
+      "loss": 0.3703,
+      "step": 5968
+    },
+    {
+      "epoch": 0.43892933303919407,
+      "grad_norm": 2.4223484992980957,
+      "learning_rate": 0.00012441243033825173,
+      "loss": 0.4031,
+      "step": 5969
+    },
+    {
+      "epoch": 0.43900286785793075,
+      "grad_norm": 0.5479739904403687,
+      "learning_rate": 0.00012438933404713192,
+      "loss": 0.3554,
+      "step": 5970
+    },
+    {
+      "epoch": 0.4390764026766674,
+      "grad_norm": 1.9446601867675781,
+      "learning_rate": 0.0001243662363726234,
+      "loss": 0.3933,
+      "step": 5971
+    },
+    {
+      "epoch": 0.4391499374954041,
+      "grad_norm": 0.9012075662612915,
+      "learning_rate": 0.00012434313731603619,
+      "loss": 0.3874,
+      "step": 5972
+    },
+    {
+      "epoch": 0.4392234723141408,
+      "grad_norm": 2.9173319339752197,
+      "learning_rate": 0.00012432003687868056,
+      "loss": 0.3859,
+      "step": 5973
+    },
+    {
+      "epoch": 0.4392970071328774,
+      "grad_norm": 3.1832404136657715,
+      "learning_rate": 0.00012429693506186678,
+      "loss": 0.3855,
+      "step": 5974
+    },
+    {
+      "epoch": 0.43937054195161407,
+      "grad_norm": 0.6436663269996643,
+      "learning_rate": 0.0001242738318669052,
+      "loss": 0.3791,
+      "step": 5975
+    },
+    {
+      "epoch": 0.43944407677035074,
+      "grad_norm": 2.8878772258758545,
+      "learning_rate": 0.00012425072729510626,
+      "loss": 0.3802,
+      "step": 5976
+    },
+    {
+      "epoch": 0.4395176115890874,
+      "grad_norm": 2.9075913429260254,
+      "learning_rate": 0.00012422762134778047,
+      "loss": 0.385,
+      "step": 5977
+    },
+    {
+      "epoch": 0.4395911464078241,
+      "grad_norm": 0.7999001741409302,
+      "learning_rate": 0.0001242045140262384,
+      "loss": 0.3713,
+      "step": 5978
+    },
+    {
+      "epoch": 0.43966468122656077,
+      "grad_norm": 2.405146837234497,
+      "learning_rate": 0.00012418140533179076,
+      "loss": 0.3973,
+      "step": 5979
+    },
+    {
+      "epoch": 0.43973821604529745,
+      "grad_norm": 2.0819146633148193,
+      "learning_rate": 0.00012415829526574835,
+      "loss": 0.403,
+      "step": 5980
+    },
+    {
+      "epoch": 0.4398117508640341,
+      "grad_norm": 1.8015068769454956,
+      "learning_rate": 0.0001241351838294219,
+      "loss": 0.4014,
+      "step": 5981
+    },
+    {
+      "epoch": 0.4398852856827708,
+      "grad_norm": 1.8618046045303345,
+      "learning_rate": 0.00012411207102412233,
+      "loss": 0.3946,
+      "step": 5982
+    },
+    {
+      "epoch": 0.4399588205015075,
+      "grad_norm": 1.0091818571090698,
+      "learning_rate": 0.00012408895685116065,
+      "loss": 0.4032,
+      "step": 5983
+    },
+    {
+      "epoch": 0.44003235532024415,
+      "grad_norm": 1.4132481813430786,
+      "learning_rate": 0.00012406584131184793,
+      "loss": 0.3507,
+      "step": 5984
+    },
+    {
+      "epoch": 0.4401058901389808,
+      "grad_norm": 0.6846411228179932,
+      "learning_rate": 0.0001240427244074953,
+      "loss": 0.3552,
+      "step": 5985
+    },
+    {
+      "epoch": 0.4401794249577175,
+      "grad_norm": 1.6641840934753418,
+      "learning_rate": 0.00012401960613941398,
+      "loss": 0.3702,
+      "step": 5986
+    },
+    {
+      "epoch": 0.4402529597764542,
+      "grad_norm": 0.6420547366142273,
+      "learning_rate": 0.0001239964865089152,
+      "loss": 0.387,
+      "step": 5987
+    },
+    {
+      "epoch": 0.4403264945951908,
+      "grad_norm": 2.0151443481445312,
+      "learning_rate": 0.0001239733655173104,
+      "loss": 0.3902,
+      "step": 5988
+    },
+    {
+      "epoch": 0.4404000294139275,
+      "grad_norm": 1.761855959892273,
+      "learning_rate": 0.00012395024316591105,
+      "loss": 0.3898,
+      "step": 5989
+    },
+    {
+      "epoch": 0.44047356423266415,
+      "grad_norm": 1.4089616537094116,
+      "learning_rate": 0.0001239271194560286,
+      "loss": 0.3718,
+      "step": 5990
+    },
+    {
+      "epoch": 0.4405470990514008,
+      "grad_norm": 1.7137305736541748,
+      "learning_rate": 0.00012390399438897467,
+      "loss": 0.4028,
+      "step": 5991
+    },
+    {
+      "epoch": 0.4406206338701375,
+      "grad_norm": 0.9711654782295227,
+      "learning_rate": 0.00012388086796606096,
+      "loss": 0.3883,
+      "step": 5992
+    },
+    {
+      "epoch": 0.4406941686888742,
+      "grad_norm": 0.7140021324157715,
+      "learning_rate": 0.00012385774018859922,
+      "loss": 0.3781,
+      "step": 5993
+    },
+    {
+      "epoch": 0.44076770350761085,
+      "grad_norm": 0.9851916432380676,
+      "learning_rate": 0.0001238346110579013,
+      "loss": 0.3708,
+      "step": 5994
+    },
+    {
+      "epoch": 0.44084123832634753,
+      "grad_norm": 0.6886529922485352,
+      "learning_rate": 0.00012381148057527905,
+      "loss": 0.3558,
+      "step": 5995
+    },
+    {
+      "epoch": 0.4409147731450842,
+      "grad_norm": 0.6543463468551636,
+      "learning_rate": 0.0001237883487420445,
+      "loss": 0.3465,
+      "step": 5996
+    },
+    {
+      "epoch": 0.4409883079638209,
+      "grad_norm": 1.0763760805130005,
+      "learning_rate": 0.00012376521555950974,
+      "loss": 0.3709,
+      "step": 5997
+    },
+    {
+      "epoch": 0.44106184278255756,
+      "grad_norm": 0.5582436919212341,
+      "learning_rate": 0.00012374208102898683,
+      "loss": 0.3689,
+      "step": 5998
+    },
+    {
+      "epoch": 0.44113537760129423,
+      "grad_norm": 0.6879895329475403,
+      "learning_rate": 0.00012371894515178804,
+      "loss": 0.3776,
+      "step": 5999
+    },
+    {
+      "epoch": 0.4412089124200309,
+      "grad_norm": 0.5740577578544617,
+      "learning_rate": 0.00012369580792922567,
+      "loss": 0.3587,
+      "step": 6000
+    },
+    {
+      "epoch": 0.4412824472387676,
+      "grad_norm": 0.5458384156227112,
+      "learning_rate": 0.00012367266936261204,
+      "loss": 0.3899,
+      "step": 6001
+    },
+    {
+      "epoch": 0.4413559820575042,
+      "grad_norm": 0.5480378270149231,
+      "learning_rate": 0.00012364952945325962,
+      "loss": 0.3725,
+      "step": 6002
+    },
+    {
+      "epoch": 0.4414295168762409,
+      "grad_norm": 0.7895516753196716,
+      "learning_rate": 0.00012362638820248096,
+      "loss": 0.3666,
+      "step": 6003
+    },
+    {
+      "epoch": 0.44150305169497756,
+      "grad_norm": 0.9104695916175842,
+      "learning_rate": 0.0001236032456115886,
+      "loss": 0.3406,
+      "step": 6004
+    },
+    {
+      "epoch": 0.44157658651371423,
+      "grad_norm": 0.5748873353004456,
+      "learning_rate": 0.00012358010168189528,
+      "loss": 0.3926,
+      "step": 6005
+    },
+    {
+      "epoch": 0.4416501213324509,
+      "grad_norm": 1.0510332584381104,
+      "learning_rate": 0.00012355695641471366,
+      "loss": 0.405,
+      "step": 6006
+    },
+    {
+      "epoch": 0.4417236561511876,
+      "grad_norm": 0.9277142286300659,
+      "learning_rate": 0.00012353380981135661,
+      "loss": 0.3701,
+      "step": 6007
+    },
+    {
+      "epoch": 0.44179719096992426,
+      "grad_norm": 0.8375948667526245,
+      "learning_rate": 0.00012351066187313707,
+      "loss": 0.3612,
+      "step": 6008
+    },
+    {
+      "epoch": 0.44187072578866093,
+      "grad_norm": 0.6838443875312805,
+      "learning_rate": 0.00012348751260136796,
+      "loss": 0.3836,
+      "step": 6009
+    },
+    {
+      "epoch": 0.4419442606073976,
+      "grad_norm": 0.9474415183067322,
+      "learning_rate": 0.00012346436199736237,
+      "loss": 0.3743,
+      "step": 6010
+    },
+    {
+      "epoch": 0.4420177954261343,
+      "grad_norm": 1.017304539680481,
+      "learning_rate": 0.00012344121006243334,
+      "loss": 0.3776,
+      "step": 6011
+    },
+    {
+      "epoch": 0.44209133024487096,
+      "grad_norm": 0.5182919502258301,
+      "learning_rate": 0.0001234180567978942,
+      "loss": 0.3658,
+      "step": 6012
+    },
+    {
+      "epoch": 0.44216486506360764,
+      "grad_norm": 1.1718809604644775,
+      "learning_rate": 0.00012339490220505812,
+      "loss": 0.3923,
+      "step": 6013
+    },
+    {
+      "epoch": 0.4422383998823443,
+      "grad_norm": 0.9113196134567261,
+      "learning_rate": 0.00012337174628523852,
+      "loss": 0.3488,
+      "step": 6014
+    },
+    {
+      "epoch": 0.442311934701081,
+      "grad_norm": 0.5659571290016174,
+      "learning_rate": 0.00012334858903974882,
+      "loss": 0.3695,
+      "step": 6015
+    },
+    {
+      "epoch": 0.4423854695198176,
+      "grad_norm": 1.2299556732177734,
+      "learning_rate": 0.00012332543046990248,
+      "loss": 0.3619,
+      "step": 6016
+    },
+    {
+      "epoch": 0.4424590043385543,
+      "grad_norm": 0.5921229124069214,
+      "learning_rate": 0.0001233022705770131,
+      "loss": 0.3439,
+      "step": 6017
+    },
+    {
+      "epoch": 0.44253253915729096,
+      "grad_norm": 1.0303611755371094,
+      "learning_rate": 0.00012327910936239437,
+      "loss": 0.3697,
+      "step": 6018
+    },
+    {
+      "epoch": 0.44260607397602764,
+      "grad_norm": 1.3760707378387451,
+      "learning_rate": 0.00012325594682735996,
+      "loss": 0.3812,
+      "step": 6019
+    },
+    {
+      "epoch": 0.4426796087947643,
+      "grad_norm": 0.6038748025894165,
+      "learning_rate": 0.00012323278297322375,
+      "loss": 0.3672,
+      "step": 6020
+    },
+    {
+      "epoch": 0.442753143613501,
+      "grad_norm": 1.631163477897644,
+      "learning_rate": 0.00012320961780129953,
+      "loss": 0.3739,
+      "step": 6021
+    },
+    {
+      "epoch": 0.44282667843223766,
+      "grad_norm": 0.7421682476997375,
+      "learning_rate": 0.0001231864513129013,
+      "loss": 0.3899,
+      "step": 6022
+    },
+    {
+      "epoch": 0.44290021325097434,
+      "grad_norm": 1.8277913331985474,
+      "learning_rate": 0.00012316328350934312,
+      "loss": 0.379,
+      "step": 6023
+    },
+    {
+      "epoch": 0.442973748069711,
+      "grad_norm": 1.360600471496582,
+      "learning_rate": 0.00012314011439193903,
+      "loss": 0.3532,
+      "step": 6024
+    },
+    {
+      "epoch": 0.4430472828884477,
+      "grad_norm": 1.7395445108413696,
+      "learning_rate": 0.00012311694396200323,
+      "loss": 0.3955,
+      "step": 6025
+    },
+    {
+      "epoch": 0.44312081770718437,
+      "grad_norm": 1.875069499015808,
+      "learning_rate": 0.00012309377222084994,
+      "loss": 0.3683,
+      "step": 6026
+    },
+    {
+      "epoch": 0.44319435252592104,
+      "grad_norm": 1.4154644012451172,
+      "learning_rate": 0.00012307059916979356,
+      "loss": 0.3475,
+      "step": 6027
+    },
+    {
+      "epoch": 0.4432678873446577,
+      "grad_norm": 1.17668616771698,
+      "learning_rate": 0.00012304742481014845,
+      "loss": 0.361,
+      "step": 6028
+    },
+    {
+      "epoch": 0.4433414221633944,
+      "grad_norm": 1.4294553995132446,
+      "learning_rate": 0.00012302424914322903,
+      "loss": 0.3612,
+      "step": 6029
+    },
+    {
+      "epoch": 0.443414956982131,
+      "grad_norm": 1.0353325605392456,
+      "learning_rate": 0.00012300107217034996,
+      "loss": 0.3692,
+      "step": 6030
+    },
+    {
+      "epoch": 0.4434884918008677,
+      "grad_norm": 2.0858185291290283,
+      "learning_rate": 0.00012297789389282578,
+      "loss": 0.3809,
+      "step": 6031
+    },
+    {
+      "epoch": 0.44356202661960437,
+      "grad_norm": 1.7050902843475342,
+      "learning_rate": 0.00012295471431197122,
+      "loss": 0.3773,
+      "step": 6032
+    },
+    {
+      "epoch": 0.44363556143834104,
+      "grad_norm": 1.6994500160217285,
+      "learning_rate": 0.00012293153342910103,
+      "loss": 0.3569,
+      "step": 6033
+    },
+    {
+      "epoch": 0.4437090962570777,
+      "grad_norm": 1.6486248970031738,
+      "learning_rate": 0.00012290835124553003,
+      "loss": 0.4156,
+      "step": 6034
+    },
+    {
+      "epoch": 0.4437826310758144,
+      "grad_norm": 1.2629139423370361,
+      "learning_rate": 0.00012288516776257322,
+      "loss": 0.3811,
+      "step": 6035
+    },
+    {
+      "epoch": 0.44385616589455107,
+      "grad_norm": 1.2565760612487793,
+      "learning_rate": 0.00012286198298154551,
+      "loss": 0.3716,
+      "step": 6036
+    },
+    {
+      "epoch": 0.44392970071328774,
+      "grad_norm": 1.4731825590133667,
+      "learning_rate": 0.00012283879690376198,
+      "loss": 0.3467,
+      "step": 6037
+    },
+    {
+      "epoch": 0.4440032355320244,
+      "grad_norm": 1.226193904876709,
+      "learning_rate": 0.00012281560953053785,
+      "loss": 0.3625,
+      "step": 6038
+    },
+    {
+      "epoch": 0.4440767703507611,
+      "grad_norm": 1.522558569908142,
+      "learning_rate": 0.00012279242086318823,
+      "loss": 0.3681,
+      "step": 6039
+    },
+    {
+      "epoch": 0.44415030516949777,
+      "grad_norm": 1.2027863264083862,
+      "learning_rate": 0.00012276923090302844,
+      "loss": 0.3529,
+      "step": 6040
+    },
+    {
+      "epoch": 0.44422383998823445,
+      "grad_norm": 1.8517543077468872,
+      "learning_rate": 0.00012274603965137387,
+      "loss": 0.3491,
+      "step": 6041
+    },
+    {
+      "epoch": 0.4442973748069711,
+      "grad_norm": 1.7216517925262451,
+      "learning_rate": 0.0001227228471095399,
+      "loss": 0.4023,
+      "step": 6042
+    },
+    {
+      "epoch": 0.4443709096257078,
+      "grad_norm": 1.5152612924575806,
+      "learning_rate": 0.00012269965327884209,
+      "loss": 0.3837,
+      "step": 6043
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 1.5946229696273804,
+      "learning_rate": 0.00012267645816059597,
+      "loss": 0.3398,
+      "step": 6044
+    },
+    {
+      "epoch": 0.4445179792631811,
+      "grad_norm": 1.2698004245758057,
+      "learning_rate": 0.0001226532617561172,
+      "loss": 0.3693,
+      "step": 6045
+    },
+    {
+      "epoch": 0.44459151408191777,
+      "grad_norm": 1.3565841913223267,
+      "learning_rate": 0.00012263006406672156,
+      "loss": 0.3755,
+      "step": 6046
+    },
+    {
+      "epoch": 0.44466504890065445,
+      "grad_norm": 1.3603850603103638,
+      "learning_rate": 0.00012260686509372478,
+      "loss": 0.4128,
+      "step": 6047
+    },
+    {
+      "epoch": 0.4447385837193911,
+      "grad_norm": 1.0426996946334839,
+      "learning_rate": 0.00012258366483844274,
+      "loss": 0.3979,
+      "step": 6048
+    },
+    {
+      "epoch": 0.4448121185381278,
+      "grad_norm": 1.7196707725524902,
+      "learning_rate": 0.0001225604633021914,
+      "loss": 0.3686,
+      "step": 6049
+    },
+    {
+      "epoch": 0.4448856533568645,
+      "grad_norm": 1.3262958526611328,
+      "learning_rate": 0.0001225372604862868,
+      "loss": 0.3893,
+      "step": 6050
+    },
+    {
+      "epoch": 0.44495918817560115,
+      "grad_norm": 1.7946810722351074,
+      "learning_rate": 0.00012251405639204495,
+      "loss": 0.3572,
+      "step": 6051
+    },
+    {
+      "epoch": 0.4450327229943378,
+      "grad_norm": 1.754517674446106,
+      "learning_rate": 0.0001224908510207821,
+      "loss": 0.3759,
+      "step": 6052
+    },
+    {
+      "epoch": 0.4451062578130745,
+      "grad_norm": 0.9772958159446716,
+      "learning_rate": 0.00012246764437381444,
+      "loss": 0.3763,
+      "step": 6053
+    },
+    {
+      "epoch": 0.4451797926318112,
+      "grad_norm": 1.2089987993240356,
+      "learning_rate": 0.00012244443645245828,
+      "loss": 0.3696,
+      "step": 6054
+    },
+    {
+      "epoch": 0.44525332745054785,
+      "grad_norm": 1.4767966270446777,
+      "learning_rate": 0.00012242122725803,
+      "loss": 0.3735,
+      "step": 6055
+    },
+    {
+      "epoch": 0.44532686226928453,
+      "grad_norm": 1.1622192859649658,
+      "learning_rate": 0.00012239801679184603,
+      "loss": 0.3494,
+      "step": 6056
+    },
+    {
+      "epoch": 0.4454003970880212,
+      "grad_norm": 1.6696099042892456,
+      "learning_rate": 0.00012237480505522292,
+      "loss": 0.3745,
+      "step": 6057
+    },
+    {
+      "epoch": 0.4454739319067578,
+      "grad_norm": 1.612338900566101,
+      "learning_rate": 0.00012235159204947727,
+      "loss": 0.3561,
+      "step": 6058
+    },
+    {
+      "epoch": 0.4455474667254945,
+      "grad_norm": 1.1547735929489136,
+      "learning_rate": 0.00012232837777592572,
+      "loss": 0.3576,
+      "step": 6059
+    },
+    {
+      "epoch": 0.4456210015442312,
+      "grad_norm": 1.1671134233474731,
+      "learning_rate": 0.000122305162235885,
+      "loss": 0.3592,
+      "step": 6060
+    },
+    {
+      "epoch": 0.44569453636296785,
+      "grad_norm": 1.6042696237564087,
+      "learning_rate": 0.00012228194543067195,
+      "loss": 0.3403,
+      "step": 6061
+    },
+    {
+      "epoch": 0.44576807118170453,
+      "grad_norm": 1.5368865728378296,
+      "learning_rate": 0.00012225872736160344,
+      "loss": 0.365,
+      "step": 6062
+    },
+    {
+      "epoch": 0.4458416060004412,
+      "grad_norm": 1.6172103881835938,
+      "learning_rate": 0.00012223550802999643,
+      "loss": 0.3489,
+      "step": 6063
+    },
+    {
+      "epoch": 0.4459151408191779,
+      "grad_norm": 1.4274412393569946,
+      "learning_rate": 0.00012221228743716794,
+      "loss": 0.3504,
+      "step": 6064
+    },
+    {
+      "epoch": 0.44598867563791456,
+      "grad_norm": 1.4510018825531006,
+      "learning_rate": 0.00012218906558443505,
+      "loss": 0.4052,
+      "step": 6065
+    },
+    {
+      "epoch": 0.44606221045665123,
+      "grad_norm": 1.1614375114440918,
+      "learning_rate": 0.00012216584247311496,
+      "loss": 0.3629,
+      "step": 6066
+    },
+    {
+      "epoch": 0.4461357452753879,
+      "grad_norm": 1.7109190225601196,
+      "learning_rate": 0.00012214261810452486,
+      "loss": 0.3535,
+      "step": 6067
+    },
+    {
+      "epoch": 0.4462092800941246,
+      "grad_norm": 1.535132884979248,
+      "learning_rate": 0.00012211939247998215,
+      "loss": 0.3586,
+      "step": 6068
+    },
+    {
+      "epoch": 0.44628281491286126,
+      "grad_norm": 1.2948333024978638,
+      "learning_rate": 0.00012209616560080412,
+      "loss": 0.376,
+      "step": 6069
+    },
+    {
+      "epoch": 0.44635634973159793,
+      "grad_norm": 1.3390389680862427,
+      "learning_rate": 0.00012207293746830826,
+      "loss": 0.3659,
+      "step": 6070
+    },
+    {
+      "epoch": 0.4464298845503346,
+      "grad_norm": 1.3358814716339111,
+      "learning_rate": 0.00012204970808381208,
+      "loss": 0.3962,
+      "step": 6071
+    },
+    {
+      "epoch": 0.44650341936907123,
+      "grad_norm": 1.0807411670684814,
+      "learning_rate": 0.00012202647744863323,
+      "loss": 0.3735,
+      "step": 6072
+    },
+    {
+      "epoch": 0.4465769541878079,
+      "grad_norm": 1.7479511499404907,
+      "learning_rate": 0.00012200324556408929,
+      "loss": 0.3655,
+      "step": 6073
+    },
+    {
+      "epoch": 0.4466504890065446,
+      "grad_norm": 1.551942229270935,
+      "learning_rate": 0.00012198001243149806,
+      "loss": 0.3592,
+      "step": 6074
+    },
+    {
+      "epoch": 0.44672402382528126,
+      "grad_norm": 1.4618464708328247,
+      "learning_rate": 0.00012195677805217732,
+      "loss": 0.3705,
+      "step": 6075
+    },
+    {
+      "epoch": 0.44679755864401793,
+      "grad_norm": 1.2451074123382568,
+      "learning_rate": 0.00012193354242744495,
+      "loss": 0.422,
+      "step": 6076
+    },
+    {
+      "epoch": 0.4468710934627546,
+      "grad_norm": 1.3355586528778076,
+      "learning_rate": 0.00012191030555861894,
+      "loss": 0.351,
+      "step": 6077
+    },
+    {
+      "epoch": 0.4469446282814913,
+      "grad_norm": 1.3274075984954834,
+      "learning_rate": 0.00012188706744701723,
+      "loss": 0.3927,
+      "step": 6078
+    },
+    {
+      "epoch": 0.44701816310022796,
+      "grad_norm": 1.4909077882766724,
+      "learning_rate": 0.000121863828093958,
+      "loss": 0.3859,
+      "step": 6079
+    },
+    {
+      "epoch": 0.44709169791896464,
+      "grad_norm": 1.4274543523788452,
+      "learning_rate": 0.00012184058750075933,
+      "loss": 0.3684,
+      "step": 6080
+    },
+    {
+      "epoch": 0.4471652327377013,
+      "grad_norm": 1.5126034021377563,
+      "learning_rate": 0.0001218173456687395,
+      "loss": 0.3526,
+      "step": 6081
+    },
+    {
+      "epoch": 0.447238767556438,
+      "grad_norm": 1.3694792985916138,
+      "learning_rate": 0.00012179410259921682,
+      "loss": 0.3903,
+      "step": 6082
+    },
+    {
+      "epoch": 0.44731230237517466,
+      "grad_norm": 1.2333742380142212,
+      "learning_rate": 0.00012177085829350961,
+      "loss": 0.3572,
+      "step": 6083
+    },
+    {
+      "epoch": 0.44738583719391134,
+      "grad_norm": 1.267262578010559,
+      "learning_rate": 0.00012174761275293636,
+      "loss": 0.3543,
+      "step": 6084
+    },
+    {
+      "epoch": 0.447459372012648,
+      "grad_norm": 1.525601863861084,
+      "learning_rate": 0.00012172436597881556,
+      "loss": 0.4024,
+      "step": 6085
+    },
+    {
+      "epoch": 0.44753290683138464,
+      "grad_norm": 1.3925718069076538,
+      "learning_rate": 0.0001217011179724658,
+      "loss": 0.3959,
+      "step": 6086
+    },
+    {
+      "epoch": 0.4476064416501213,
+      "grad_norm": 1.4545295238494873,
+      "learning_rate": 0.00012167786873520573,
+      "loss": 0.3898,
+      "step": 6087
+    },
+    {
+      "epoch": 0.447679976468858,
+      "grad_norm": 1.2093968391418457,
+      "learning_rate": 0.00012165461826835403,
+      "loss": 0.3948,
+      "step": 6088
+    },
+    {
+      "epoch": 0.44775351128759466,
+      "grad_norm": 1.5673469305038452,
+      "learning_rate": 0.00012163136657322955,
+      "loss": 0.4047,
+      "step": 6089
+    },
+    {
+      "epoch": 0.44782704610633134,
+      "grad_norm": 1.000958800315857,
+      "learning_rate": 0.00012160811365115114,
+      "loss": 0.3295,
+      "step": 6090
+    },
+    {
+      "epoch": 0.447900580925068,
+      "grad_norm": 1.6868795156478882,
+      "learning_rate": 0.00012158485950343772,
+      "loss": 0.3697,
+      "step": 6091
+    },
+    {
+      "epoch": 0.4479741157438047,
+      "grad_norm": 1.1949756145477295,
+      "learning_rate": 0.0001215616041314083,
+      "loss": 0.3598,
+      "step": 6092
+    },
+    {
+      "epoch": 0.44804765056254137,
+      "grad_norm": 1.726575493812561,
+      "learning_rate": 0.00012153834753638192,
+      "loss": 0.3589,
+      "step": 6093
+    },
+    {
+      "epoch": 0.44812118538127804,
+      "grad_norm": 1.469313383102417,
+      "learning_rate": 0.00012151508971967774,
+      "loss": 0.3796,
+      "step": 6094
+    },
+    {
+      "epoch": 0.4481947202000147,
+      "grad_norm": 1.1451746225357056,
+      "learning_rate": 0.00012149183068261496,
+      "loss": 0.3552,
+      "step": 6095
+    },
+    {
+      "epoch": 0.4482682550187514,
+      "grad_norm": 1.2884976863861084,
+      "learning_rate": 0.00012146857042651285,
+      "loss": 0.409,
+      "step": 6096
+    },
+    {
+      "epoch": 0.44834178983748807,
+      "grad_norm": 1.4108457565307617,
+      "learning_rate": 0.00012144530895269082,
+      "loss": 0.3431,
+      "step": 6097
+    },
+    {
+      "epoch": 0.44841532465622475,
+      "grad_norm": 1.079412817955017,
+      "learning_rate": 0.00012142204626246818,
+      "loss": 0.3667,
+      "step": 6098
+    },
+    {
+      "epoch": 0.4484888594749614,
+      "grad_norm": 1.6017910242080688,
+      "learning_rate": 0.0001213987823571645,
+      "loss": 0.3653,
+      "step": 6099
+    },
+    {
+      "epoch": 0.44856239429369804,
+      "grad_norm": 1.4347933530807495,
+      "learning_rate": 0.0001213755172380993,
+      "loss": 0.3942,
+      "step": 6100
+    },
+    {
+      "epoch": 0.4486359291124347,
+      "grad_norm": 2.0026209354400635,
+      "learning_rate": 0.00012135225090659217,
+      "loss": 0.3421,
+      "step": 6101
+    },
+    {
+      "epoch": 0.4487094639311714,
+      "grad_norm": 2.021918773651123,
+      "learning_rate": 0.0001213289833639629,
+      "loss": 0.3572,
+      "step": 6102
+    },
+    {
+      "epoch": 0.44878299874990807,
+      "grad_norm": 0.9357815980911255,
+      "learning_rate": 0.00012130571461153113,
+      "loss": 0.4051,
+      "step": 6103
+    },
+    {
+      "epoch": 0.44885653356864474,
+      "grad_norm": 1.2569633722305298,
+      "learning_rate": 0.00012128244465061675,
+      "loss": 0.3571,
+      "step": 6104
+    },
+    {
+      "epoch": 0.4489300683873814,
+      "grad_norm": 0.9906185269355774,
+      "learning_rate": 0.00012125917348253965,
+      "loss": 0.3644,
+      "step": 6105
+    },
+    {
+      "epoch": 0.4490036032061181,
+      "grad_norm": 0.7510443329811096,
+      "learning_rate": 0.0001212359011086198,
+      "loss": 0.4115,
+      "step": 6106
+    },
+    {
+      "epoch": 0.44907713802485477,
+      "grad_norm": 1.1904433965682983,
+      "learning_rate": 0.00012121262753017724,
+      "loss": 0.4051,
+      "step": 6107
+    },
+    {
+      "epoch": 0.44915067284359145,
+      "grad_norm": 0.8315442800521851,
+      "learning_rate": 0.00012118935274853205,
+      "loss": 0.3785,
+      "step": 6108
+    },
+    {
+      "epoch": 0.4492242076623281,
+      "grad_norm": 0.6125947833061218,
+      "learning_rate": 0.0001211660767650044,
+      "loss": 0.3741,
+      "step": 6109
+    },
+    {
+      "epoch": 0.4492977424810648,
+      "grad_norm": 1.1456331014633179,
+      "learning_rate": 0.00012114279958091457,
+      "loss": 0.3592,
+      "step": 6110
+    },
+    {
+      "epoch": 0.4493712772998015,
+      "grad_norm": 0.5216671228408813,
+      "learning_rate": 0.00012111952119758279,
+      "loss": 0.3673,
+      "step": 6111
+    },
+    {
+      "epoch": 0.44944481211853815,
+      "grad_norm": 0.8468093276023865,
+      "learning_rate": 0.0001210962416163295,
+      "loss": 0.359,
+      "step": 6112
+    },
+    {
+      "epoch": 0.4495183469372748,
+      "grad_norm": 0.6058722734451294,
+      "learning_rate": 0.00012107296083847513,
+      "loss": 0.3727,
+      "step": 6113
+    },
+    {
+      "epoch": 0.44959188175601145,
+      "grad_norm": 0.6687707304954529,
+      "learning_rate": 0.00012104967886534018,
+      "loss": 0.3473,
+      "step": 6114
+    },
+    {
+      "epoch": 0.4496654165747481,
+      "grad_norm": 0.6553505063056946,
+      "learning_rate": 0.00012102639569824524,
+      "loss": 0.3886,
+      "step": 6115
+    },
+    {
+      "epoch": 0.4497389513934848,
+      "grad_norm": 0.6527856588363647,
+      "learning_rate": 0.0001210031113385109,
+      "loss": 0.3684,
+      "step": 6116
+    },
+    {
+      "epoch": 0.4498124862122215,
+      "grad_norm": 0.9310654997825623,
+      "learning_rate": 0.00012097982578745797,
+      "loss": 0.3615,
+      "step": 6117
+    },
+    {
+      "epoch": 0.44988602103095815,
+      "grad_norm": 0.6004503965377808,
+      "learning_rate": 0.00012095653904640716,
+      "loss": 0.389,
+      "step": 6118
+    },
+    {
+      "epoch": 0.4499595558496948,
+      "grad_norm": 0.587963342666626,
+      "learning_rate": 0.00012093325111667932,
+      "loss": 0.3713,
+      "step": 6119
+    },
+    {
+      "epoch": 0.4500330906684315,
+      "grad_norm": 0.5486131906509399,
+      "learning_rate": 0.00012090996199959544,
+      "loss": 0.3721,
+      "step": 6120
+    },
+    {
+      "epoch": 0.4501066254871682,
+      "grad_norm": 0.46291878819465637,
+      "learning_rate": 0.00012088667169647639,
+      "loss": 0.3664,
+      "step": 6121
+    },
+    {
+      "epoch": 0.45018016030590485,
+      "grad_norm": 0.5127429962158203,
+      "learning_rate": 0.0001208633802086433,
+      "loss": 0.3878,
+      "step": 6122
+    },
+    {
+      "epoch": 0.45025369512464153,
+      "grad_norm": 1.0055522918701172,
+      "learning_rate": 0.00012084008753741727,
+      "loss": 0.3566,
+      "step": 6123
+    },
+    {
+      "epoch": 0.4503272299433782,
+      "grad_norm": 0.570457398891449,
+      "learning_rate": 0.00012081679368411946,
+      "loss": 0.3506,
+      "step": 6124
+    },
+    {
+      "epoch": 0.4504007647621149,
+      "grad_norm": 0.7130426168441772,
+      "learning_rate": 0.00012079349865007117,
+      "loss": 0.3526,
+      "step": 6125
+    },
+    {
+      "epoch": 0.45047429958085156,
+      "grad_norm": 0.8721097707748413,
+      "learning_rate": 0.00012077020243659369,
+      "loss": 0.3505,
+      "step": 6126
+    },
+    {
+      "epoch": 0.45054783439958823,
+      "grad_norm": 0.5523410439491272,
+      "learning_rate": 0.0001207469050450084,
+      "loss": 0.3633,
+      "step": 6127
+    },
+    {
+      "epoch": 0.45062136921832485,
+      "grad_norm": 0.6443420648574829,
+      "learning_rate": 0.00012072360647663679,
+      "loss": 0.3976,
+      "step": 6128
+    },
+    {
+      "epoch": 0.45069490403706153,
+      "grad_norm": 0.6700137257575989,
+      "learning_rate": 0.00012070030673280031,
+      "loss": 0.3816,
+      "step": 6129
+    },
+    {
+      "epoch": 0.4507684388557982,
+      "grad_norm": 0.4705854058265686,
+      "learning_rate": 0.0001206770058148206,
+      "loss": 0.372,
+      "step": 6130
+    },
+    {
+      "epoch": 0.4508419736745349,
+      "grad_norm": 0.7872517704963684,
+      "learning_rate": 0.0001206537037240193,
+      "loss": 0.3826,
+      "step": 6131
+    },
+    {
+      "epoch": 0.45091550849327156,
+      "grad_norm": 0.9612182378768921,
+      "learning_rate": 0.00012063040046171814,
+      "loss": 0.4002,
+      "step": 6132
+    },
+    {
+      "epoch": 0.45098904331200823,
+      "grad_norm": 0.5080069303512573,
+      "learning_rate": 0.00012060709602923887,
+      "loss": 0.3502,
+      "step": 6133
+    },
+    {
+      "epoch": 0.4510625781307449,
+      "grad_norm": 0.7458832263946533,
+      "learning_rate": 0.00012058379042790338,
+      "loss": 0.3558,
+      "step": 6134
+    },
+    {
+      "epoch": 0.4511361129494816,
+      "grad_norm": 0.6548380851745605,
+      "learning_rate": 0.00012056048365903356,
+      "loss": 0.3572,
+      "step": 6135
+    },
+    {
+      "epoch": 0.45120964776821826,
+      "grad_norm": 0.6683770418167114,
+      "learning_rate": 0.00012053717572395142,
+      "loss": 0.3644,
+      "step": 6136
+    },
+    {
+      "epoch": 0.45128318258695493,
+      "grad_norm": 0.8044237494468689,
+      "learning_rate": 0.00012051386662397902,
+      "loss": 0.4013,
+      "step": 6137
+    },
+    {
+      "epoch": 0.4513567174056916,
+      "grad_norm": 0.5429604053497314,
+      "learning_rate": 0.00012049055636043843,
+      "loss": 0.423,
+      "step": 6138
+    },
+    {
+      "epoch": 0.4514302522244283,
+      "grad_norm": 0.8461694121360779,
+      "learning_rate": 0.00012046724493465185,
+      "loss": 0.3851,
+      "step": 6139
+    },
+    {
+      "epoch": 0.45150378704316496,
+      "grad_norm": 0.6063132882118225,
+      "learning_rate": 0.00012044393234794155,
+      "loss": 0.3671,
+      "step": 6140
+    },
+    {
+      "epoch": 0.45157732186190164,
+      "grad_norm": 0.762172281742096,
+      "learning_rate": 0.00012042061860162986,
+      "loss": 0.3666,
+      "step": 6141
+    },
+    {
+      "epoch": 0.45165085668063826,
+      "grad_norm": 0.6041750311851501,
+      "learning_rate": 0.00012039730369703908,
+      "loss": 0.3344,
+      "step": 6142
+    },
+    {
+      "epoch": 0.45172439149937493,
+      "grad_norm": 0.6230520606040955,
+      "learning_rate": 0.00012037398763549175,
+      "loss": 0.3817,
+      "step": 6143
+    },
+    {
+      "epoch": 0.4517979263181116,
+      "grad_norm": 0.6350012421607971,
+      "learning_rate": 0.0001203506704183103,
+      "loss": 0.3664,
+      "step": 6144
+    },
+    {
+      "epoch": 0.4518714611368483,
+      "grad_norm": 0.45818042755126953,
+      "learning_rate": 0.00012032735204681737,
+      "loss": 0.3729,
+      "step": 6145
+    },
+    {
+      "epoch": 0.45194499595558496,
+      "grad_norm": 0.47990089654922485,
+      "learning_rate": 0.00012030403252233559,
+      "loss": 0.3786,
+      "step": 6146
+    },
+    {
+      "epoch": 0.45201853077432164,
+      "grad_norm": 0.4941905438899994,
+      "learning_rate": 0.00012028071184618763,
+      "loss": 0.3424,
+      "step": 6147
+    },
+    {
+      "epoch": 0.4520920655930583,
+      "grad_norm": 0.5931134819984436,
+      "learning_rate": 0.00012025739001969632,
+      "loss": 0.3433,
+      "step": 6148
+    },
+    {
+      "epoch": 0.452165600411795,
+      "grad_norm": 0.5385184288024902,
+      "learning_rate": 0.00012023406704418447,
+      "loss": 0.3581,
+      "step": 6149
+    },
+    {
+      "epoch": 0.45223913523053166,
+      "grad_norm": 0.6839372515678406,
+      "learning_rate": 0.00012021074292097496,
+      "loss": 0.3567,
+      "step": 6150
+    },
+    {
+      "epoch": 0.45231267004926834,
+      "grad_norm": 0.8475883603096008,
+      "learning_rate": 0.00012018741765139081,
+      "loss": 0.3534,
+      "step": 6151
+    },
+    {
+      "epoch": 0.452386204868005,
+      "grad_norm": 0.49836936593055725,
+      "learning_rate": 0.00012016409123675498,
+      "loss": 0.3627,
+      "step": 6152
+    },
+    {
+      "epoch": 0.4524597396867417,
+      "grad_norm": 0.717572033405304,
+      "learning_rate": 0.00012014076367839066,
+      "loss": 0.3458,
+      "step": 6153
+    },
+    {
+      "epoch": 0.45253327450547837,
+      "grad_norm": 1.1096607446670532,
+      "learning_rate": 0.00012011743497762095,
+      "loss": 0.3816,
+      "step": 6154
+    },
+    {
+      "epoch": 0.45260680932421504,
+      "grad_norm": 0.7983670830726624,
+      "learning_rate": 0.0001200941051357691,
+      "loss": 0.4013,
+      "step": 6155
+    },
+    {
+      "epoch": 0.45268034414295166,
+      "grad_norm": 0.6045268177986145,
+      "learning_rate": 0.00012007077415415842,
+      "loss": 0.3882,
+      "step": 6156
+    },
+    {
+      "epoch": 0.45275387896168834,
+      "grad_norm": 0.5852932333946228,
+      "learning_rate": 0.0001200474420341122,
+      "loss": 0.3638,
+      "step": 6157
+    },
+    {
+      "epoch": 0.452827413780425,
+      "grad_norm": 0.818597674369812,
+      "learning_rate": 0.00012002410877695396,
+      "loss": 0.3774,
+      "step": 6158
+    },
+    {
+      "epoch": 0.4529009485991617,
+      "grad_norm": 1.0860739946365356,
+      "learning_rate": 0.00012000077438400712,
+      "loss": 0.3815,
+      "step": 6159
+    },
+    {
+      "epoch": 0.45297448341789837,
+      "grad_norm": 0.9441291093826294,
+      "learning_rate": 0.00011997743885659522,
+      "loss": 0.3445,
+      "step": 6160
+    },
+    {
+      "epoch": 0.45304801823663504,
+      "grad_norm": 0.8071250915527344,
+      "learning_rate": 0.00011995410219604198,
+      "loss": 0.3524,
+      "step": 6161
+    },
+    {
+      "epoch": 0.4531215530553717,
+      "grad_norm": 0.7014995813369751,
+      "learning_rate": 0.00011993076440367092,
+      "loss": 0.344,
+      "step": 6162
+    },
+    {
+      "epoch": 0.4531950878741084,
+      "grad_norm": 0.8541041016578674,
+      "learning_rate": 0.00011990742548080592,
+      "loss": 0.356,
+      "step": 6163
+    },
+    {
+      "epoch": 0.45326862269284507,
+      "grad_norm": 1.1164796352386475,
+      "learning_rate": 0.00011988408542877074,
+      "loss": 0.3438,
+      "step": 6164
+    },
+    {
+      "epoch": 0.45334215751158174,
+      "grad_norm": 0.7705814242362976,
+      "learning_rate": 0.0001198607442488892,
+      "loss": 0.3404,
+      "step": 6165
+    },
+    {
+      "epoch": 0.4534156923303184,
+      "grad_norm": 0.9493677616119385,
+      "learning_rate": 0.00011983740194248534,
+      "loss": 0.364,
+      "step": 6166
+    },
+    {
+      "epoch": 0.4534892271490551,
+      "grad_norm": 0.6628478765487671,
+      "learning_rate": 0.00011981405851088307,
+      "loss": 0.3793,
+      "step": 6167
+    },
+    {
+      "epoch": 0.45356276196779177,
+      "grad_norm": 0.6380237936973572,
+      "learning_rate": 0.00011979071395540653,
+      "loss": 0.3638,
+      "step": 6168
+    },
+    {
+      "epoch": 0.45363629678652845,
+      "grad_norm": 0.7542129158973694,
+      "learning_rate": 0.0001197673682773798,
+      "loss": 0.3511,
+      "step": 6169
+    },
+    {
+      "epoch": 0.45370983160526507,
+      "grad_norm": 0.5334354639053345,
+      "learning_rate": 0.00011974402147812704,
+      "loss": 0.348,
+      "step": 6170
+    },
+    {
+      "epoch": 0.45378336642400174,
+      "grad_norm": 0.7543604969978333,
+      "learning_rate": 0.00011972067355897258,
+      "loss": 0.3659,
+      "step": 6171
+    },
+    {
+      "epoch": 0.4538569012427384,
+      "grad_norm": 0.6270601749420166,
+      "learning_rate": 0.00011969732452124069,
+      "loss": 0.3405,
+      "step": 6172
+    },
+    {
+      "epoch": 0.4539304360614751,
+      "grad_norm": 0.593454122543335,
+      "learning_rate": 0.00011967397436625577,
+      "loss": 0.3753,
+      "step": 6173
+    },
+    {
+      "epoch": 0.45400397088021177,
+      "grad_norm": 0.5589761734008789,
+      "learning_rate": 0.00011965062309534227,
+      "loss": 0.3622,
+      "step": 6174
+    },
+    {
+      "epoch": 0.45407750569894845,
+      "grad_norm": 0.5700258612632751,
+      "learning_rate": 0.00011962727070982467,
+      "loss": 0.3971,
+      "step": 6175
+    },
+    {
+      "epoch": 0.4541510405176851,
+      "grad_norm": 0.5637813806533813,
+      "learning_rate": 0.00011960391721102759,
+      "loss": 0.3767,
+      "step": 6176
+    },
+    {
+      "epoch": 0.4542245753364218,
+      "grad_norm": 0.5582564473152161,
+      "learning_rate": 0.00011958056260027561,
+      "loss": 0.3434,
+      "step": 6177
+    },
+    {
+      "epoch": 0.4542981101551585,
+      "grad_norm": 0.4832659065723419,
+      "learning_rate": 0.00011955720687889349,
+      "loss": 0.3586,
+      "step": 6178
+    },
+    {
+      "epoch": 0.45437164497389515,
+      "grad_norm": 0.6175673007965088,
+      "learning_rate": 0.00011953385004820592,
+      "loss": 0.3647,
+      "step": 6179
+    },
+    {
+      "epoch": 0.4544451797926318,
+      "grad_norm": 0.5804812908172607,
+      "learning_rate": 0.00011951049210953776,
+      "loss": 0.3503,
+      "step": 6180
+    },
+    {
+      "epoch": 0.4545187146113685,
+      "grad_norm": 0.4833614230155945,
+      "learning_rate": 0.0001194871330642139,
+      "loss": 0.3538,
+      "step": 6181
+    },
+    {
+      "epoch": 0.4545922494301052,
+      "grad_norm": 0.6926799416542053,
+      "learning_rate": 0.00011946377291355932,
+      "loss": 0.4087,
+      "step": 6182
+    },
+    {
+      "epoch": 0.45466578424884185,
+      "grad_norm": 0.7855358123779297,
+      "learning_rate": 0.00011944041165889897,
+      "loss": 0.3984,
+      "step": 6183
+    },
+    {
+      "epoch": 0.4547393190675785,
+      "grad_norm": 1.0643057823181152,
+      "learning_rate": 0.00011941704930155797,
+      "loss": 0.3936,
+      "step": 6184
+    },
+    {
+      "epoch": 0.45481285388631515,
+      "grad_norm": 1.1821717023849487,
+      "learning_rate": 0.00011939368584286144,
+      "loss": 0.3633,
+      "step": 6185
+    },
+    {
+      "epoch": 0.4548863887050518,
+      "grad_norm": 0.5782965421676636,
+      "learning_rate": 0.00011937032128413457,
+      "loss": 0.3565,
+      "step": 6186
+    },
+    {
+      "epoch": 0.4549599235237885,
+      "grad_norm": 1.56307053565979,
+      "learning_rate": 0.00011934695562670264,
+      "loss": 0.3459,
+      "step": 6187
+    },
+    {
+      "epoch": 0.4550334583425252,
+      "grad_norm": 0.5923351645469666,
+      "learning_rate": 0.00011932358887189094,
+      "loss": 0.3727,
+      "step": 6188
+    },
+    {
+      "epoch": 0.45510699316126185,
+      "grad_norm": 1.3372844457626343,
+      "learning_rate": 0.00011930022102102494,
+      "loss": 0.3657,
+      "step": 6189
+    },
+    {
+      "epoch": 0.45518052797999853,
+      "grad_norm": 0.9065720438957214,
+      "learning_rate": 0.00011927685207543,
+      "loss": 0.3649,
+      "step": 6190
+    },
+    {
+      "epoch": 0.4552540627987352,
+      "grad_norm": 0.6095065474510193,
+      "learning_rate": 0.00011925348203643169,
+      "loss": 0.3778,
+      "step": 6191
+    },
+    {
+      "epoch": 0.4553275976174719,
+      "grad_norm": 0.4915781617164612,
+      "learning_rate": 0.00011923011090535555,
+      "loss": 0.391,
+      "step": 6192
+    },
+    {
+      "epoch": 0.45540113243620856,
+      "grad_norm": 0.6706447601318359,
+      "learning_rate": 0.00011920673868352721,
+      "loss": 0.3363,
+      "step": 6193
+    },
+    {
+      "epoch": 0.45547466725494523,
+      "grad_norm": 0.6309241652488708,
+      "learning_rate": 0.0001191833653722724,
+      "loss": 0.3657,
+      "step": 6194
+    },
+    {
+      "epoch": 0.4555482020736819,
+      "grad_norm": 0.5612021088600159,
+      "learning_rate": 0.00011915999097291686,
+      "loss": 0.389,
+      "step": 6195
+    },
+    {
+      "epoch": 0.4556217368924186,
+      "grad_norm": 0.6886342167854309,
+      "learning_rate": 0.00011913661548678641,
+      "loss": 0.3548,
+      "step": 6196
+    },
+    {
+      "epoch": 0.45569527171115526,
+      "grad_norm": 1.0993925333023071,
+      "learning_rate": 0.00011911323891520694,
+      "loss": 0.3406,
+      "step": 6197
+    },
+    {
+      "epoch": 0.4557688065298919,
+      "grad_norm": 0.7125188112258911,
+      "learning_rate": 0.00011908986125950435,
+      "loss": 0.3438,
+      "step": 6198
+    },
+    {
+      "epoch": 0.45584234134862855,
+      "grad_norm": 0.8485866785049438,
+      "learning_rate": 0.00011906648252100472,
+      "loss": 0.3943,
+      "step": 6199
+    },
+    {
+      "epoch": 0.45591587616736523,
+      "grad_norm": 0.9397155046463013,
+      "learning_rate": 0.00011904310270103407,
+      "loss": 0.3426,
+      "step": 6200
+    },
+    {
+      "epoch": 0.4559894109861019,
+      "grad_norm": 0.4765337109565735,
+      "learning_rate": 0.00011901972180091852,
+      "loss": 0.3877,
+      "step": 6201
+    },
+    {
+      "epoch": 0.4560629458048386,
+      "grad_norm": 1.0392591953277588,
+      "learning_rate": 0.00011899633982198432,
+      "loss": 0.3462,
+      "step": 6202
+    },
+    {
+      "epoch": 0.45613648062357526,
+      "grad_norm": 0.989059567451477,
+      "learning_rate": 0.00011897295676555765,
+      "loss": 0.3547,
+      "step": 6203
+    },
+    {
+      "epoch": 0.45621001544231193,
+      "grad_norm": 0.5425757765769958,
+      "learning_rate": 0.00011894957263296484,
+      "loss": 0.3462,
+      "step": 6204
+    },
+    {
+      "epoch": 0.4562835502610486,
+      "grad_norm": 0.7110049724578857,
+      "learning_rate": 0.0001189261874255323,
+      "loss": 0.3517,
+      "step": 6205
+    },
+    {
+      "epoch": 0.4563570850797853,
+      "grad_norm": 0.7386878132820129,
+      "learning_rate": 0.0001189028011445864,
+      "loss": 0.3477,
+      "step": 6206
+    },
+    {
+      "epoch": 0.45643061989852196,
+      "grad_norm": 0.7724016904830933,
+      "learning_rate": 0.00011887941379145371,
+      "loss": 0.3424,
+      "step": 6207
+    },
+    {
+      "epoch": 0.45650415471725864,
+      "grad_norm": 0.6281265616416931,
+      "learning_rate": 0.00011885602536746072,
+      "loss": 0.3768,
+      "step": 6208
+    },
+    {
+      "epoch": 0.4565776895359953,
+      "grad_norm": 0.5813159346580505,
+      "learning_rate": 0.00011883263587393409,
+      "loss": 0.3533,
+      "step": 6209
+    },
+    {
+      "epoch": 0.456651224354732,
+      "grad_norm": 0.6054369211196899,
+      "learning_rate": 0.00011880924531220045,
+      "loss": 0.3692,
+      "step": 6210
+    },
+    {
+      "epoch": 0.45672475917346866,
+      "grad_norm": 0.4462224245071411,
+      "learning_rate": 0.00011878585368358659,
+      "loss": 0.3513,
+      "step": 6211
+    },
+    {
+      "epoch": 0.4567982939922053,
+      "grad_norm": 0.6805505156517029,
+      "learning_rate": 0.00011876246098941927,
+      "loss": 0.3876,
+      "step": 6212
+    },
+    {
+      "epoch": 0.45687182881094196,
+      "grad_norm": 1.034709095954895,
+      "learning_rate": 0.00011873906723102537,
+      "loss": 0.3619,
+      "step": 6213
+    },
+    {
+      "epoch": 0.45694536362967864,
+      "grad_norm": 0.9891403317451477,
+      "learning_rate": 0.00011871567240973181,
+      "loss": 0.3484,
+      "step": 6214
+    },
+    {
+      "epoch": 0.4570188984484153,
+      "grad_norm": 0.5528017282485962,
+      "learning_rate": 0.00011869227652686554,
+      "loss": 0.3567,
+      "step": 6215
+    },
+    {
+      "epoch": 0.457092433267152,
+      "grad_norm": 0.4463767409324646,
+      "learning_rate": 0.00011866887958375361,
+      "loss": 0.3667,
+      "step": 6216
+    },
+    {
+      "epoch": 0.45716596808588866,
+      "grad_norm": 0.5159367918968201,
+      "learning_rate": 0.00011864548158172315,
+      "loss": 0.3598,
+      "step": 6217
+    },
+    {
+      "epoch": 0.45723950290462534,
+      "grad_norm": 0.5123791694641113,
+      "learning_rate": 0.0001186220825221013,
+      "loss": 0.3558,
+      "step": 6218
+    },
+    {
+      "epoch": 0.457313037723362,
+      "grad_norm": 0.4317947030067444,
+      "learning_rate": 0.00011859868240621528,
+      "loss": 0.3656,
+      "step": 6219
+    },
+    {
+      "epoch": 0.4573865725420987,
+      "grad_norm": 0.5213778614997864,
+      "learning_rate": 0.00011857528123539236,
+      "loss": 0.3525,
+      "step": 6220
+    },
+    {
+      "epoch": 0.45746010736083537,
+      "grad_norm": 0.51774001121521,
+      "learning_rate": 0.00011855187901095986,
+      "loss": 0.336,
+      "step": 6221
+    },
+    {
+      "epoch": 0.45753364217957204,
+      "grad_norm": 0.5871605277061462,
+      "learning_rate": 0.00011852847573424521,
+      "loss": 0.3441,
+      "step": 6222
+    },
+    {
+      "epoch": 0.4576071769983087,
+      "grad_norm": 0.8633671402931213,
+      "learning_rate": 0.0001185050714065759,
+      "loss": 0.367,
+      "step": 6223
+    },
+    {
+      "epoch": 0.4576807118170454,
+      "grad_norm": 0.6594526171684265,
+      "learning_rate": 0.00011848166602927937,
+      "loss": 0.3472,
+      "step": 6224
+    },
+    {
+      "epoch": 0.45775424663578207,
+      "grad_norm": 0.5934797525405884,
+      "learning_rate": 0.00011845825960368326,
+      "loss": 0.3534,
+      "step": 6225
+    },
+    {
+      "epoch": 0.4578277814545187,
+      "grad_norm": 0.665091335773468,
+      "learning_rate": 0.00011843485213111518,
+      "loss": 0.3712,
+      "step": 6226
+    },
+    {
+      "epoch": 0.45790131627325537,
+      "grad_norm": 0.995705783367157,
+      "learning_rate": 0.00011841144361290284,
+      "loss": 0.3444,
+      "step": 6227
+    },
+    {
+      "epoch": 0.45797485109199204,
+      "grad_norm": 1.0905507802963257,
+      "learning_rate": 0.00011838803405037398,
+      "loss": 0.3367,
+      "step": 6228
+    },
+    {
+      "epoch": 0.4580483859107287,
+      "grad_norm": 0.8010570406913757,
+      "learning_rate": 0.00011836462344485642,
+      "loss": 0.3486,
+      "step": 6229
+    },
+    {
+      "epoch": 0.4581219207294654,
+      "grad_norm": 0.9078811407089233,
+      "learning_rate": 0.00011834121179767803,
+      "loss": 0.3667,
+      "step": 6230
+    },
+    {
+      "epoch": 0.45819545554820207,
+      "grad_norm": 1.1089524030685425,
+      "learning_rate": 0.00011831779911016676,
+      "loss": 0.3658,
+      "step": 6231
+    },
+    {
+      "epoch": 0.45826899036693874,
+      "grad_norm": 0.9316732287406921,
+      "learning_rate": 0.00011829438538365061,
+      "loss": 0.3617,
+      "step": 6232
+    },
+    {
+      "epoch": 0.4583425251856754,
+      "grad_norm": 0.6151084899902344,
+      "learning_rate": 0.0001182709706194576,
+      "loss": 0.3473,
+      "step": 6233
+    },
+    {
+      "epoch": 0.4584160600044121,
+      "grad_norm": 0.8707422018051147,
+      "learning_rate": 0.00011824755481891583,
+      "loss": 0.3555,
+      "step": 6234
+    },
+    {
+      "epoch": 0.45848959482314877,
+      "grad_norm": 0.7750018835067749,
+      "learning_rate": 0.00011822413798335353,
+      "loss": 0.3504,
+      "step": 6235
+    },
+    {
+      "epoch": 0.45856312964188545,
+      "grad_norm": 0.7767665982246399,
+      "learning_rate": 0.00011820072011409888,
+      "loss": 0.3742,
+      "step": 6236
+    },
+    {
+      "epoch": 0.4586366644606221,
+      "grad_norm": 0.7899577021598816,
+      "learning_rate": 0.00011817730121248019,
+      "loss": 0.3492,
+      "step": 6237
+    },
+    {
+      "epoch": 0.4587101992793588,
+      "grad_norm": 0.48993930220603943,
+      "learning_rate": 0.00011815388127982576,
+      "loss": 0.3509,
+      "step": 6238
+    },
+    {
+      "epoch": 0.4587837340980955,
+      "grad_norm": 0.7312673330307007,
+      "learning_rate": 0.00011813046031746402,
+      "loss": 0.3372,
+      "step": 6239
+    },
+    {
+      "epoch": 0.4588572689168321,
+      "grad_norm": 1.0920871496200562,
+      "learning_rate": 0.00011810703832672347,
+      "loss": 0.3388,
+      "step": 6240
+    },
+    {
+      "epoch": 0.45893080373556877,
+      "grad_norm": 0.950035810470581,
+      "learning_rate": 0.00011808361530893259,
+      "loss": 0.3603,
+      "step": 6241
+    },
+    {
+      "epoch": 0.45900433855430545,
+      "grad_norm": 0.8629716038703918,
+      "learning_rate": 0.00011806019126541994,
+      "loss": 0.3475,
+      "step": 6242
+    },
+    {
+      "epoch": 0.4590778733730421,
+      "grad_norm": 0.947162926197052,
+      "learning_rate": 0.00011803676619751423,
+      "loss": 0.3671,
+      "step": 6243
+    },
+    {
+      "epoch": 0.4591514081917788,
+      "grad_norm": 1.5727835893630981,
+      "learning_rate": 0.00011801334010654405,
+      "loss": 0.3503,
+      "step": 6244
+    },
+    {
+      "epoch": 0.4592249430105155,
+      "grad_norm": 0.6672695875167847,
+      "learning_rate": 0.00011798991299383826,
+      "loss": 0.3386,
+      "step": 6245
+    },
+    {
+      "epoch": 0.45929847782925215,
+      "grad_norm": 1.529799222946167,
+      "learning_rate": 0.00011796648486072559,
+      "loss": 0.3603,
+      "step": 6246
+    },
+    {
+      "epoch": 0.4593720126479888,
+      "grad_norm": 0.8284086585044861,
+      "learning_rate": 0.00011794305570853492,
+      "loss": 0.3696,
+      "step": 6247
+    },
+    {
+      "epoch": 0.4594455474667255,
+      "grad_norm": 1.7140145301818848,
+      "learning_rate": 0.00011791962553859525,
+      "loss": 0.3698,
+      "step": 6248
+    },
+    {
+      "epoch": 0.4595190822854622,
+      "grad_norm": 1.1718047857284546,
+      "learning_rate": 0.00011789619435223546,
+      "loss": 0.3756,
+      "step": 6249
+    },
+    {
+      "epoch": 0.45959261710419885,
+      "grad_norm": 1.2844130992889404,
+      "learning_rate": 0.00011787276215078466,
+      "loss": 0.3785,
+      "step": 6250
+    },
+    {
+      "epoch": 0.45966615192293553,
+      "grad_norm": 0.9949660301208496,
+      "learning_rate": 0.00011784932893557194,
+      "loss": 0.3436,
+      "step": 6251
+    },
+    {
+      "epoch": 0.4597396867416722,
+      "grad_norm": 1.406160831451416,
+      "learning_rate": 0.00011782589470792642,
+      "loss": 0.3862,
+      "step": 6252
+    },
+    {
+      "epoch": 0.4598132215604089,
+      "grad_norm": 0.7310903668403625,
+      "learning_rate": 0.00011780245946917734,
+      "loss": 0.3668,
+      "step": 6253
+    },
+    {
+      "epoch": 0.4598867563791455,
+      "grad_norm": 1.6667264699935913,
+      "learning_rate": 0.000117779023220654,
+      "loss": 0.3655,
+      "step": 6254
+    },
+    {
+      "epoch": 0.4599602911978822,
+      "grad_norm": 1.10541832447052,
+      "learning_rate": 0.00011775558596368568,
+      "loss": 0.3835,
+      "step": 6255
+    },
+    {
+      "epoch": 0.46003382601661885,
+      "grad_norm": 1.5324389934539795,
+      "learning_rate": 0.0001177321476996018,
+      "loss": 0.3753,
+      "step": 6256
+    },
+    {
+      "epoch": 0.46010736083535553,
+      "grad_norm": 1.5224992036819458,
+      "learning_rate": 0.00011770870842973177,
+      "loss": 0.4268,
+      "step": 6257
+    },
+    {
+      "epoch": 0.4601808956540922,
+      "grad_norm": 1.1013344526290894,
+      "learning_rate": 0.0001176852681554051,
+      "loss": 0.3932,
+      "step": 6258
+    },
+    {
+      "epoch": 0.4602544304728289,
+      "grad_norm": 1.2478052377700806,
+      "learning_rate": 0.00011766182687795137,
+      "loss": 0.3722,
+      "step": 6259
+    },
+    {
+      "epoch": 0.46032796529156556,
+      "grad_norm": 0.9212057590484619,
+      "learning_rate": 0.0001176383845987002,
+      "loss": 0.3631,
+      "step": 6260
+    },
+    {
+      "epoch": 0.46040150011030223,
+      "grad_norm": 1.7510472536087036,
+      "learning_rate": 0.00011761494131898123,
+      "loss": 0.3646,
+      "step": 6261
+    },
+    {
+      "epoch": 0.4604750349290389,
+      "grad_norm": 0.6957036256790161,
+      "learning_rate": 0.00011759149704012418,
+      "loss": 0.3423,
+      "step": 6262
+    },
+    {
+      "epoch": 0.4605485697477756,
+      "grad_norm": 2.084439754486084,
+      "learning_rate": 0.00011756805176345886,
+      "loss": 0.3625,
+      "step": 6263
+    },
+    {
+      "epoch": 0.46062210456651226,
+      "grad_norm": 1.0408072471618652,
+      "learning_rate": 0.00011754460549031512,
+      "loss": 0.3767,
+      "step": 6264
+    },
+    {
+      "epoch": 0.46069563938524893,
+      "grad_norm": 2.450429916381836,
+      "learning_rate": 0.0001175211582220228,
+      "loss": 0.3755,
+      "step": 6265
+    },
+    {
+      "epoch": 0.4607691742039856,
+      "grad_norm": 2.8796000480651855,
+      "learning_rate": 0.00011749770995991196,
+      "loss": 0.3413,
+      "step": 6266
+    },
+    {
+      "epoch": 0.4608427090227223,
+      "grad_norm": 1.0839293003082275,
+      "learning_rate": 0.00011747426070531249,
+      "loss": 0.3467,
+      "step": 6267
+    },
+    {
+      "epoch": 0.4609162438414589,
+      "grad_norm": 1.8136755228042603,
+      "learning_rate": 0.00011745081045955452,
+      "loss": 0.3914,
+      "step": 6268
+    },
+    {
+      "epoch": 0.4609897786601956,
+      "grad_norm": 1.7463805675506592,
+      "learning_rate": 0.0001174273592239682,
+      "loss": 0.3883,
+      "step": 6269
+    },
+    {
+      "epoch": 0.46106331347893226,
+      "grad_norm": 1.398087501525879,
+      "learning_rate": 0.00011740390699988362,
+      "loss": 0.3714,
+      "step": 6270
+    },
+    {
+      "epoch": 0.46113684829766893,
+      "grad_norm": 1.6386921405792236,
+      "learning_rate": 0.00011738045378863112,
+      "loss": 0.374,
+      "step": 6271
+    },
+    {
+      "epoch": 0.4612103831164056,
+      "grad_norm": 1.7128411531448364,
+      "learning_rate": 0.0001173569995915409,
+      "loss": 0.3781,
+      "step": 6272
+    },
+    {
+      "epoch": 0.4612839179351423,
+      "grad_norm": 1.140944004058838,
+      "learning_rate": 0.00011733354440994338,
+      "loss": 0.3675,
+      "step": 6273
+    },
+    {
+      "epoch": 0.46135745275387896,
+      "grad_norm": 1.0774496793746948,
+      "learning_rate": 0.00011731008824516892,
+      "loss": 0.3444,
+      "step": 6274
+    },
+    {
+      "epoch": 0.46143098757261564,
+      "grad_norm": 0.9977837800979614,
+      "learning_rate": 0.00011728663109854793,
+      "loss": 0.3555,
+      "step": 6275
+    },
+    {
+      "epoch": 0.4615045223913523,
+      "grad_norm": 1.0025955438613892,
+      "learning_rate": 0.00011726317297141104,
+      "loss": 0.3752,
+      "step": 6276
+    },
+    {
+      "epoch": 0.461578057210089,
+      "grad_norm": 1.980357050895691,
+      "learning_rate": 0.00011723971386508877,
+      "loss": 0.4012,
+      "step": 6277
+    },
+    {
+      "epoch": 0.46165159202882566,
+      "grad_norm": 1.2442679405212402,
+      "learning_rate": 0.0001172162537809117,
+      "loss": 0.3591,
+      "step": 6278
+    },
+    {
+      "epoch": 0.46172512684756234,
+      "grad_norm": 2.157864809036255,
+      "learning_rate": 0.00011719279272021056,
+      "loss": 0.3568,
+      "step": 6279
+    },
+    {
+      "epoch": 0.461798661666299,
+      "grad_norm": 2.0982611179351807,
+      "learning_rate": 0.00011716933068431606,
+      "loss": 0.3577,
+      "step": 6280
+    },
+    {
+      "epoch": 0.4618721964850357,
+      "grad_norm": 1.637620210647583,
+      "learning_rate": 0.00011714586767455902,
+      "loss": 0.3615,
+      "step": 6281
+    },
+    {
+      "epoch": 0.4619457313037723,
+      "grad_norm": 1.4100263118743896,
+      "learning_rate": 0.00011712240369227028,
+      "loss": 0.3822,
+      "step": 6282
+    },
+    {
+      "epoch": 0.462019266122509,
+      "grad_norm": 1.9073642492294312,
+      "learning_rate": 0.00011709893873878069,
+      "loss": 0.3681,
+      "step": 6283
+    },
+    {
+      "epoch": 0.46209280094124566,
+      "grad_norm": 1.1209079027175903,
+      "learning_rate": 0.00011707547281542131,
+      "loss": 0.3822,
+      "step": 6284
+    },
+    {
+      "epoch": 0.46216633575998234,
+      "grad_norm": 2.972618341445923,
+      "learning_rate": 0.00011705200592352303,
+      "loss": 0.38,
+      "step": 6285
+    },
+    {
+      "epoch": 0.462239870578719,
+      "grad_norm": 2.799818277359009,
+      "learning_rate": 0.00011702853806441703,
+      "loss": 0.3459,
+      "step": 6286
+    },
+    {
+      "epoch": 0.4623134053974557,
+      "grad_norm": 0.9335697293281555,
+      "learning_rate": 0.00011700506923943434,
+      "loss": 0.3683,
+      "step": 6287
+    },
+    {
+      "epoch": 0.46238694021619237,
+      "grad_norm": 1.9884064197540283,
+      "learning_rate": 0.00011698159944990619,
+      "loss": 0.4271,
+      "step": 6288
+    },
+    {
+      "epoch": 0.46246047503492904,
+      "grad_norm": 0.9640592932701111,
+      "learning_rate": 0.00011695812869716382,
+      "loss": 0.4314,
+      "step": 6289
+    },
+    {
+      "epoch": 0.4625340098536657,
+      "grad_norm": 1.8692984580993652,
+      "learning_rate": 0.00011693465698253847,
+      "loss": 0.4155,
+      "step": 6290
+    },
+    {
+      "epoch": 0.4626075446724024,
+      "grad_norm": 1.8110517263412476,
+      "learning_rate": 0.00011691118430736152,
+      "loss": 0.4171,
+      "step": 6291
+    },
+    {
+      "epoch": 0.46268107949113907,
+      "grad_norm": 1.1012272834777832,
+      "learning_rate": 0.00011688771067296433,
+      "loss": 0.3749,
+      "step": 6292
+    },
+    {
+      "epoch": 0.46275461430987574,
+      "grad_norm": 1.28475022315979,
+      "learning_rate": 0.00011686423608067837,
+      "loss": 0.3749,
+      "step": 6293
+    },
+    {
+      "epoch": 0.4628281491286124,
+      "grad_norm": 1.7396913766860962,
+      "learning_rate": 0.00011684076053183517,
+      "loss": 0.3765,
+      "step": 6294
+    },
+    {
+      "epoch": 0.4629016839473491,
+      "grad_norm": 1.7956327199935913,
+      "learning_rate": 0.00011681728402776623,
+      "loss": 0.4131,
+      "step": 6295
+    },
+    {
+      "epoch": 0.4629752187660857,
+      "grad_norm": 1.6717603206634521,
+      "learning_rate": 0.00011679380656980321,
+      "loss": 0.4041,
+      "step": 6296
+    },
+    {
+      "epoch": 0.4630487535848224,
+      "grad_norm": 2.3627827167510986,
+      "learning_rate": 0.00011677032815927775,
+      "loss": 0.402,
+      "step": 6297
+    },
+    {
+      "epoch": 0.46312228840355907,
+      "grad_norm": 1.8629138469696045,
+      "learning_rate": 0.00011674684879752156,
+      "loss": 0.358,
+      "step": 6298
+    },
+    {
+      "epoch": 0.46319582322229574,
+      "grad_norm": 1.48030686378479,
+      "learning_rate": 0.00011672336848586644,
+      "loss": 0.4085,
+      "step": 6299
+    },
+    {
+      "epoch": 0.4632693580410324,
+      "grad_norm": 1.5685673952102661,
+      "learning_rate": 0.00011669988722564425,
+      "loss": 0.3748,
+      "step": 6300
+    },
+    {
+      "epoch": 0.4633428928597691,
+      "grad_norm": 1.3945047855377197,
+      "learning_rate": 0.0001166764050181868,
+      "loss": 0.3812,
+      "step": 6301
+    },
+    {
+      "epoch": 0.46341642767850577,
+      "grad_norm": 2.4545226097106934,
+      "learning_rate": 0.00011665292186482606,
+      "loss": 0.3687,
+      "step": 6302
+    },
+    {
+      "epoch": 0.46348996249724245,
+      "grad_norm": 1.7099908590316772,
+      "learning_rate": 0.00011662943776689401,
+      "loss": 0.3935,
+      "step": 6303
+    },
+    {
+      "epoch": 0.4635634973159791,
+      "grad_norm": 2.264186143875122,
+      "learning_rate": 0.00011660595272572272,
+      "loss": 0.3618,
+      "step": 6304
+    },
+    {
+      "epoch": 0.4636370321347158,
+      "grad_norm": 2.8830997943878174,
+      "learning_rate": 0.00011658246674264425,
+      "loss": 0.3574,
+      "step": 6305
+    },
+    {
+      "epoch": 0.4637105669534525,
+      "grad_norm": 0.8811777830123901,
+      "learning_rate": 0.00011655897981899076,
+      "loss": 0.3753,
+      "step": 6306
+    },
+    {
+      "epoch": 0.46378410177218915,
+      "grad_norm": 2.5403518676757812,
+      "learning_rate": 0.0001165354919560945,
+      "loss": 0.3839,
+      "step": 6307
+    },
+    {
+      "epoch": 0.4638576365909258,
+      "grad_norm": 2.9298431873321533,
+      "learning_rate": 0.00011651200315528763,
+      "loss": 0.3895,
+      "step": 6308
+    },
+    {
+      "epoch": 0.4639311714096625,
+      "grad_norm": 1.1470447778701782,
+      "learning_rate": 0.00011648851341790257,
+      "loss": 0.3441,
+      "step": 6309
+    },
+    {
+      "epoch": 0.4640047062283991,
+      "grad_norm": 2.424952507019043,
+      "learning_rate": 0.00011646502274527159,
+      "loss": 0.3718,
+      "step": 6310
+    },
+    {
+      "epoch": 0.4640782410471358,
+      "grad_norm": 1.9604947566986084,
+      "learning_rate": 0.00011644153113872714,
+      "loss": 0.3643,
+      "step": 6311
+    },
+    {
+      "epoch": 0.4641517758658725,
+      "grad_norm": 1.2452553510665894,
+      "learning_rate": 0.0001164180385996017,
+      "loss": 0.4108,
+      "step": 6312
+    },
+    {
+      "epoch": 0.46422531068460915,
+      "grad_norm": 1.3672728538513184,
+      "learning_rate": 0.0001163945451292278,
+      "loss": 0.3644,
+      "step": 6313
+    },
+    {
+      "epoch": 0.4642988455033458,
+      "grad_norm": 0.9871485233306885,
+      "learning_rate": 0.00011637105072893799,
+      "loss": 0.3464,
+      "step": 6314
+    },
+    {
+      "epoch": 0.4643723803220825,
+      "grad_norm": 1.135166049003601,
+      "learning_rate": 0.0001163475554000649,
+      "loss": 0.3793,
+      "step": 6315
+    },
+    {
+      "epoch": 0.4644459151408192,
+      "grad_norm": 1.0990195274353027,
+      "learning_rate": 0.00011632405914394122,
+      "loss": 0.3565,
+      "step": 6316
+    },
+    {
+      "epoch": 0.46451944995955585,
+      "grad_norm": 0.7841278910636902,
+      "learning_rate": 0.00011630056196189967,
+      "loss": 0.4019,
+      "step": 6317
+    },
+    {
+      "epoch": 0.46459298477829253,
+      "grad_norm": 1.4879816770553589,
+      "learning_rate": 0.00011627706385527306,
+      "loss": 0.3821,
+      "step": 6318
+    },
+    {
+      "epoch": 0.4646665195970292,
+      "grad_norm": 0.952788770198822,
+      "learning_rate": 0.0001162535648253942,
+      "loss": 0.3629,
+      "step": 6319
+    },
+    {
+      "epoch": 0.4647400544157659,
+      "grad_norm": 1.751371145248413,
+      "learning_rate": 0.000116230064873596,
+      "loss": 0.3738,
+      "step": 6320
+    },
+    {
+      "epoch": 0.46481358923450256,
+      "grad_norm": 1.421355962753296,
+      "learning_rate": 0.00011620656400121137,
+      "loss": 0.3712,
+      "step": 6321
+    },
+    {
+      "epoch": 0.46488712405323923,
+      "grad_norm": 1.2807424068450928,
+      "learning_rate": 0.00011618306220957333,
+      "loss": 0.3757,
+      "step": 6322
+    },
+    {
+      "epoch": 0.4649606588719759,
+      "grad_norm": 1.0017268657684326,
+      "learning_rate": 0.00011615955950001493,
+      "loss": 0.3861,
+      "step": 6323
+    },
+    {
+      "epoch": 0.4650341936907125,
+      "grad_norm": 1.2452362775802612,
+      "learning_rate": 0.0001161360558738693,
+      "loss": 0.3659,
+      "step": 6324
+    },
+    {
+      "epoch": 0.4651077285094492,
+      "grad_norm": 0.9998116493225098,
+      "learning_rate": 0.0001161125513324695,
+      "loss": 0.3684,
+      "step": 6325
+    },
+    {
+      "epoch": 0.4651812633281859,
+      "grad_norm": 0.8719884157180786,
+      "learning_rate": 0.00011608904587714879,
+      "loss": 0.3898,
+      "step": 6326
+    },
+    {
+      "epoch": 0.46525479814692255,
+      "grad_norm": 1.271941065788269,
+      "learning_rate": 0.00011606553950924045,
+      "loss": 0.3652,
+      "step": 6327
+    },
+    {
+      "epoch": 0.46532833296565923,
+      "grad_norm": 0.7825437784194946,
+      "learning_rate": 0.00011604203223007772,
+      "loss": 0.3395,
+      "step": 6328
+    },
+    {
+      "epoch": 0.4654018677843959,
+      "grad_norm": 1.2656804323196411,
+      "learning_rate": 0.00011601852404099398,
+      "loss": 0.3784,
+      "step": 6329
+    },
+    {
+      "epoch": 0.4654754026031326,
+      "grad_norm": 1.1670985221862793,
+      "learning_rate": 0.0001159950149433227,
+      "loss": 0.3487,
+      "step": 6330
+    },
+    {
+      "epoch": 0.46554893742186926,
+      "grad_norm": 1.3864103555679321,
+      "learning_rate": 0.00011597150493839725,
+      "loss": 0.3761,
+      "step": 6331
+    },
+    {
+      "epoch": 0.46562247224060593,
+      "grad_norm": 1.3200836181640625,
+      "learning_rate": 0.00011594799402755121,
+      "loss": 0.3805,
+      "step": 6332
+    },
+    {
+      "epoch": 0.4656960070593426,
+      "grad_norm": 1.2221260070800781,
+      "learning_rate": 0.00011592448221211812,
+      "loss": 0.3952,
+      "step": 6333
+    },
+    {
+      "epoch": 0.4657695418780793,
+      "grad_norm": 1.0786011219024658,
+      "learning_rate": 0.00011590096949343156,
+      "loss": 0.4118,
+      "step": 6334
+    },
+    {
+      "epoch": 0.46584307669681596,
+      "grad_norm": 0.959138810634613,
+      "learning_rate": 0.00011587745587282526,
+      "loss": 0.367,
+      "step": 6335
+    },
+    {
+      "epoch": 0.46591661151555264,
+      "grad_norm": 1.0285403728485107,
+      "learning_rate": 0.0001158539413516329,
+      "loss": 0.3546,
+      "step": 6336
+    },
+    {
+      "epoch": 0.4659901463342893,
+      "grad_norm": 0.7812040448188782,
+      "learning_rate": 0.00011583042593118826,
+      "loss": 0.3541,
+      "step": 6337
+    },
+    {
+      "epoch": 0.46606368115302593,
+      "grad_norm": 1.026471495628357,
+      "learning_rate": 0.00011580690961282516,
+      "loss": 0.3824,
+      "step": 6338
+    },
+    {
+      "epoch": 0.4661372159717626,
+      "grad_norm": 0.868392288684845,
+      "learning_rate": 0.00011578339239787743,
+      "loss": 0.3822,
+      "step": 6339
+    },
+    {
+      "epoch": 0.4662107507904993,
+      "grad_norm": 1.4735033512115479,
+      "learning_rate": 0.00011575987428767905,
+      "loss": 0.3744,
+      "step": 6340
+    },
+    {
+      "epoch": 0.46628428560923596,
+      "grad_norm": 1.1510437726974487,
+      "learning_rate": 0.000115736355283564,
+      "loss": 0.3697,
+      "step": 6341
+    },
+    {
+      "epoch": 0.46635782042797264,
+      "grad_norm": 0.8683717846870422,
+      "learning_rate": 0.00011571283538686622,
+      "loss": 0.3655,
+      "step": 6342
+    },
+    {
+      "epoch": 0.4664313552467093,
+      "grad_norm": 1.3830277919769287,
+      "learning_rate": 0.00011568931459891987,
+      "loss": 0.3453,
+      "step": 6343
+    },
+    {
+      "epoch": 0.466504890065446,
+      "grad_norm": 0.6826279759407043,
+      "learning_rate": 0.000115665792921059,
+      "loss": 0.3755,
+      "step": 6344
+    },
+    {
+      "epoch": 0.46657842488418266,
+      "grad_norm": 1.1544381380081177,
+      "learning_rate": 0.00011564227035461786,
+      "loss": 0.4123,
+      "step": 6345
+    },
+    {
+      "epoch": 0.46665195970291934,
+      "grad_norm": 0.8923050761222839,
+      "learning_rate": 0.00011561874690093062,
+      "loss": 0.3535,
+      "step": 6346
+    },
+    {
+      "epoch": 0.466725494521656,
+      "grad_norm": 0.9445260167121887,
+      "learning_rate": 0.00011559522256133155,
+      "loss": 0.3921,
+      "step": 6347
+    },
+    {
+      "epoch": 0.4667990293403927,
+      "grad_norm": 1.1497992277145386,
+      "learning_rate": 0.00011557169733715504,
+      "loss": 0.3804,
+      "step": 6348
+    },
+    {
+      "epoch": 0.46687256415912937,
+      "grad_norm": 0.6425493955612183,
+      "learning_rate": 0.0001155481712297354,
+      "loss": 0.3512,
+      "step": 6349
+    },
+    {
+      "epoch": 0.46694609897786604,
+      "grad_norm": 1.0071300268173218,
+      "learning_rate": 0.0001155246442404071,
+      "loss": 0.3682,
+      "step": 6350
+    },
+    {
+      "epoch": 0.4670196337966027,
+      "grad_norm": 1.0724595785140991,
+      "learning_rate": 0.00011550111637050459,
+      "loss": 0.3728,
+      "step": 6351
+    },
+    {
+      "epoch": 0.46709316861533934,
+      "grad_norm": 0.6263882517814636,
+      "learning_rate": 0.00011547758762136238,
+      "loss": 0.3563,
+      "step": 6352
+    },
+    {
+      "epoch": 0.467166703434076,
+      "grad_norm": 1.2784879207611084,
+      "learning_rate": 0.00011545405799431508,
+      "loss": 0.3799,
+      "step": 6353
+    },
+    {
+      "epoch": 0.4672402382528127,
+      "grad_norm": 0.8252459764480591,
+      "learning_rate": 0.00011543052749069729,
+      "loss": 0.3454,
+      "step": 6354
+    },
+    {
+      "epoch": 0.46731377307154937,
+      "grad_norm": 0.6907427310943604,
+      "learning_rate": 0.00011540699611184371,
+      "loss": 0.3647,
+      "step": 6355
+    },
+    {
+      "epoch": 0.46738730789028604,
+      "grad_norm": 1.045098066329956,
+      "learning_rate": 0.00011538346385908907,
+      "loss": 0.3637,
+      "step": 6356
+    },
+    {
+      "epoch": 0.4674608427090227,
+      "grad_norm": 0.5766198039054871,
+      "learning_rate": 0.00011535993073376807,
+      "loss": 0.3359,
+      "step": 6357
+    },
+    {
+      "epoch": 0.4675343775277594,
+      "grad_norm": 0.714160680770874,
+      "learning_rate": 0.00011533639673721564,
+      "loss": 0.3421,
+      "step": 6358
+    },
+    {
+      "epoch": 0.46760791234649607,
+      "grad_norm": 0.7781263589859009,
+      "learning_rate": 0.00011531286187076658,
+      "loss": 0.373,
+      "step": 6359
+    },
+    {
+      "epoch": 0.46768144716523274,
+      "grad_norm": 0.6667099595069885,
+      "learning_rate": 0.00011528932613575585,
+      "loss": 0.3504,
+      "step": 6360
+    },
+    {
+      "epoch": 0.4677549819839694,
+      "grad_norm": 0.6041437983512878,
+      "learning_rate": 0.00011526578953351839,
+      "loss": 0.3528,
+      "step": 6361
+    },
+    {
+      "epoch": 0.4678285168027061,
+      "grad_norm": 0.9238623380661011,
+      "learning_rate": 0.00011524225206538921,
+      "loss": 0.3563,
+      "step": 6362
+    },
+    {
+      "epoch": 0.46790205162144277,
+      "grad_norm": 1.1122595071792603,
+      "learning_rate": 0.00011521871373270345,
+      "loss": 0.3714,
+      "step": 6363
+    },
+    {
+      "epoch": 0.46797558644017945,
+      "grad_norm": 0.810469925403595,
+      "learning_rate": 0.00011519517453679617,
+      "loss": 0.3448,
+      "step": 6364
+    },
+    {
+      "epoch": 0.4680491212589161,
+      "grad_norm": 0.5414056777954102,
+      "learning_rate": 0.00011517163447900257,
+      "loss": 0.3577,
+      "step": 6365
+    },
+    {
+      "epoch": 0.46812265607765274,
+      "grad_norm": 1.0580976009368896,
+      "learning_rate": 0.00011514809356065783,
+      "loss": 0.3637,
+      "step": 6366
+    },
+    {
+      "epoch": 0.4681961908963894,
+      "grad_norm": 0.9105299711227417,
+      "learning_rate": 0.00011512455178309725,
+      "loss": 0.3387,
+      "step": 6367
+    },
+    {
+      "epoch": 0.4682697257151261,
+      "grad_norm": 0.7266759276390076,
+      "learning_rate": 0.00011510100914765611,
+      "loss": 0.3433,
+      "step": 6368
+    },
+    {
+      "epoch": 0.46834326053386277,
+      "grad_norm": 0.47904327511787415,
+      "learning_rate": 0.00011507746565566981,
+      "loss": 0.3594,
+      "step": 6369
+    },
+    {
+      "epoch": 0.46841679535259945,
+      "grad_norm": 0.6811612844467163,
+      "learning_rate": 0.00011505392130847375,
+      "loss": 0.3593,
+      "step": 6370
+    },
+    {
+      "epoch": 0.4684903301713361,
+      "grad_norm": 0.7075286507606506,
+      "learning_rate": 0.00011503037610740342,
+      "loss": 0.369,
+      "step": 6371
+    },
+    {
+      "epoch": 0.4685638649900728,
+      "grad_norm": 0.4825374484062195,
+      "learning_rate": 0.00011500683005379423,
+      "loss": 0.3402,
+      "step": 6372
+    },
+    {
+      "epoch": 0.4686373998088095,
+      "grad_norm": 0.521576464176178,
+      "learning_rate": 0.00011498328314898183,
+      "loss": 0.3706,
+      "step": 6373
+    },
+    {
+      "epoch": 0.46871093462754615,
+      "grad_norm": 0.7213711142539978,
+      "learning_rate": 0.0001149597353943018,
+      "loss": 0.3672,
+      "step": 6374
+    },
+    {
+      "epoch": 0.4687844694462828,
+      "grad_norm": 0.4882444143295288,
+      "learning_rate": 0.00011493618679108977,
+      "loss": 0.3702,
+      "step": 6375
+    },
+    {
+      "epoch": 0.4688580042650195,
+      "grad_norm": 0.6332077383995056,
+      "learning_rate": 0.00011491263734068149,
+      "loss": 0.3789,
+      "step": 6376
+    },
+    {
+      "epoch": 0.4689315390837562,
+      "grad_norm": 0.7036481499671936,
+      "learning_rate": 0.00011488908704441268,
+      "loss": 0.3619,
+      "step": 6377
+    },
+    {
+      "epoch": 0.46900507390249285,
+      "grad_norm": 0.6078009605407715,
+      "learning_rate": 0.00011486553590361912,
+      "loss": 0.3381,
+      "step": 6378
+    },
+    {
+      "epoch": 0.46907860872122953,
+      "grad_norm": 0.5756479501724243,
+      "learning_rate": 0.00011484198391963669,
+      "loss": 0.3495,
+      "step": 6379
+    },
+    {
+      "epoch": 0.46915214353996615,
+      "grad_norm": 0.8945436477661133,
+      "learning_rate": 0.00011481843109380127,
+      "loss": 0.3477,
+      "step": 6380
+    },
+    {
+      "epoch": 0.4692256783587028,
+      "grad_norm": 1.025556206703186,
+      "learning_rate": 0.00011479487742744878,
+      "loss": 0.3544,
+      "step": 6381
+    },
+    {
+      "epoch": 0.4692992131774395,
+      "grad_norm": 0.8235267400741577,
+      "learning_rate": 0.00011477132292191526,
+      "loss": 0.3673,
+      "step": 6382
+    },
+    {
+      "epoch": 0.4693727479961762,
+      "grad_norm": 0.5523346066474915,
+      "learning_rate": 0.00011474776757853672,
+      "loss": 0.3472,
+      "step": 6383
+    },
+    {
+      "epoch": 0.46944628281491285,
+      "grad_norm": 0.6100131273269653,
+      "learning_rate": 0.00011472421139864922,
+      "loss": 0.367,
+      "step": 6384
+    },
+    {
+      "epoch": 0.4695198176336495,
+      "grad_norm": 0.5598767995834351,
+      "learning_rate": 0.00011470065438358888,
+      "loss": 0.3647,
+      "step": 6385
+    },
+    {
+      "epoch": 0.4695933524523862,
+      "grad_norm": 0.5487914681434631,
+      "learning_rate": 0.00011467709653469194,
+      "loss": 0.3526,
+      "step": 6386
+    },
+    {
+      "epoch": 0.4696668872711229,
+      "grad_norm": 0.5581538677215576,
+      "learning_rate": 0.0001146535378532946,
+      "loss": 0.3374,
+      "step": 6387
+    },
+    {
+      "epoch": 0.46974042208985956,
+      "grad_norm": 0.6417310833930969,
+      "learning_rate": 0.0001146299783407331,
+      "loss": 0.3455,
+      "step": 6388
+    },
+    {
+      "epoch": 0.46981395690859623,
+      "grad_norm": 0.8341915607452393,
+      "learning_rate": 0.00011460641799834384,
+      "loss": 0.3466,
+      "step": 6389
+    },
+    {
+      "epoch": 0.4698874917273329,
+      "grad_norm": 0.9700922966003418,
+      "learning_rate": 0.0001145828568274631,
+      "loss": 0.3543,
+      "step": 6390
+    },
+    {
+      "epoch": 0.4699610265460696,
+      "grad_norm": 0.9483770132064819,
+      "learning_rate": 0.00011455929482942735,
+      "loss": 0.345,
+      "step": 6391
+    },
+    {
+      "epoch": 0.47003456136480626,
+      "grad_norm": 0.4473118782043457,
+      "learning_rate": 0.00011453573200557302,
+      "loss": 0.3306,
+      "step": 6392
+    },
+    {
+      "epoch": 0.47010809618354293,
+      "grad_norm": 1.0381044149398804,
+      "learning_rate": 0.00011451216835723662,
+      "loss": 0.3867,
+      "step": 6393
+    },
+    {
+      "epoch": 0.47018163100227955,
+      "grad_norm": 0.7680667042732239,
+      "learning_rate": 0.00011448860388575473,
+      "loss": 0.3374,
+      "step": 6394
+    },
+    {
+      "epoch": 0.47025516582101623,
+      "grad_norm": 0.6945741176605225,
+      "learning_rate": 0.00011446503859246394,
+      "loss": 0.3415,
+      "step": 6395
+    },
+    {
+      "epoch": 0.4703287006397529,
+      "grad_norm": 1.294927716255188,
+      "learning_rate": 0.0001144414724787009,
+      "loss": 0.3279,
+      "step": 6396
+    },
+    {
+      "epoch": 0.4704022354584896,
+      "grad_norm": 0.8160262703895569,
+      "learning_rate": 0.00011441790554580227,
+      "loss": 0.3496,
+      "step": 6397
+    },
+    {
+      "epoch": 0.47047577027722626,
+      "grad_norm": 0.8273621201515198,
+      "learning_rate": 0.00011439433779510484,
+      "loss": 0.3442,
+      "step": 6398
+    },
+    {
+      "epoch": 0.47054930509596293,
+      "grad_norm": 1.5975579023361206,
+      "learning_rate": 0.00011437076922794538,
+      "loss": 0.3336,
+      "step": 6399
+    },
+    {
+      "epoch": 0.4706228399146996,
+      "grad_norm": 0.676043689250946,
+      "learning_rate": 0.00011434719984566074,
+      "loss": 0.341,
+      "step": 6400
+    },
+    {
+      "epoch": 0.4706963747334363,
+      "grad_norm": 1.7063782215118408,
+      "learning_rate": 0.00011432362964958775,
+      "loss": 0.3582,
+      "step": 6401
+    },
+    {
+      "epoch": 0.47076990955217296,
+      "grad_norm": 0.7023589611053467,
+      "learning_rate": 0.00011430005864106341,
+      "loss": 0.3382,
+      "step": 6402
+    },
+    {
+      "epoch": 0.47084344437090964,
+      "grad_norm": 1.695648193359375,
+      "learning_rate": 0.0001142764868214246,
+      "loss": 0.3485,
+      "step": 6403
+    },
+    {
+      "epoch": 0.4709169791896463,
+      "grad_norm": 1.0410014390945435,
+      "learning_rate": 0.00011425291419200843,
+      "loss": 0.3459,
+      "step": 6404
+    },
+    {
+      "epoch": 0.470990514008383,
+      "grad_norm": 2.8860185146331787,
+      "learning_rate": 0.00011422934075415193,
+      "loss": 0.3498,
+      "step": 6405
+    },
+    {
+      "epoch": 0.47106404882711966,
+      "grad_norm": 2.6915066242218018,
+      "learning_rate": 0.00011420576650919219,
+      "loss": 0.3316,
+      "step": 6406
+    },
+    {
+      "epoch": 0.47113758364585634,
+      "grad_norm": 0.7950438857078552,
+      "learning_rate": 0.00011418219145846638,
+      "loss": 0.3412,
+      "step": 6407
+    },
+    {
+      "epoch": 0.47121111846459296,
+      "grad_norm": 2.3941738605499268,
+      "learning_rate": 0.0001141586156033117,
+      "loss": 0.3629,
+      "step": 6408
+    },
+    {
+      "epoch": 0.47128465328332964,
+      "grad_norm": 1.8061208724975586,
+      "learning_rate": 0.00011413503894506541,
+      "loss": 0.3462,
+      "step": 6409
+    },
+    {
+      "epoch": 0.4713581881020663,
+      "grad_norm": 1.7510210275650024,
+      "learning_rate": 0.0001141114614850648,
+      "loss": 0.3598,
+      "step": 6410
+    },
+    {
+      "epoch": 0.471431722920803,
+      "grad_norm": 1.811170220375061,
+      "learning_rate": 0.00011408788322464717,
+      "loss": 0.3465,
+      "step": 6411
+    },
+    {
+      "epoch": 0.47150525773953966,
+      "grad_norm": 1.593695878982544,
+      "learning_rate": 0.00011406430416514999,
+      "loss": 0.367,
+      "step": 6412
+    },
+    {
+      "epoch": 0.47157879255827634,
+      "grad_norm": 1.4459583759307861,
+      "learning_rate": 0.00011404072430791059,
+      "loss": 0.372,
+      "step": 6413
+    },
+    {
+      "epoch": 0.471652327377013,
+      "grad_norm": 2.0202648639678955,
+      "learning_rate": 0.0001140171436542665,
+      "loss": 0.3533,
+      "step": 6414
+    },
+    {
+      "epoch": 0.4717258621957497,
+      "grad_norm": 1.7203288078308105,
+      "learning_rate": 0.00011399356220555524,
+      "loss": 0.3389,
+      "step": 6415
+    },
+    {
+      "epoch": 0.47179939701448637,
+      "grad_norm": 1.8229951858520508,
+      "learning_rate": 0.00011396997996311434,
+      "loss": 0.3424,
+      "step": 6416
+    },
+    {
+      "epoch": 0.47187293183322304,
+      "grad_norm": 1.6988964080810547,
+      "learning_rate": 0.00011394639692828146,
+      "loss": 0.3668,
+      "step": 6417
+    },
+    {
+      "epoch": 0.4719464666519597,
+      "grad_norm": 0.8956049680709839,
+      "learning_rate": 0.00011392281310239421,
+      "loss": 0.3686,
+      "step": 6418
+    },
+    {
+      "epoch": 0.4720200014706964,
+      "grad_norm": 1.0016320943832397,
+      "learning_rate": 0.00011389922848679033,
+      "loss": 0.3474,
+      "step": 6419
+    },
+    {
+      "epoch": 0.47209353628943307,
+      "grad_norm": 1.2358428239822388,
+      "learning_rate": 0.00011387564308280754,
+      "loss": 0.374,
+      "step": 6420
+    },
+    {
+      "epoch": 0.47216707110816974,
+      "grad_norm": 0.8556691408157349,
+      "learning_rate": 0.0001138520568917836,
+      "loss": 0.3344,
+      "step": 6421
+    },
+    {
+      "epoch": 0.47224060592690636,
+      "grad_norm": 1.8571592569351196,
+      "learning_rate": 0.00011382846991505638,
+      "loss": 0.341,
+      "step": 6422
+    },
+    {
+      "epoch": 0.47231414074564304,
+      "grad_norm": 1.3451274633407593,
+      "learning_rate": 0.00011380488215396379,
+      "loss": 0.3466,
+      "step": 6423
+    },
+    {
+      "epoch": 0.4723876755643797,
+      "grad_norm": 1.7864357233047485,
+      "learning_rate": 0.00011378129360984369,
+      "loss": 0.3657,
+      "step": 6424
+    },
+    {
+      "epoch": 0.4724612103831164,
+      "grad_norm": 1.8947925567626953,
+      "learning_rate": 0.00011375770428403408,
+      "loss": 0.3641,
+      "step": 6425
+    },
+    {
+      "epoch": 0.47253474520185307,
+      "grad_norm": 1.0818872451782227,
+      "learning_rate": 0.00011373411417787295,
+      "loss": 0.368,
+      "step": 6426
+    },
+    {
+      "epoch": 0.47260828002058974,
+      "grad_norm": 0.7641738653182983,
+      "learning_rate": 0.00011371052329269835,
+      "loss": 0.3569,
+      "step": 6427
+    },
+    {
+      "epoch": 0.4726818148393264,
+      "grad_norm": 1.014693021774292,
+      "learning_rate": 0.00011368693162984844,
+      "loss": 0.3303,
+      "step": 6428
+    },
+    {
+      "epoch": 0.4727553496580631,
+      "grad_norm": 0.5018914937973022,
+      "learning_rate": 0.0001136633391906613,
+      "loss": 0.3611,
+      "step": 6429
+    },
+    {
+      "epoch": 0.47282888447679977,
+      "grad_norm": 0.613017201423645,
+      "learning_rate": 0.00011363974597647517,
+      "loss": 0.3279,
+      "step": 6430
+    },
+    {
+      "epoch": 0.47290241929553645,
+      "grad_norm": 0.5009536743164062,
+      "learning_rate": 0.00011361615198862823,
+      "loss": 0.3577,
+      "step": 6431
+    },
+    {
+      "epoch": 0.4729759541142731,
+      "grad_norm": 0.5991023182868958,
+      "learning_rate": 0.00011359255722845879,
+      "loss": 0.3476,
+      "step": 6432
+    },
+    {
+      "epoch": 0.4730494889330098,
+      "grad_norm": 0.626431405544281,
+      "learning_rate": 0.00011356896169730516,
+      "loss": 0.3577,
+      "step": 6433
+    },
+    {
+      "epoch": 0.4731230237517465,
+      "grad_norm": 0.689420223236084,
+      "learning_rate": 0.00011354536539650567,
+      "loss": 0.3314,
+      "step": 6434
+    },
+    {
+      "epoch": 0.47319655857048315,
+      "grad_norm": 0.5817624926567078,
+      "learning_rate": 0.0001135217683273988,
+      "loss": 0.3521,
+      "step": 6435
+    },
+    {
+      "epoch": 0.47327009338921977,
+      "grad_norm": 0.5320832133293152,
+      "learning_rate": 0.00011349817049132294,
+      "loss": 0.3346,
+      "step": 6436
+    },
+    {
+      "epoch": 0.47334362820795645,
+      "grad_norm": 0.44672757387161255,
+      "learning_rate": 0.00011347457188961661,
+      "loss": 0.3347,
+      "step": 6437
+    },
+    {
+      "epoch": 0.4734171630266931,
+      "grad_norm": 0.6608941555023193,
+      "learning_rate": 0.00011345097252361837,
+      "loss": 0.345,
+      "step": 6438
+    },
+    {
+      "epoch": 0.4734906978454298,
+      "grad_norm": 0.4946008026599884,
+      "learning_rate": 0.00011342737239466672,
+      "loss": 0.331,
+      "step": 6439
+    },
+    {
+      "epoch": 0.4735642326641665,
+      "grad_norm": 0.8214064240455627,
+      "learning_rate": 0.00011340377150410035,
+      "loss": 0.3514,
+      "step": 6440
+    },
+    {
+      "epoch": 0.47363776748290315,
+      "grad_norm": 0.667453408241272,
+      "learning_rate": 0.00011338016985325794,
+      "loss": 0.3369,
+      "step": 6441
+    },
+    {
+      "epoch": 0.4737113023016398,
+      "grad_norm": 0.5965724587440491,
+      "learning_rate": 0.00011335656744347818,
+      "loss": 0.3506,
+      "step": 6442
+    },
+    {
+      "epoch": 0.4737848371203765,
+      "grad_norm": 0.7255462408065796,
+      "learning_rate": 0.00011333296427609981,
+      "loss": 0.36,
+      "step": 6443
+    },
+    {
+      "epoch": 0.4738583719391132,
+      "grad_norm": 0.6307930946350098,
+      "learning_rate": 0.0001133093603524616,
+      "loss": 0.3675,
+      "step": 6444
+    },
+    {
+      "epoch": 0.47393190675784985,
+      "grad_norm": 0.6780118346214294,
+      "learning_rate": 0.00011328575567390245,
+      "loss": 0.3489,
+      "step": 6445
+    },
+    {
+      "epoch": 0.47400544157658653,
+      "grad_norm": 0.5586619973182678,
+      "learning_rate": 0.00011326215024176124,
+      "loss": 0.3809,
+      "step": 6446
+    },
+    {
+      "epoch": 0.4740789763953232,
+      "grad_norm": 0.626480758190155,
+      "learning_rate": 0.00011323854405737684,
+      "loss": 0.3851,
+      "step": 6447
+    },
+    {
+      "epoch": 0.4741525112140599,
+      "grad_norm": 0.9779590368270874,
+      "learning_rate": 0.00011321493712208828,
+      "loss": 0.3547,
+      "step": 6448
+    },
+    {
+      "epoch": 0.47422604603279656,
+      "grad_norm": 0.8594125509262085,
+      "learning_rate": 0.0001131913294372345,
+      "loss": 0.3585,
+      "step": 6449
+    },
+    {
+      "epoch": 0.4742995808515332,
+      "grad_norm": 0.5590276122093201,
+      "learning_rate": 0.00011316772100415462,
+      "loss": 0.3696,
+      "step": 6450
+    },
+    {
+      "epoch": 0.47437311567026985,
+      "grad_norm": 0.8510725498199463,
+      "learning_rate": 0.00011314411182418771,
+      "loss": 0.3717,
+      "step": 6451
+    },
+    {
+      "epoch": 0.4744466504890065,
+      "grad_norm": 1.1186732053756714,
+      "learning_rate": 0.00011312050189867286,
+      "loss": 0.3738,
+      "step": 6452
+    },
+    {
+      "epoch": 0.4745201853077432,
+      "grad_norm": 0.5768918395042419,
+      "learning_rate": 0.00011309689122894938,
+      "loss": 0.3453,
+      "step": 6453
+    },
+    {
+      "epoch": 0.4745937201264799,
+      "grad_norm": 1.5595993995666504,
+      "learning_rate": 0.00011307327981635633,
+      "loss": 0.3514,
+      "step": 6454
+    },
+    {
+      "epoch": 0.47466725494521655,
+      "grad_norm": 0.6877975463867188,
+      "learning_rate": 0.0001130496676622331,
+      "loss": 0.3743,
+      "step": 6455
+    },
+    {
+      "epoch": 0.47474078976395323,
+      "grad_norm": 2.057023763656616,
+      "learning_rate": 0.00011302605476791894,
+      "loss": 0.3682,
+      "step": 6456
+    },
+    {
+      "epoch": 0.4748143245826899,
+      "grad_norm": 1.4734551906585693,
+      "learning_rate": 0.00011300244113475318,
+      "loss": 0.3952,
+      "step": 6457
+    },
+    {
+      "epoch": 0.4748878594014266,
+      "grad_norm": 2.2723987102508545,
+      "learning_rate": 0.00011297882676407528,
+      "loss": 0.3465,
+      "step": 6458
+    },
+    {
+      "epoch": 0.47496139422016326,
+      "grad_norm": 2.3637406826019287,
+      "learning_rate": 0.0001129552116572246,
+      "loss": 0.3685,
+      "step": 6459
+    },
+    {
+      "epoch": 0.47503492903889993,
+      "grad_norm": 0.5755724906921387,
+      "learning_rate": 0.00011293159581554064,
+      "loss": 0.3614,
+      "step": 6460
+    },
+    {
+      "epoch": 0.4751084638576366,
+      "grad_norm": 2.1746108531951904,
+      "learning_rate": 0.00011290797924036293,
+      "loss": 0.3684,
+      "step": 6461
+    },
+    {
+      "epoch": 0.4751819986763733,
+      "grad_norm": 0.6869356632232666,
+      "learning_rate": 0.000112884361933031,
+      "loss": 0.3618,
+      "step": 6462
+    },
+    {
+      "epoch": 0.47525553349510996,
+      "grad_norm": 2.394284725189209,
+      "learning_rate": 0.00011286074389488448,
+      "loss": 0.3525,
+      "step": 6463
+    },
+    {
+      "epoch": 0.4753290683138466,
+      "grad_norm": 2.081216335296631,
+      "learning_rate": 0.00011283712512726299,
+      "loss": 0.3919,
+      "step": 6464
+    },
+    {
+      "epoch": 0.47540260313258326,
+      "grad_norm": 0.7820133566856384,
+      "learning_rate": 0.00011281350563150623,
+      "loss": 0.3843,
+      "step": 6465
+    },
+    {
+      "epoch": 0.47547613795131993,
+      "grad_norm": 1.4690145254135132,
+      "learning_rate": 0.0001127898854089539,
+      "loss": 0.3673,
+      "step": 6466
+    },
+    {
+      "epoch": 0.4755496727700566,
+      "grad_norm": 0.7119136452674866,
+      "learning_rate": 0.00011276626446094573,
+      "loss": 0.3861,
+      "step": 6467
+    },
+    {
+      "epoch": 0.4756232075887933,
+      "grad_norm": 1.383021593093872,
+      "learning_rate": 0.00011274264278882161,
+      "loss": 0.3652,
+      "step": 6468
+    },
+    {
+      "epoch": 0.47569674240752996,
+      "grad_norm": 1.3557339906692505,
+      "learning_rate": 0.00011271902039392135,
+      "loss": 0.3886,
+      "step": 6469
+    },
+    {
+      "epoch": 0.47577027722626664,
+      "grad_norm": 0.695876955986023,
+      "learning_rate": 0.0001126953972775848,
+      "loss": 0.3769,
+      "step": 6470
+    },
+    {
+      "epoch": 0.4758438120450033,
+      "grad_norm": 1.787514328956604,
+      "learning_rate": 0.00011267177344115194,
+      "loss": 0.3868,
+      "step": 6471
+    },
+    {
+      "epoch": 0.47591734686374,
+      "grad_norm": 0.9940299987792969,
+      "learning_rate": 0.00011264814888596272,
+      "loss": 0.3539,
+      "step": 6472
+    },
+    {
+      "epoch": 0.47599088168247666,
+      "grad_norm": 2.2750613689422607,
+      "learning_rate": 0.00011262452361335711,
+      "loss": 0.3516,
+      "step": 6473
+    },
+    {
+      "epoch": 0.47606441650121334,
+      "grad_norm": 2.4438276290893555,
+      "learning_rate": 0.00011260089762467524,
+      "loss": 0.3715,
+      "step": 6474
+    },
+    {
+      "epoch": 0.47613795131995,
+      "grad_norm": 0.6611111164093018,
+      "learning_rate": 0.00011257727092125714,
+      "loss": 0.36,
+      "step": 6475
+    },
+    {
+      "epoch": 0.4762114861386867,
+      "grad_norm": 1.5266249179840088,
+      "learning_rate": 0.00011255364350444297,
+      "loss": 0.3563,
+      "step": 6476
+    },
+    {
+      "epoch": 0.47628502095742337,
+      "grad_norm": 0.6268367171287537,
+      "learning_rate": 0.00011253001537557288,
+      "loss": 0.3738,
+      "step": 6477
+    },
+    {
+      "epoch": 0.47635855577616,
+      "grad_norm": 0.8936167359352112,
+      "learning_rate": 0.0001125063865359871,
+      "loss": 0.3642,
+      "step": 6478
+    },
+    {
+      "epoch": 0.47643209059489666,
+      "grad_norm": 1.3022080659866333,
+      "learning_rate": 0.00011248275698702589,
+      "loss": 0.3798,
+      "step": 6479
+    },
+    {
+      "epoch": 0.47650562541363334,
+      "grad_norm": 0.6298601627349854,
+      "learning_rate": 0.0001124591267300295,
+      "loss": 0.3709,
+      "step": 6480
+    },
+    {
+      "epoch": 0.47657916023237,
+      "grad_norm": 1.380429983139038,
+      "learning_rate": 0.00011243549576633832,
+      "loss": 0.3304,
+      "step": 6481
+    },
+    {
+      "epoch": 0.4766526950511067,
+      "grad_norm": 0.6151654720306396,
+      "learning_rate": 0.00011241186409729268,
+      "loss": 0.3423,
+      "step": 6482
+    },
+    {
+      "epoch": 0.47672622986984337,
+      "grad_norm": 1.3482890129089355,
+      "learning_rate": 0.00011238823172423303,
+      "loss": 0.3485,
+      "step": 6483
+    },
+    {
+      "epoch": 0.47679976468858004,
+      "grad_norm": 0.6233728528022766,
+      "learning_rate": 0.00011236459864849977,
+      "loss": 0.3577,
+      "step": 6484
+    },
+    {
+      "epoch": 0.4768732995073167,
+      "grad_norm": 1.574273705482483,
+      "learning_rate": 0.00011234096487143344,
+      "loss": 0.3225,
+      "step": 6485
+    },
+    {
+      "epoch": 0.4769468343260534,
+      "grad_norm": 0.6472114324569702,
+      "learning_rate": 0.00011231733039437458,
+      "loss": 0.3549,
+      "step": 6486
+    },
+    {
+      "epoch": 0.47702036914479007,
+      "grad_norm": 1.5810860395431519,
+      "learning_rate": 0.0001122936952186637,
+      "loss": 0.3578,
+      "step": 6487
+    },
+    {
+      "epoch": 0.47709390396352674,
+      "grad_norm": 0.8814293742179871,
+      "learning_rate": 0.0001122700593456415,
+      "loss": 0.334,
+      "step": 6488
+    },
+    {
+      "epoch": 0.4771674387822634,
+      "grad_norm": 2.142334222793579,
+      "learning_rate": 0.00011224642277664857,
+      "loss": 0.3525,
+      "step": 6489
+    },
+    {
+      "epoch": 0.4772409736010001,
+      "grad_norm": 1.8698065280914307,
+      "learning_rate": 0.00011222278551302558,
+      "loss": 0.3487,
+      "step": 6490
+    },
+    {
+      "epoch": 0.47731450841973677,
+      "grad_norm": 1.5252610445022583,
+      "learning_rate": 0.00011219914755611334,
+      "loss": 0.3584,
+      "step": 6491
+    },
+    {
+      "epoch": 0.4773880432384734,
+      "grad_norm": 1.2672306299209595,
+      "learning_rate": 0.00011217550890725256,
+      "loss": 0.3643,
+      "step": 6492
+    },
+    {
+      "epoch": 0.47746157805721007,
+      "grad_norm": 1.219599723815918,
+      "learning_rate": 0.00011215186956778405,
+      "loss": 0.3637,
+      "step": 6493
+    },
+    {
+      "epoch": 0.47753511287594674,
+      "grad_norm": 0.9134689569473267,
+      "learning_rate": 0.00011212822953904872,
+      "loss": 0.3825,
+      "step": 6494
+    },
+    {
+      "epoch": 0.4776086476946834,
+      "grad_norm": 1.7088569402694702,
+      "learning_rate": 0.00011210458882238737,
+      "loss": 0.3626,
+      "step": 6495
+    },
+    {
+      "epoch": 0.4776821825134201,
+      "grad_norm": 1.2951550483703613,
+      "learning_rate": 0.00011208094741914097,
+      "loss": 0.3606,
+      "step": 6496
+    },
+    {
+      "epoch": 0.47775571733215677,
+      "grad_norm": 1.5848946571350098,
+      "learning_rate": 0.00011205730533065051,
+      "loss": 0.3578,
+      "step": 6497
+    },
+    {
+      "epoch": 0.47782925215089345,
+      "grad_norm": 1.5606495141983032,
+      "learning_rate": 0.00011203366255825692,
+      "loss": 0.3478,
+      "step": 6498
+    },
+    {
+      "epoch": 0.4779027869696301,
+      "grad_norm": 1.272545337677002,
+      "learning_rate": 0.00011201001910330133,
+      "loss": 0.358,
+      "step": 6499
+    },
+    {
+      "epoch": 0.4779763217883668,
+      "grad_norm": 1.0087202787399292,
+      "learning_rate": 0.00011198637496712477,
+      "loss": 0.3199,
+      "step": 6500
+    },
+    {
+      "epoch": 0.4780498566071035,
+      "grad_norm": 1.5922521352767944,
+      "learning_rate": 0.00011196273015106837,
+      "loss": 0.3919,
+      "step": 6501
+    },
+    {
+      "epoch": 0.47812339142584015,
+      "grad_norm": 1.7364858388900757,
+      "learning_rate": 0.00011193908465647332,
+      "loss": 0.3696,
+      "step": 6502
+    },
+    {
+      "epoch": 0.4781969262445768,
+      "grad_norm": 1.2408910989761353,
+      "learning_rate": 0.00011191543848468074,
+      "loss": 0.3523,
+      "step": 6503
+    },
+    {
+      "epoch": 0.4782704610633135,
+      "grad_norm": 1.251861333847046,
+      "learning_rate": 0.00011189179163703194,
+      "loss": 0.3513,
+      "step": 6504
+    },
+    {
+      "epoch": 0.4783439958820502,
+      "grad_norm": 1.7326555252075195,
+      "learning_rate": 0.00011186814411486817,
+      "loss": 0.3794,
+      "step": 6505
+    },
+    {
+      "epoch": 0.4784175307007868,
+      "grad_norm": 1.4633290767669678,
+      "learning_rate": 0.00011184449591953072,
+      "loss": 0.3586,
+      "step": 6506
+    },
+    {
+      "epoch": 0.4784910655195235,
+      "grad_norm": 1.758223056793213,
+      "learning_rate": 0.000111820847052361,
+      "loss": 0.3954,
+      "step": 6507
+    },
+    {
+      "epoch": 0.47856460033826015,
+      "grad_norm": 1.6517916917800903,
+      "learning_rate": 0.0001117971975147003,
+      "loss": 0.3444,
+      "step": 6508
+    },
+    {
+      "epoch": 0.4786381351569968,
+      "grad_norm": 0.9317672848701477,
+      "learning_rate": 0.00011177354730789016,
+      "loss": 0.4165,
+      "step": 6509
+    },
+    {
+      "epoch": 0.4787116699757335,
+      "grad_norm": 1.1221022605895996,
+      "learning_rate": 0.00011174989643327198,
+      "loss": 0.3584,
+      "step": 6510
+    },
+    {
+      "epoch": 0.4787852047944702,
+      "grad_norm": 1.4398300647735596,
+      "learning_rate": 0.00011172624489218728,
+      "loss": 0.3684,
+      "step": 6511
+    },
+    {
+      "epoch": 0.47885873961320685,
+      "grad_norm": 1.1049264669418335,
+      "learning_rate": 0.00011170259268597758,
+      "loss": 0.3524,
+      "step": 6512
+    },
+    {
+      "epoch": 0.4789322744319435,
+      "grad_norm": 1.8196868896484375,
+      "learning_rate": 0.00011167893981598447,
+      "loss": 0.3799,
+      "step": 6513
+    },
+    {
+      "epoch": 0.4790058092506802,
+      "grad_norm": 1.7297797203063965,
+      "learning_rate": 0.00011165528628354957,
+      "loss": 0.3628,
+      "step": 6514
+    },
+    {
+      "epoch": 0.4790793440694169,
+      "grad_norm": 1.179380178451538,
+      "learning_rate": 0.00011163163209001456,
+      "loss": 0.3462,
+      "step": 6515
+    },
+    {
+      "epoch": 0.47915287888815355,
+      "grad_norm": 1.0519917011260986,
+      "learning_rate": 0.00011160797723672105,
+      "loss": 0.3441,
+      "step": 6516
+    },
+    {
+      "epoch": 0.47922641370689023,
+      "grad_norm": 1.7697153091430664,
+      "learning_rate": 0.00011158432172501087,
+      "loss": 0.3744,
+      "step": 6517
+    },
+    {
+      "epoch": 0.4792999485256269,
+      "grad_norm": 1.2056694030761719,
+      "learning_rate": 0.0001115606655562257,
+      "loss": 0.3313,
+      "step": 6518
+    },
+    {
+      "epoch": 0.4793734833443636,
+      "grad_norm": 1.6434377431869507,
+      "learning_rate": 0.0001115370087317074,
+      "loss": 0.3149,
+      "step": 6519
+    },
+    {
+      "epoch": 0.4794470181631002,
+      "grad_norm": 1.9542101621627808,
+      "learning_rate": 0.0001115133512527978,
+      "loss": 0.3402,
+      "step": 6520
+    },
+    {
+      "epoch": 0.4795205529818369,
+      "grad_norm": 1.0054240226745605,
+      "learning_rate": 0.00011148969312083872,
+      "loss": 0.3554,
+      "step": 6521
+    },
+    {
+      "epoch": 0.47959408780057355,
+      "grad_norm": 0.9951702356338501,
+      "learning_rate": 0.00011146603433717218,
+      "loss": 0.361,
+      "step": 6522
+    },
+    {
+      "epoch": 0.47966762261931023,
+      "grad_norm": 1.389779806137085,
+      "learning_rate": 0.00011144237490314005,
+      "loss": 0.3481,
+      "step": 6523
+    },
+    {
+      "epoch": 0.4797411574380469,
+      "grad_norm": 1.0690706968307495,
+      "learning_rate": 0.00011141871482008432,
+      "loss": 0.3243,
+      "step": 6524
+    },
+    {
+      "epoch": 0.4798146922567836,
+      "grad_norm": 1.7578816413879395,
+      "learning_rate": 0.00011139505408934706,
+      "loss": 0.3563,
+      "step": 6525
+    },
+    {
+      "epoch": 0.47988822707552026,
+      "grad_norm": 1.649330735206604,
+      "learning_rate": 0.00011137139271227028,
+      "loss": 0.3655,
+      "step": 6526
+    },
+    {
+      "epoch": 0.47996176189425693,
+      "grad_norm": 1.0661989450454712,
+      "learning_rate": 0.00011134773069019612,
+      "loss": 0.341,
+      "step": 6527
+    },
+    {
+      "epoch": 0.4800352967129936,
+      "grad_norm": 0.9688524603843689,
+      "learning_rate": 0.00011132406802446671,
+      "loss": 0.3967,
+      "step": 6528
+    },
+    {
+      "epoch": 0.4801088315317303,
+      "grad_norm": 1.7305535078048706,
+      "learning_rate": 0.0001113004047164242,
+      "loss": 0.3651,
+      "step": 6529
+    },
+    {
+      "epoch": 0.48018236635046696,
+      "grad_norm": 1.4272650480270386,
+      "learning_rate": 0.00011127674076741081,
+      "loss": 0.3679,
+      "step": 6530
+    },
+    {
+      "epoch": 0.48025590116920364,
+      "grad_norm": 1.8700480461120605,
+      "learning_rate": 0.00011125307617876876,
+      "loss": 0.3781,
+      "step": 6531
+    },
+    {
+      "epoch": 0.4803294359879403,
+      "grad_norm": 1.8800688982009888,
+      "learning_rate": 0.00011122941095184039,
+      "loss": 0.3675,
+      "step": 6532
+    },
+    {
+      "epoch": 0.480402970806677,
+      "grad_norm": 0.9943934082984924,
+      "learning_rate": 0.00011120574508796797,
+      "loss": 0.3648,
+      "step": 6533
+    },
+    {
+      "epoch": 0.4804765056254136,
+      "grad_norm": 1.1164125204086304,
+      "learning_rate": 0.00011118207858849383,
+      "loss": 0.3394,
+      "step": 6534
+    },
+    {
+      "epoch": 0.4805500404441503,
+      "grad_norm": 1.3361400365829468,
+      "learning_rate": 0.00011115841145476045,
+      "loss": 0.3675,
+      "step": 6535
+    },
+    {
+      "epoch": 0.48062357526288696,
+      "grad_norm": 0.8321199417114258,
+      "learning_rate": 0.00011113474368811014,
+      "loss": 0.3722,
+      "step": 6536
+    },
+    {
+      "epoch": 0.48069711008162364,
+      "grad_norm": 1.6507107019424438,
+      "learning_rate": 0.00011111107528988546,
+      "loss": 0.3544,
+      "step": 6537
+    },
+    {
+      "epoch": 0.4807706449003603,
+      "grad_norm": 1.159364938735962,
+      "learning_rate": 0.00011108740626142886,
+      "loss": 0.3661,
+      "step": 6538
+    },
+    {
+      "epoch": 0.480844179719097,
+      "grad_norm": 1.6838350296020508,
+      "learning_rate": 0.00011106373660408284,
+      "loss": 0.3858,
+      "step": 6539
+    },
+    {
+      "epoch": 0.48091771453783366,
+      "grad_norm": 1.5165119171142578,
+      "learning_rate": 0.00011104006631919004,
+      "loss": 0.3576,
+      "step": 6540
+    },
+    {
+      "epoch": 0.48099124935657034,
+      "grad_norm": 1.308843970298767,
+      "learning_rate": 0.00011101639540809301,
+      "loss": 0.3634,
+      "step": 6541
+    },
+    {
+      "epoch": 0.481064784175307,
+      "grad_norm": 1.049223780632019,
+      "learning_rate": 0.00011099272387213442,
+      "loss": 0.3602,
+      "step": 6542
+    },
+    {
+      "epoch": 0.4811383189940437,
+      "grad_norm": 1.7838454246520996,
+      "learning_rate": 0.00011096905171265693,
+      "loss": 0.3574,
+      "step": 6543
+    },
+    {
+      "epoch": 0.48121185381278037,
+      "grad_norm": 1.7751262187957764,
+      "learning_rate": 0.00011094537893100321,
+      "loss": 0.3714,
+      "step": 6544
+    },
+    {
+      "epoch": 0.48128538863151704,
+      "grad_norm": 1.412024736404419,
+      "learning_rate": 0.00011092170552851608,
+      "loss": 0.3799,
+      "step": 6545
+    },
+    {
+      "epoch": 0.4813589234502537,
+      "grad_norm": 1.2364633083343506,
+      "learning_rate": 0.00011089803150653829,
+      "loss": 0.377,
+      "step": 6546
+    },
+    {
+      "epoch": 0.4814324582689904,
+      "grad_norm": 1.27782142162323,
+      "learning_rate": 0.00011087435686641265,
+      "loss": 0.3404,
+      "step": 6547
+    },
+    {
+      "epoch": 0.481505993087727,
+      "grad_norm": 1.3547998666763306,
+      "learning_rate": 0.00011085068160948202,
+      "loss": 0.3549,
+      "step": 6548
+    },
+    {
+      "epoch": 0.4815795279064637,
+      "grad_norm": 1.2770905494689941,
+      "learning_rate": 0.00011082700573708924,
+      "loss": 0.3795,
+      "step": 6549
+    },
+    {
+      "epoch": 0.48165306272520036,
+      "grad_norm": 1.1836752891540527,
+      "learning_rate": 0.0001108033292505773,
+      "loss": 0.3604,
+      "step": 6550
+    },
+    {
+      "epoch": 0.48172659754393704,
+      "grad_norm": 1.5751628875732422,
+      "learning_rate": 0.00011077965215128911,
+      "loss": 0.3614,
+      "step": 6551
+    },
+    {
+      "epoch": 0.4818001323626737,
+      "grad_norm": 1.1448884010314941,
+      "learning_rate": 0.00011075597444056768,
+      "loss": 0.3779,
+      "step": 6552
+    },
+    {
+      "epoch": 0.4818736671814104,
+      "grad_norm": 1.5752078294754028,
+      "learning_rate": 0.00011073229611975606,
+      "loss": 0.3567,
+      "step": 6553
+    },
+    {
+      "epoch": 0.48194720200014707,
+      "grad_norm": 1.202325463294983,
+      "learning_rate": 0.00011070861719019722,
+      "loss": 0.3776,
+      "step": 6554
+    },
+    {
+      "epoch": 0.48202073681888374,
+      "grad_norm": 1.6928480863571167,
+      "learning_rate": 0.00011068493765323435,
+      "loss": 0.366,
+      "step": 6555
+    },
+    {
+      "epoch": 0.4820942716376204,
+      "grad_norm": 1.593935489654541,
+      "learning_rate": 0.00011066125751021052,
+      "loss": 0.3829,
+      "step": 6556
+    },
+    {
+      "epoch": 0.4821678064563571,
+      "grad_norm": 1.6233108043670654,
+      "learning_rate": 0.00011063757676246891,
+      "loss": 0.3814,
+      "step": 6557
+    },
+    {
+      "epoch": 0.48224134127509377,
+      "grad_norm": 1.7346705198287964,
+      "learning_rate": 0.00011061389541135276,
+      "loss": 0.3653,
+      "step": 6558
+    },
+    {
+      "epoch": 0.48231487609383045,
+      "grad_norm": 1.113072156906128,
+      "learning_rate": 0.00011059021345820521,
+      "loss": 0.3937,
+      "step": 6559
+    },
+    {
+      "epoch": 0.4823884109125671,
+      "grad_norm": 0.9684404730796814,
+      "learning_rate": 0.00011056653090436962,
+      "loss": 0.3392,
+      "step": 6560
+    },
+    {
+      "epoch": 0.4824619457313038,
+      "grad_norm": 1.4340364933013916,
+      "learning_rate": 0.00011054284775118922,
+      "loss": 0.3936,
+      "step": 6561
+    },
+    {
+      "epoch": 0.4825354805500404,
+      "grad_norm": 0.9922011494636536,
+      "learning_rate": 0.00011051916400000737,
+      "loss": 0.3766,
+      "step": 6562
+    },
+    {
+      "epoch": 0.4826090153687771,
+      "grad_norm": 1.9103896617889404,
+      "learning_rate": 0.00011049547965216745,
+      "loss": 0.342,
+      "step": 6563
+    },
+    {
+      "epoch": 0.48268255018751377,
+      "grad_norm": 1.8280268907546997,
+      "learning_rate": 0.00011047179470901284,
+      "loss": 0.3587,
+      "step": 6564
+    },
+    {
+      "epoch": 0.48275608500625045,
+      "grad_norm": 1.16150963306427,
+      "learning_rate": 0.00011044810917188697,
+      "loss": 0.3447,
+      "step": 6565
+    },
+    {
+      "epoch": 0.4828296198249871,
+      "grad_norm": 1.3103257417678833,
+      "learning_rate": 0.00011042442304213335,
+      "loss": 0.3973,
+      "step": 6566
+    },
+    {
+      "epoch": 0.4829031546437238,
+      "grad_norm": 1.132474660873413,
+      "learning_rate": 0.00011040073632109542,
+      "loss": 0.3831,
+      "step": 6567
+    },
+    {
+      "epoch": 0.4829766894624605,
+      "grad_norm": 0.913303792476654,
+      "learning_rate": 0.00011037704901011676,
+      "loss": 0.3548,
+      "step": 6568
+    },
+    {
+      "epoch": 0.48305022428119715,
+      "grad_norm": 1.4507077932357788,
+      "learning_rate": 0.00011035336111054096,
+      "loss": 0.378,
+      "step": 6569
+    },
+    {
+      "epoch": 0.4831237590999338,
+      "grad_norm": 0.8300760388374329,
+      "learning_rate": 0.00011032967262371155,
+      "loss": 0.3739,
+      "step": 6570
+    },
+    {
+      "epoch": 0.4831972939186705,
+      "grad_norm": 1.678439974784851,
+      "learning_rate": 0.00011030598355097224,
+      "loss": 0.3564,
+      "step": 6571
+    },
+    {
+      "epoch": 0.4832708287374072,
+      "grad_norm": 1.1094540357589722,
+      "learning_rate": 0.00011028229389366662,
+      "loss": 0.3554,
+      "step": 6572
+    },
+    {
+      "epoch": 0.48334436355614385,
+      "grad_norm": 2.1585116386413574,
+      "learning_rate": 0.00011025860365313848,
+      "loss": 0.3559,
+      "step": 6573
+    },
+    {
+      "epoch": 0.48341789837488053,
+      "grad_norm": 1.6977189779281616,
+      "learning_rate": 0.00011023491283073151,
+      "loss": 0.3544,
+      "step": 6574
+    },
+    {
+      "epoch": 0.4834914331936172,
+      "grad_norm": 1.1602743864059448,
+      "learning_rate": 0.00011021122142778945,
+      "loss": 0.358,
+      "step": 6575
+    },
+    {
+      "epoch": 0.4835649680123538,
+      "grad_norm": 1.1484191417694092,
+      "learning_rate": 0.0001101875294456562,
+      "loss": 0.365,
+      "step": 6576
+    },
+    {
+      "epoch": 0.4836385028310905,
+      "grad_norm": 1.744558334350586,
+      "learning_rate": 0.00011016383688567546,
+      "loss": 0.4247,
+      "step": 6577
+    },
+    {
+      "epoch": 0.4837120376498272,
+      "grad_norm": 1.4004101753234863,
+      "learning_rate": 0.00011014014374919122,
+      "loss": 0.411,
+      "step": 6578
+    },
+    {
+      "epoch": 0.48378557246856385,
+      "grad_norm": 1.6536494493484497,
+      "learning_rate": 0.0001101164500375473,
+      "loss": 0.4279,
+      "step": 6579
+    },
+    {
+      "epoch": 0.4838591072873005,
+      "grad_norm": 1.7462927103042603,
+      "learning_rate": 0.00011009275575208764,
+      "loss": 0.3688,
+      "step": 6580
+    },
+    {
+      "epoch": 0.4839326421060372,
+      "grad_norm": 1.3496601581573486,
+      "learning_rate": 0.00011006906089415627,
+      "loss": 0.3877,
+      "step": 6581
+    },
+    {
+      "epoch": 0.4840061769247739,
+      "grad_norm": 1.444054365158081,
+      "learning_rate": 0.00011004536546509709,
+      "loss": 0.3697,
+      "step": 6582
+    },
+    {
+      "epoch": 0.48407971174351055,
+      "grad_norm": 1.1283388137817383,
+      "learning_rate": 0.00011002166946625421,
+      "loss": 0.3548,
+      "step": 6583
+    },
+    {
+      "epoch": 0.48415324656224723,
+      "grad_norm": 1.0878881216049194,
+      "learning_rate": 0.00010999797289897164,
+      "loss": 0.3639,
+      "step": 6584
+    },
+    {
+      "epoch": 0.4842267813809839,
+      "grad_norm": 1.3191925287246704,
+      "learning_rate": 0.0001099742757645935,
+      "loss": 0.3712,
+      "step": 6585
+    },
+    {
+      "epoch": 0.4843003161997206,
+      "grad_norm": 1.0794892311096191,
+      "learning_rate": 0.00010995057806446392,
+      "loss": 0.3783,
+      "step": 6586
+    },
+    {
+      "epoch": 0.48437385101845726,
+      "grad_norm": 1.7437009811401367,
+      "learning_rate": 0.00010992687979992706,
+      "loss": 0.3952,
+      "step": 6587
+    },
+    {
+      "epoch": 0.48444738583719393,
+      "grad_norm": 1.4378113746643066,
+      "learning_rate": 0.0001099031809723271,
+      "loss": 0.3701,
+      "step": 6588
+    },
+    {
+      "epoch": 0.4845209206559306,
+      "grad_norm": 1.611843228340149,
+      "learning_rate": 0.00010987948158300825,
+      "loss": 0.3453,
+      "step": 6589
+    },
+    {
+      "epoch": 0.48459445547466723,
+      "grad_norm": 1.3362417221069336,
+      "learning_rate": 0.00010985578163331478,
+      "loss": 0.3643,
+      "step": 6590
+    },
+    {
+      "epoch": 0.4846679902934039,
+      "grad_norm": 1.4339872598648071,
+      "learning_rate": 0.00010983208112459097,
+      "loss": 0.3599,
+      "step": 6591
+    },
+    {
+      "epoch": 0.4847415251121406,
+      "grad_norm": 1.3197669982910156,
+      "learning_rate": 0.00010980838005818116,
+      "loss": 0.3631,
+      "step": 6592
+    },
+    {
+      "epoch": 0.48481505993087726,
+      "grad_norm": 1.2044730186462402,
+      "learning_rate": 0.00010978467843542968,
+      "loss": 0.3449,
+      "step": 6593
+    },
+    {
+      "epoch": 0.48488859474961393,
+      "grad_norm": 1.0924413204193115,
+      "learning_rate": 0.00010976097625768092,
+      "loss": 0.3611,
+      "step": 6594
+    },
+    {
+      "epoch": 0.4849621295683506,
+      "grad_norm": 1.5209636688232422,
+      "learning_rate": 0.00010973727352627925,
+      "loss": 0.3859,
+      "step": 6595
+    },
+    {
+      "epoch": 0.4850356643870873,
+      "grad_norm": 1.494430661201477,
+      "learning_rate": 0.00010971357024256916,
+      "loss": 0.3767,
+      "step": 6596
+    },
+    {
+      "epoch": 0.48510919920582396,
+      "grad_norm": 1.3342406749725342,
+      "learning_rate": 0.00010968986640789513,
+      "loss": 0.3534,
+      "step": 6597
+    },
+    {
+      "epoch": 0.48518273402456064,
+      "grad_norm": 1.472252607345581,
+      "learning_rate": 0.00010966616202360165,
+      "loss": 0.3598,
+      "step": 6598
+    },
+    {
+      "epoch": 0.4852562688432973,
+      "grad_norm": 0.9780088067054749,
+      "learning_rate": 0.00010964245709103327,
+      "loss": 0.3882,
+      "step": 6599
+    },
+    {
+      "epoch": 0.485329803662034,
+      "grad_norm": 1.1021552085876465,
+      "learning_rate": 0.00010961875161153452,
+      "loss": 0.3557,
+      "step": 6600
+    },
+    {
+      "epoch": 0.48540333848077066,
+      "grad_norm": 1.5003646612167358,
+      "learning_rate": 0.00010959504558645004,
+      "loss": 0.3878,
+      "step": 6601
+    },
+    {
+      "epoch": 0.48547687329950734,
+      "grad_norm": 1.0533802509307861,
+      "learning_rate": 0.00010957133901712446,
+      "loss": 0.3812,
+      "step": 6602
+    },
+    {
+      "epoch": 0.485550408118244,
+      "grad_norm": 1.7154145240783691,
+      "learning_rate": 0.0001095476319049024,
+      "loss": 0.3641,
+      "step": 6603
+    },
+    {
+      "epoch": 0.48562394293698063,
+      "grad_norm": 1.398852825164795,
+      "learning_rate": 0.00010952392425112861,
+      "loss": 0.3851,
+      "step": 6604
+    },
+    {
+      "epoch": 0.4856974777557173,
+      "grad_norm": 1.6558713912963867,
+      "learning_rate": 0.00010950021605714778,
+      "loss": 0.3815,
+      "step": 6605
+    },
+    {
+      "epoch": 0.485771012574454,
+      "grad_norm": 1.6472728252410889,
+      "learning_rate": 0.00010947650732430468,
+      "loss": 0.3725,
+      "step": 6606
+    },
+    {
+      "epoch": 0.48584454739319066,
+      "grad_norm": 1.038486123085022,
+      "learning_rate": 0.00010945279805394406,
+      "loss": 0.3743,
+      "step": 6607
+    },
+    {
+      "epoch": 0.48591808221192734,
+      "grad_norm": 1.1314597129821777,
+      "learning_rate": 0.00010942908824741076,
+      "loss": 0.3921,
+      "step": 6608
+    },
+    {
+      "epoch": 0.485991617030664,
+      "grad_norm": 1.3320293426513672,
+      "learning_rate": 0.00010940537790604963,
+      "loss": 0.3519,
+      "step": 6609
+    },
+    {
+      "epoch": 0.4860651518494007,
+      "grad_norm": 1.050055742263794,
+      "learning_rate": 0.00010938166703120555,
+      "loss": 0.373,
+      "step": 6610
+    },
+    {
+      "epoch": 0.48613868666813737,
+      "grad_norm": 2.340153455734253,
+      "learning_rate": 0.00010935795562422342,
+      "loss": 0.3613,
+      "step": 6611
+    },
+    {
+      "epoch": 0.48621222148687404,
+      "grad_norm": 2.184659481048584,
+      "learning_rate": 0.00010933424368644816,
+      "loss": 0.3998,
+      "step": 6612
+    },
+    {
+      "epoch": 0.4862857563056107,
+      "grad_norm": 0.7612171173095703,
+      "learning_rate": 0.00010931053121922472,
+      "loss": 0.3949,
+      "step": 6613
+    },
+    {
+      "epoch": 0.4863592911243474,
+      "grad_norm": 2.1237258911132812,
+      "learning_rate": 0.00010928681822389815,
+      "loss": 0.3882,
+      "step": 6614
+    },
+    {
+      "epoch": 0.48643282594308407,
+      "grad_norm": 0.8275244235992432,
+      "learning_rate": 0.00010926310470181346,
+      "loss": 0.3892,
+      "step": 6615
+    },
+    {
+      "epoch": 0.48650636076182074,
+      "grad_norm": 1.5145549774169922,
+      "learning_rate": 0.00010923939065431565,
+      "loss": 0.3531,
+      "step": 6616
+    },
+    {
+      "epoch": 0.4865798955805574,
+      "grad_norm": 1.302581787109375,
+      "learning_rate": 0.00010921567608274991,
+      "loss": 0.3737,
+      "step": 6617
+    },
+    {
+      "epoch": 0.48665343039929404,
+      "grad_norm": 1.1668572425842285,
+      "learning_rate": 0.00010919196098846127,
+      "loss": 0.386,
+      "step": 6618
+    },
+    {
+      "epoch": 0.4867269652180307,
+      "grad_norm": 1.2487152814865112,
+      "learning_rate": 0.00010916824537279491,
+      "loss": 0.3626,
+      "step": 6619
+    },
+    {
+      "epoch": 0.4868005000367674,
+      "grad_norm": 1.2679682970046997,
+      "learning_rate": 0.000109144529237096,
+      "loss": 0.3868,
+      "step": 6620
+    },
+    {
+      "epoch": 0.48687403485550407,
+      "grad_norm": 1.2098082304000854,
+      "learning_rate": 0.00010912081258270972,
+      "loss": 0.4093,
+      "step": 6621
+    },
+    {
+      "epoch": 0.48694756967424074,
+      "grad_norm": 1.3701062202453613,
+      "learning_rate": 0.00010909709541098136,
+      "loss": 0.3729,
+      "step": 6622
+    },
+    {
+      "epoch": 0.4870211044929774,
+      "grad_norm": 1.1515161991119385,
+      "learning_rate": 0.00010907337772325609,
+      "loss": 0.3673,
+      "step": 6623
+    },
+    {
+      "epoch": 0.4870946393117141,
+      "grad_norm": 1.301115870475769,
+      "learning_rate": 0.00010904965952087931,
+      "loss": 0.387,
+      "step": 6624
+    },
+    {
+      "epoch": 0.48716817413045077,
+      "grad_norm": 0.9396963119506836,
+      "learning_rate": 0.00010902594080519627,
+      "loss": 0.3757,
+      "step": 6625
+    },
+    {
+      "epoch": 0.48724170894918745,
+      "grad_norm": 1.2665737867355347,
+      "learning_rate": 0.00010900222157755234,
+      "loss": 0.3685,
+      "step": 6626
+    },
+    {
+      "epoch": 0.4873152437679241,
+      "grad_norm": 1.0557612180709839,
+      "learning_rate": 0.00010897850183929293,
+      "loss": 0.392,
+      "step": 6627
+    },
+    {
+      "epoch": 0.4873887785866608,
+      "grad_norm": 0.8316888809204102,
+      "learning_rate": 0.0001089547815917634,
+      "loss": 0.4012,
+      "step": 6628
+    },
+    {
+      "epoch": 0.4874623134053975,
+      "grad_norm": 0.9362977147102356,
+      "learning_rate": 0.0001089310608363092,
+      "loss": 0.3911,
+      "step": 6629
+    },
+    {
+      "epoch": 0.48753584822413415,
+      "grad_norm": 0.7548674941062927,
+      "learning_rate": 0.00010890733957427581,
+      "loss": 0.3759,
+      "step": 6630
+    },
+    {
+      "epoch": 0.4876093830428708,
+      "grad_norm": 1.0554343461990356,
+      "learning_rate": 0.00010888361780700871,
+      "loss": 0.379,
+      "step": 6631
+    },
+    {
+      "epoch": 0.48768291786160745,
+      "grad_norm": 0.7697526216506958,
+      "learning_rate": 0.00010885989553585345,
+      "loss": 0.3712,
+      "step": 6632
+    },
+    {
+      "epoch": 0.4877564526803441,
+      "grad_norm": 0.7613747119903564,
+      "learning_rate": 0.00010883617276215556,
+      "loss": 0.3769,
+      "step": 6633
+    },
+    {
+      "epoch": 0.4878299874990808,
+      "grad_norm": 1.0611449480056763,
+      "learning_rate": 0.00010881244948726063,
+      "loss": 0.3681,
+      "step": 6634
+    },
+    {
+      "epoch": 0.4879035223178175,
+      "grad_norm": 0.5617631077766418,
+      "learning_rate": 0.00010878872571251425,
+      "loss": 0.3525,
+      "step": 6635
+    },
+    {
+      "epoch": 0.48797705713655415,
+      "grad_norm": 1.1440975666046143,
+      "learning_rate": 0.00010876500143926209,
+      "loss": 0.3937,
+      "step": 6636
+    },
+    {
+      "epoch": 0.4880505919552908,
+      "grad_norm": 0.6019554138183594,
+      "learning_rate": 0.00010874127666884978,
+      "loss": 0.4106,
+      "step": 6637
+    },
+    {
+      "epoch": 0.4881241267740275,
+      "grad_norm": 0.960480809211731,
+      "learning_rate": 0.00010871755140262305,
+      "loss": 0.3913,
+      "step": 6638
+    },
+    {
+      "epoch": 0.4881976615927642,
+      "grad_norm": 0.7443310022354126,
+      "learning_rate": 0.00010869382564192759,
+      "loss": 0.4044,
+      "step": 6639
+    },
+    {
+      "epoch": 0.48827119641150085,
+      "grad_norm": 0.6075333952903748,
+      "learning_rate": 0.0001086700993881092,
+      "loss": 0.3668,
+      "step": 6640
+    },
+    {
+      "epoch": 0.4883447312302375,
+      "grad_norm": 0.6012383103370667,
+      "learning_rate": 0.00010864637264251356,
+      "loss": 0.3959,
+      "step": 6641
+    },
+    {
+      "epoch": 0.4884182660489742,
+      "grad_norm": 0.6704562902450562,
+      "learning_rate": 0.00010862264540648659,
+      "loss": 0.3903,
+      "step": 6642
+    },
+    {
+      "epoch": 0.4884918008677109,
+      "grad_norm": 0.4765653908252716,
+      "learning_rate": 0.00010859891768137407,
+      "loss": 0.388,
+      "step": 6643
+    },
+    {
+      "epoch": 0.48856533568644755,
+      "grad_norm": 0.747025728225708,
+      "learning_rate": 0.00010857518946852181,
+      "loss": 0.361,
+      "step": 6644
+    },
+    {
+      "epoch": 0.48863887050518423,
+      "grad_norm": 0.7676913738250732,
+      "learning_rate": 0.0001085514607692758,
+      "loss": 0.3701,
+      "step": 6645
+    },
+    {
+      "epoch": 0.48871240532392085,
+      "grad_norm": 0.9545100331306458,
+      "learning_rate": 0.00010852773158498193,
+      "loss": 0.3804,
+      "step": 6646
+    },
+    {
+      "epoch": 0.4887859401426575,
+      "grad_norm": 0.5239514112472534,
+      "learning_rate": 0.00010850400191698609,
+      "loss": 0.3616,
+      "step": 6647
+    },
+    {
+      "epoch": 0.4888594749613942,
+      "grad_norm": 0.8227052092552185,
+      "learning_rate": 0.00010848027176663431,
+      "loss": 0.3732,
+      "step": 6648
+    },
+    {
+      "epoch": 0.4889330097801309,
+      "grad_norm": 0.604997456073761,
+      "learning_rate": 0.00010845654113527253,
+      "loss": 0.3703,
+      "step": 6649
+    },
+    {
+      "epoch": 0.48900654459886755,
+      "grad_norm": 0.7740441560745239,
+      "learning_rate": 0.00010843281002424684,
+      "loss": 0.3687,
+      "step": 6650
+    },
+    {
+      "epoch": 0.48908007941760423,
+      "grad_norm": 1.064470648765564,
+      "learning_rate": 0.00010840907843490326,
+      "loss": 0.3925,
+      "step": 6651
+    },
+    {
+      "epoch": 0.4891536142363409,
+      "grad_norm": 0.5289266705513,
+      "learning_rate": 0.00010838534636858788,
+      "loss": 0.373,
+      "step": 6652
+    },
+    {
+      "epoch": 0.4892271490550776,
+      "grad_norm": 0.8663640022277832,
+      "learning_rate": 0.00010836161382664681,
+      "loss": 0.4016,
+      "step": 6653
+    },
+    {
+      "epoch": 0.48930068387381426,
+      "grad_norm": 0.6355032324790955,
+      "learning_rate": 0.00010833788081042615,
+      "loss": 0.3842,
+      "step": 6654
+    },
+    {
+      "epoch": 0.48937421869255093,
+      "grad_norm": 0.8897100687026978,
+      "learning_rate": 0.00010831414732127212,
+      "loss": 0.3591,
+      "step": 6655
+    },
+    {
+      "epoch": 0.4894477535112876,
+      "grad_norm": 1.1753257513046265,
+      "learning_rate": 0.00010829041336053087,
+      "loss": 0.3664,
+      "step": 6656
+    },
+    {
+      "epoch": 0.4895212883300243,
+      "grad_norm": 0.7642697095870972,
+      "learning_rate": 0.00010826667892954863,
+      "loss": 0.3788,
+      "step": 6657
+    },
+    {
+      "epoch": 0.48959482314876096,
+      "grad_norm": 0.8818247318267822,
+      "learning_rate": 0.00010824294402967164,
+      "loss": 0.3596,
+      "step": 6658
+    },
+    {
+      "epoch": 0.48966835796749764,
+      "grad_norm": 0.7959984540939331,
+      "learning_rate": 0.00010821920866224613,
+      "loss": 0.3998,
+      "step": 6659
+    },
+    {
+      "epoch": 0.48974189278623426,
+      "grad_norm": 0.7239900827407837,
+      "learning_rate": 0.00010819547282861848,
+      "loss": 0.3874,
+      "step": 6660
+    },
+    {
+      "epoch": 0.48981542760497093,
+      "grad_norm": 1.4080150127410889,
+      "learning_rate": 0.00010817173653013494,
+      "loss": 0.3529,
+      "step": 6661
+    },
+    {
+      "epoch": 0.4898889624237076,
+      "grad_norm": 0.4969620108604431,
+      "learning_rate": 0.00010814799976814187,
+      "loss": 0.3704,
+      "step": 6662
+    },
+    {
+      "epoch": 0.4899624972424443,
+      "grad_norm": 1.665450096130371,
+      "learning_rate": 0.0001081242625439857,
+      "loss": 0.3992,
+      "step": 6663
+    },
+    {
+      "epoch": 0.49003603206118096,
+      "grad_norm": 0.49376189708709717,
+      "learning_rate": 0.00010810052485901272,
+      "loss": 0.3731,
+      "step": 6664
+    },
+    {
+      "epoch": 0.49010956687991764,
+      "grad_norm": 1.3748871088027954,
+      "learning_rate": 0.00010807678671456944,
+      "loss": 0.3767,
+      "step": 6665
+    },
+    {
+      "epoch": 0.4901831016986543,
+      "grad_norm": 0.5038865804672241,
+      "learning_rate": 0.00010805304811200231,
+      "loss": 0.3833,
+      "step": 6666
+    },
+    {
+      "epoch": 0.490256636517391,
+      "grad_norm": 1.5035758018493652,
+      "learning_rate": 0.00010802930905265778,
+      "loss": 0.3963,
+      "step": 6667
+    },
+    {
+      "epoch": 0.49033017133612766,
+      "grad_norm": 0.9026893973350525,
+      "learning_rate": 0.00010800556953788238,
+      "loss": 0.3854,
+      "step": 6668
+    },
+    {
+      "epoch": 0.49040370615486434,
+      "grad_norm": 0.6276199221611023,
+      "learning_rate": 0.00010798182956902263,
+      "loss": 0.4364,
+      "step": 6669
+    },
+    {
+      "epoch": 0.490477240973601,
+      "grad_norm": 0.7900773286819458,
+      "learning_rate": 0.00010795808914742509,
+      "loss": 0.3963,
+      "step": 6670
+    },
+    {
+      "epoch": 0.4905507757923377,
+      "grad_norm": 0.7878774404525757,
+      "learning_rate": 0.00010793434827443631,
+      "loss": 0.3744,
+      "step": 6671
+    },
+    {
+      "epoch": 0.49062431061107437,
+      "grad_norm": 0.8635752201080322,
+      "learning_rate": 0.00010791060695140292,
+      "loss": 0.3668,
+      "step": 6672
+    },
+    {
+      "epoch": 0.49069784542981104,
+      "grad_norm": 0.7378910779953003,
+      "learning_rate": 0.00010788686517967157,
+      "loss": 0.403,
+      "step": 6673
+    },
+    {
+      "epoch": 0.49077138024854766,
+      "grad_norm": 0.7976657152175903,
+      "learning_rate": 0.00010786312296058892,
+      "loss": 0.4145,
+      "step": 6674
+    },
+    {
+      "epoch": 0.49084491506728434,
+      "grad_norm": 0.9590200781822205,
+      "learning_rate": 0.0001078393802955016,
+      "loss": 0.378,
+      "step": 6675
+    },
+    {
+      "epoch": 0.490918449886021,
+      "grad_norm": 0.7053317427635193,
+      "learning_rate": 0.00010781563718575638,
+      "loss": 0.3673,
+      "step": 6676
+    },
+    {
+      "epoch": 0.4909919847047577,
+      "grad_norm": 1.2785788774490356,
+      "learning_rate": 0.00010779189363269994,
+      "loss": 0.369,
+      "step": 6677
+    },
+    {
+      "epoch": 0.49106551952349436,
+      "grad_norm": 0.761652410030365,
+      "learning_rate": 0.00010776814963767912,
+      "loss": 0.3722,
+      "step": 6678
+    },
+    {
+      "epoch": 0.49113905434223104,
+      "grad_norm": 0.9270232319831848,
+      "learning_rate": 0.00010774440520204063,
+      "loss": 0.3868,
+      "step": 6679
+    },
+    {
+      "epoch": 0.4912125891609677,
+      "grad_norm": 1.097195029258728,
+      "learning_rate": 0.00010772066032713129,
+      "loss": 0.3738,
+      "step": 6680
+    },
+    {
+      "epoch": 0.4912861239797044,
+      "grad_norm": 1.2991851568222046,
+      "learning_rate": 0.000107696915014298,
+      "loss": 0.3432,
+      "step": 6681
+    },
+    {
+      "epoch": 0.49135965879844107,
+      "grad_norm": 0.793159544467926,
+      "learning_rate": 0.0001076731692648875,
+      "loss": 0.385,
+      "step": 6682
+    },
+    {
+      "epoch": 0.49143319361717774,
+      "grad_norm": 1.0709863901138306,
+      "learning_rate": 0.00010764942308024677,
+      "loss": 0.3642,
+      "step": 6683
+    },
+    {
+      "epoch": 0.4915067284359144,
+      "grad_norm": 0.8550132513046265,
+      "learning_rate": 0.0001076256764617227,
+      "loss": 0.3654,
+      "step": 6684
+    },
+    {
+      "epoch": 0.4915802632546511,
+      "grad_norm": 1.2397851943969727,
+      "learning_rate": 0.00010760192941066219,
+      "loss": 0.3748,
+      "step": 6685
+    },
+    {
+      "epoch": 0.49165379807338777,
+      "grad_norm": 0.7074998617172241,
+      "learning_rate": 0.00010757818192841224,
+      "loss": 0.3821,
+      "step": 6686
+    },
+    {
+      "epoch": 0.49172733289212445,
+      "grad_norm": 1.012372374534607,
+      "learning_rate": 0.00010755443401631981,
+      "loss": 0.368,
+      "step": 6687
+    },
+    {
+      "epoch": 0.49180086771086107,
+      "grad_norm": 0.8648008704185486,
+      "learning_rate": 0.00010753068567573193,
+      "loss": 0.4047,
+      "step": 6688
+    },
+    {
+      "epoch": 0.49187440252959774,
+      "grad_norm": 0.633052408695221,
+      "learning_rate": 0.00010750693690799559,
+      "loss": 0.393,
+      "step": 6689
+    },
+    {
+      "epoch": 0.4919479373483344,
+      "grad_norm": 0.6513110995292664,
+      "learning_rate": 0.00010748318771445784,
+      "loss": 0.4066,
+      "step": 6690
+    },
+    {
+      "epoch": 0.4920214721670711,
+      "grad_norm": 0.5243488550186157,
+      "learning_rate": 0.00010745943809646582,
+      "loss": 0.3848,
+      "step": 6691
+    },
+    {
+      "epoch": 0.49209500698580777,
+      "grad_norm": 0.8013033270835876,
+      "learning_rate": 0.00010743568805536659,
+      "loss": 0.3758,
+      "step": 6692
+    },
+    {
+      "epoch": 0.49216854180454445,
+      "grad_norm": 1.122584581375122,
+      "learning_rate": 0.0001074119375925073,
+      "loss": 0.3782,
+      "step": 6693
+    },
+    {
+      "epoch": 0.4922420766232811,
+      "grad_norm": 0.7113456130027771,
+      "learning_rate": 0.00010738818670923508,
+      "loss": 0.3635,
+      "step": 6694
+    },
+    {
+      "epoch": 0.4923156114420178,
+      "grad_norm": 0.6295540928840637,
+      "learning_rate": 0.00010736443540689708,
+      "loss": 0.3763,
+      "step": 6695
+    },
+    {
+      "epoch": 0.4923891462607545,
+      "grad_norm": 0.6996574401855469,
+      "learning_rate": 0.00010734068368684058,
+      "loss": 0.4043,
+      "step": 6696
+    },
+    {
+      "epoch": 0.49246268107949115,
+      "grad_norm": 0.5112796425819397,
+      "learning_rate": 0.00010731693155041274,
+      "loss": 0.379,
+      "step": 6697
+    },
+    {
+      "epoch": 0.4925362158982278,
+      "grad_norm": 0.5560314655303955,
+      "learning_rate": 0.00010729317899896081,
+      "loss": 0.3567,
+      "step": 6698
+    },
+    {
+      "epoch": 0.4926097507169645,
+      "grad_norm": 0.6168650388717651,
+      "learning_rate": 0.00010726942603383208,
+      "loss": 0.4022,
+      "step": 6699
+    },
+    {
+      "epoch": 0.4926832855357012,
+      "grad_norm": 0.6529182195663452,
+      "learning_rate": 0.00010724567265637382,
+      "loss": 0.3706,
+      "step": 6700
+    },
+    {
+      "epoch": 0.49275682035443785,
+      "grad_norm": 0.5462172627449036,
+      "learning_rate": 0.00010722191886793338,
+      "loss": 0.3979,
+      "step": 6701
+    },
+    {
+      "epoch": 0.4928303551731745,
+      "grad_norm": 0.5998204350471497,
+      "learning_rate": 0.00010719816466985806,
+      "loss": 0.3962,
+      "step": 6702
+    },
+    {
+      "epoch": 0.49290388999191115,
+      "grad_norm": 0.5560969710350037,
+      "learning_rate": 0.00010717441006349525,
+      "loss": 0.4003,
+      "step": 6703
+    },
+    {
+      "epoch": 0.4929774248106478,
+      "grad_norm": 0.5943616628646851,
+      "learning_rate": 0.00010715065505019235,
+      "loss": 0.4067,
+      "step": 6704
+    },
+    {
+      "epoch": 0.4930509596293845,
+      "grad_norm": 0.8101683259010315,
+      "learning_rate": 0.00010712689963129672,
+      "loss": 0.3798,
+      "step": 6705
+    },
+    {
+      "epoch": 0.4931244944481212,
+      "grad_norm": 0.6069464683532715,
+      "learning_rate": 0.00010710314380815582,
+      "loss": 0.385,
+      "step": 6706
+    },
+    {
+      "epoch": 0.49319802926685785,
+      "grad_norm": 0.4723527133464813,
+      "learning_rate": 0.00010707938758211713,
+      "loss": 0.3943,
+      "step": 6707
+    },
+    {
+      "epoch": 0.4932715640855945,
+      "grad_norm": 0.6221211552619934,
+      "learning_rate": 0.00010705563095452808,
+      "loss": 0.3847,
+      "step": 6708
+    },
+    {
+      "epoch": 0.4933450989043312,
+      "grad_norm": 0.7763896584510803,
+      "learning_rate": 0.00010703187392673622,
+      "loss": 0.3799,
+      "step": 6709
+    },
+    {
+      "epoch": 0.4934186337230679,
+      "grad_norm": 0.5191657543182373,
+      "learning_rate": 0.00010700811650008904,
+      "loss": 0.3835,
+      "step": 6710
+    },
+    {
+      "epoch": 0.49349216854180455,
+      "grad_norm": 0.7642470598220825,
+      "learning_rate": 0.0001069843586759341,
+      "loss": 0.3879,
+      "step": 6711
+    },
+    {
+      "epoch": 0.49356570336054123,
+      "grad_norm": 1.1419706344604492,
+      "learning_rate": 0.00010696060045561898,
+      "loss": 0.3994,
+      "step": 6712
+    },
+    {
+      "epoch": 0.4936392381792779,
+      "grad_norm": 0.6050761342048645,
+      "learning_rate": 0.00010693684184049122,
+      "loss": 0.3713,
+      "step": 6713
+    },
+    {
+      "epoch": 0.4937127729980146,
+      "grad_norm": 0.8644517064094543,
+      "learning_rate": 0.0001069130828318985,
+      "loss": 0.356,
+      "step": 6714
+    },
+    {
+      "epoch": 0.49378630781675126,
+      "grad_norm": 1.241900086402893,
+      "learning_rate": 0.0001068893234311884,
+      "loss": 0.361,
+      "step": 6715
+    },
+    {
+      "epoch": 0.4938598426354879,
+      "grad_norm": 0.4970303475856781,
+      "learning_rate": 0.00010686556363970864,
+      "loss": 0.3875,
+      "step": 6716
+    },
+    {
+      "epoch": 0.49393337745422455,
+      "grad_norm": 1.2793163061141968,
+      "learning_rate": 0.00010684180345880683,
+      "loss": 0.3793,
+      "step": 6717
+    },
+    {
+      "epoch": 0.49400691227296123,
+      "grad_norm": 0.7676216959953308,
+      "learning_rate": 0.0001068180428898307,
+      "loss": 0.3887,
+      "step": 6718
+    },
+    {
+      "epoch": 0.4940804470916979,
+      "grad_norm": 0.46495380997657776,
+      "learning_rate": 0.00010679428193412802,
+      "loss": 0.3756,
+      "step": 6719
+    },
+    {
+      "epoch": 0.4941539819104346,
+      "grad_norm": 0.8968109488487244,
+      "learning_rate": 0.00010677052059304649,
+      "loss": 0.3563,
+      "step": 6720
+    },
+    {
+      "epoch": 0.49422751672917126,
+      "grad_norm": 0.9782388806343079,
+      "learning_rate": 0.00010674675886793386,
+      "loss": 0.361,
+      "step": 6721
+    },
+    {
+      "epoch": 0.49430105154790793,
+      "grad_norm": 0.588751494884491,
+      "learning_rate": 0.000106722996760138,
+      "loss": 0.3642,
+      "step": 6722
+    },
+    {
+      "epoch": 0.4943745863666446,
+      "grad_norm": 0.7599904537200928,
+      "learning_rate": 0.00010669923427100661,
+      "loss": 0.3541,
+      "step": 6723
+    },
+    {
+      "epoch": 0.4944481211853813,
+      "grad_norm": 0.7569181323051453,
+      "learning_rate": 0.00010667547140188761,
+      "loss": 0.359,
+      "step": 6724
+    },
+    {
+      "epoch": 0.49452165600411796,
+      "grad_norm": 0.4489741921424866,
+      "learning_rate": 0.00010665170815412884,
+      "loss": 0.3487,
+      "step": 6725
+    },
+    {
+      "epoch": 0.49459519082285464,
+      "grad_norm": 0.6738259792327881,
+      "learning_rate": 0.00010662794452907811,
+      "loss": 0.4263,
+      "step": 6726
+    },
+    {
+      "epoch": 0.4946687256415913,
+      "grad_norm": 0.8963955640792847,
+      "learning_rate": 0.00010660418052808343,
+      "loss": 0.3953,
+      "step": 6727
+    },
+    {
+      "epoch": 0.494742260460328,
+      "grad_norm": 1.1698473691940308,
+      "learning_rate": 0.00010658041615249263,
+      "loss": 0.3704,
+      "step": 6728
+    },
+    {
+      "epoch": 0.49481579527906466,
+      "grad_norm": 0.7447865605354309,
+      "learning_rate": 0.0001065566514036537,
+      "loss": 0.3716,
+      "step": 6729
+    },
+    {
+      "epoch": 0.4948893300978013,
+      "grad_norm": 0.5361195206642151,
+      "learning_rate": 0.00010653288628291457,
+      "loss": 0.3741,
+      "step": 6730
+    },
+    {
+      "epoch": 0.49496286491653796,
+      "grad_norm": 0.8817668557167053,
+      "learning_rate": 0.00010650912079162322,
+      "loss": 0.3687,
+      "step": 6731
+    },
+    {
+      "epoch": 0.49503639973527463,
+      "grad_norm": 0.6625123620033264,
+      "learning_rate": 0.0001064853549311277,
+      "loss": 0.3801,
+      "step": 6732
+    },
+    {
+      "epoch": 0.4951099345540113,
+      "grad_norm": 0.6314841508865356,
+      "learning_rate": 0.000106461588702776,
+      "loss": 0.4013,
+      "step": 6733
+    },
+    {
+      "epoch": 0.495183469372748,
+      "grad_norm": 0.7629512548446655,
+      "learning_rate": 0.00010643782210791618,
+      "loss": 0.3933,
+      "step": 6734
+    },
+    {
+      "epoch": 0.49525700419148466,
+      "grad_norm": 0.4683447778224945,
+      "learning_rate": 0.00010641405514789629,
+      "loss": 0.3798,
+      "step": 6735
+    },
+    {
+      "epoch": 0.49533053901022134,
+      "grad_norm": 0.6614900231361389,
+      "learning_rate": 0.00010639028782406438,
+      "loss": 0.3803,
+      "step": 6736
+    },
+    {
+      "epoch": 0.495404073828958,
+      "grad_norm": 0.7423810958862305,
+      "learning_rate": 0.00010636652013776865,
+      "loss": 0.3885,
+      "step": 6737
+    },
+    {
+      "epoch": 0.4954776086476947,
+      "grad_norm": 0.4960765242576599,
+      "learning_rate": 0.00010634275209035716,
+      "loss": 0.3708,
+      "step": 6738
+    },
+    {
+      "epoch": 0.49555114346643137,
+      "grad_norm": 0.6691722869873047,
+      "learning_rate": 0.00010631898368317808,
+      "loss": 0.3809,
+      "step": 6739
+    },
+    {
+      "epoch": 0.49562467828516804,
+      "grad_norm": 1.0515515804290771,
+      "learning_rate": 0.00010629521491757957,
+      "loss": 0.3752,
+      "step": 6740
+    },
+    {
+      "epoch": 0.4956982131039047,
+      "grad_norm": 0.6753626465797424,
+      "learning_rate": 0.00010627144579490982,
+      "loss": 0.396,
+      "step": 6741
+    },
+    {
+      "epoch": 0.4957717479226414,
+      "grad_norm": 0.8408822417259216,
+      "learning_rate": 0.00010624767631651706,
+      "loss": 0.385,
+      "step": 6742
+    },
+    {
+      "epoch": 0.49584528274137807,
+      "grad_norm": 1.1617590188980103,
+      "learning_rate": 0.00010622390648374952,
+      "loss": 0.3825,
+      "step": 6743
+    },
+    {
+      "epoch": 0.4959188175601147,
+      "grad_norm": 0.4823738634586334,
+      "learning_rate": 0.0001062001362979554,
+      "loss": 0.3727,
+      "step": 6744
+    },
+    {
+      "epoch": 0.49599235237885136,
+      "grad_norm": 1.1973392963409424,
+      "learning_rate": 0.00010617636576048304,
+      "loss": 0.3725,
+      "step": 6745
+    },
+    {
+      "epoch": 0.49606588719758804,
+      "grad_norm": 1.0553213357925415,
+      "learning_rate": 0.00010615259487268068,
+      "loss": 0.3848,
+      "step": 6746
+    },
+    {
+      "epoch": 0.4961394220163247,
+      "grad_norm": 0.4804255962371826,
+      "learning_rate": 0.00010612882363589665,
+      "loss": 0.3787,
+      "step": 6747
+    },
+    {
+      "epoch": 0.4962129568350614,
+      "grad_norm": 1.244854211807251,
+      "learning_rate": 0.00010610505205147927,
+      "loss": 0.3742,
+      "step": 6748
+    },
+    {
+      "epoch": 0.49628649165379807,
+      "grad_norm": 0.709155797958374,
+      "learning_rate": 0.00010608128012077686,
+      "loss": 0.3509,
+      "step": 6749
+    },
+    {
+      "epoch": 0.49636002647253474,
+      "grad_norm": 0.708418607711792,
+      "learning_rate": 0.00010605750784513788,
+      "loss": 0.39,
+      "step": 6750
+    },
+    {
+      "epoch": 0.4964335612912714,
+      "grad_norm": 1.0876315832138062,
+      "learning_rate": 0.00010603373522591065,
+      "loss": 0.3803,
+      "step": 6751
+    },
+    {
+      "epoch": 0.4965070961100081,
+      "grad_norm": 0.6191591620445251,
+      "learning_rate": 0.00010600996226444359,
+      "loss": 0.3953,
+      "step": 6752
+    },
+    {
+      "epoch": 0.49658063092874477,
+      "grad_norm": 0.814781665802002,
+      "learning_rate": 0.00010598618896208513,
+      "loss": 0.3735,
+      "step": 6753
+    },
+    {
+      "epoch": 0.49665416574748145,
+      "grad_norm": 1.2113147974014282,
+      "learning_rate": 0.00010596241532018367,
+      "loss": 0.4017,
+      "step": 6754
+    },
+    {
+      "epoch": 0.4967277005662181,
+      "grad_norm": 0.7873516082763672,
+      "learning_rate": 0.00010593864134008777,
+      "loss": 0.3794,
+      "step": 6755
+    },
+    {
+      "epoch": 0.4968012353849548,
+      "grad_norm": 1.0288691520690918,
+      "learning_rate": 0.00010591486702314585,
+      "loss": 0.3821,
+      "step": 6756
+    },
+    {
+      "epoch": 0.4968747702036915,
+      "grad_norm": 1.5927454233169556,
+      "learning_rate": 0.00010589109237070644,
+      "loss": 0.3747,
+      "step": 6757
+    },
+    {
+      "epoch": 0.4969483050224281,
+      "grad_norm": 0.802470862865448,
+      "learning_rate": 0.00010586731738411806,
+      "loss": 0.4066,
+      "step": 6758
+    },
+    {
+      "epoch": 0.49702183984116477,
+      "grad_norm": 1.5106767416000366,
+      "learning_rate": 0.00010584354206472921,
+      "loss": 0.3645,
+      "step": 6759
+    },
+    {
+      "epoch": 0.49709537465990145,
+      "grad_norm": 1.0579848289489746,
+      "learning_rate": 0.00010581976641388851,
+      "loss": 0.4216,
+      "step": 6760
+    },
+    {
+      "epoch": 0.4971689094786381,
+      "grad_norm": 2.2994041442871094,
+      "learning_rate": 0.00010579599043294452,
+      "loss": 0.3727,
+      "step": 6761
+    },
+    {
+      "epoch": 0.4972424442973748,
+      "grad_norm": 2.1488449573516846,
+      "learning_rate": 0.00010577221412324583,
+      "loss": 0.3821,
+      "step": 6762
+    },
+    {
+      "epoch": 0.4973159791161115,
+      "grad_norm": 1.4459303617477417,
+      "learning_rate": 0.0001057484374861411,
+      "loss": 0.3908,
+      "step": 6763
+    },
+    {
+      "epoch": 0.49738951393484815,
+      "grad_norm": 1.44571852684021,
+      "learning_rate": 0.00010572466052297888,
+      "loss": 0.3719,
+      "step": 6764
+    },
+    {
+      "epoch": 0.4974630487535848,
+      "grad_norm": 1.4444057941436768,
+      "learning_rate": 0.00010570088323510788,
+      "loss": 0.3759,
+      "step": 6765
+    },
+    {
+      "epoch": 0.4975365835723215,
+      "grad_norm": 1.1461331844329834,
+      "learning_rate": 0.00010567710562387678,
+      "loss": 0.3959,
+      "step": 6766
+    },
+    {
+      "epoch": 0.4976101183910582,
+      "grad_norm": 1.3379753828048706,
+      "learning_rate": 0.00010565332769063423,
+      "loss": 0.3857,
+      "step": 6767
+    },
+    {
+      "epoch": 0.49768365320979485,
+      "grad_norm": 1.0091559886932373,
+      "learning_rate": 0.00010562954943672903,
+      "loss": 0.3748,
+      "step": 6768
+    },
+    {
+      "epoch": 0.4977571880285315,
+      "grad_norm": 1.4301483631134033,
+      "learning_rate": 0.00010560577086350977,
+      "loss": 0.3803,
+      "step": 6769
+    },
+    {
+      "epoch": 0.4978307228472682,
+      "grad_norm": 1.0632727146148682,
+      "learning_rate": 0.00010558199197232532,
+      "loss": 0.4025,
+      "step": 6770
+    },
+    {
+      "epoch": 0.4979042576660049,
+      "grad_norm": 0.6718907356262207,
+      "learning_rate": 0.00010555821276452438,
+      "loss": 0.3671,
+      "step": 6771
+    },
+    {
+      "epoch": 0.4979777924847415,
+      "grad_norm": 1.3090674877166748,
+      "learning_rate": 0.00010553443324145573,
+      "loss": 0.4068,
+      "step": 6772
+    },
+    {
+      "epoch": 0.4980513273034782,
+      "grad_norm": 0.9016309380531311,
+      "learning_rate": 0.0001055106534044682,
+      "loss": 0.4177,
+      "step": 6773
+    },
+    {
+      "epoch": 0.49812486212221485,
+      "grad_norm": 0.6252167224884033,
+      "learning_rate": 0.0001054868732549106,
+      "loss": 0.4147,
+      "step": 6774
+    },
+    {
+      "epoch": 0.4981983969409515,
+      "grad_norm": 0.7380488514900208,
+      "learning_rate": 0.00010546309279413176,
+      "loss": 0.3846,
+      "step": 6775
+    },
+    {
+      "epoch": 0.4982719317596882,
+      "grad_norm": 0.6119115352630615,
+      "learning_rate": 0.00010543931202348052,
+      "loss": 0.4285,
+      "step": 6776
+    },
+    {
+      "epoch": 0.4983454665784249,
+      "grad_norm": 0.6345200538635254,
+      "learning_rate": 0.00010541553094430577,
+      "loss": 0.4077,
+      "step": 6777
+    },
+    {
+      "epoch": 0.49841900139716155,
+      "grad_norm": 0.6942078471183777,
+      "learning_rate": 0.00010539174955795639,
+      "loss": 0.4126,
+      "step": 6778
+    },
+    {
+      "epoch": 0.49849253621589823,
+      "grad_norm": 0.6188831329345703,
+      "learning_rate": 0.00010536796786578126,
+      "loss": 0.3972,
+      "step": 6779
+    },
+    {
+      "epoch": 0.4985660710346349,
+      "grad_norm": 0.9673404097557068,
+      "learning_rate": 0.00010534418586912937,
+      "loss": 0.3902,
+      "step": 6780
+    },
+    {
+      "epoch": 0.4986396058533716,
+      "grad_norm": 0.7563518285751343,
+      "learning_rate": 0.0001053204035693496,
+      "loss": 0.3995,
+      "step": 6781
+    },
+    {
+      "epoch": 0.49871314067210826,
+      "grad_norm": 0.5526779890060425,
+      "learning_rate": 0.00010529662096779089,
+      "loss": 0.3749,
+      "step": 6782
+    },
+    {
+      "epoch": 0.49878667549084493,
+      "grad_norm": 0.5473906397819519,
+      "learning_rate": 0.00010527283806580231,
+      "loss": 0.3738,
+      "step": 6783
+    },
+    {
+      "epoch": 0.4988602103095816,
+      "grad_norm": 0.5468201637268066,
+      "learning_rate": 0.00010524905486473276,
+      "loss": 0.4093,
+      "step": 6784
+    },
+    {
+      "epoch": 0.4989337451283183,
+      "grad_norm": 0.5936571359634399,
+      "learning_rate": 0.00010522527136593125,
+      "loss": 0.3838,
+      "step": 6785
+    },
+    {
+      "epoch": 0.4990072799470549,
+      "grad_norm": 0.8434199094772339,
+      "learning_rate": 0.00010520148757074688,
+      "loss": 0.4089,
+      "step": 6786
+    },
+    {
+      "epoch": 0.4990808147657916,
+      "grad_norm": 0.6540611386299133,
+      "learning_rate": 0.00010517770348052862,
+      "loss": 0.4093,
+      "step": 6787
+    },
+    {
+      "epoch": 0.49915434958452826,
+      "grad_norm": 0.6770199537277222,
+      "learning_rate": 0.00010515391909662558,
+      "loss": 0.4012,
+      "step": 6788
+    },
+    {
+      "epoch": 0.49922788440326493,
+      "grad_norm": 0.8257825374603271,
+      "learning_rate": 0.0001051301344203868,
+      "loss": 0.3743,
+      "step": 6789
+    },
+    {
+      "epoch": 0.4993014192220016,
+      "grad_norm": 0.7651835083961487,
+      "learning_rate": 0.00010510634945316133,
+      "loss": 0.3692,
+      "step": 6790
+    },
+    {
+      "epoch": 0.4993749540407383,
+      "grad_norm": 0.7893267273902893,
+      "learning_rate": 0.00010508256419629839,
+      "loss": 0.4001,
+      "step": 6791
+    },
+    {
+      "epoch": 0.49944848885947496,
+      "grad_norm": 0.5845324993133545,
+      "learning_rate": 0.00010505877865114704,
+      "loss": 0.3737,
+      "step": 6792
+    },
+    {
+      "epoch": 0.49952202367821164,
+      "grad_norm": 1.1270127296447754,
+      "learning_rate": 0.00010503499281905642,
+      "loss": 0.3875,
+      "step": 6793
+    },
+    {
+      "epoch": 0.4995955584969483,
+      "grad_norm": 1.1187059879302979,
+      "learning_rate": 0.00010501120670137568,
+      "loss": 0.3851,
+      "step": 6794
+    },
+    {
+      "epoch": 0.499669093315685,
+      "grad_norm": 1.3758118152618408,
+      "learning_rate": 0.000104987420299454,
+      "loss": 0.4061,
+      "step": 6795
+    },
+    {
+      "epoch": 0.49974262813442166,
+      "grad_norm": 0.6574012041091919,
+      "learning_rate": 0.0001049636336146406,
+      "loss": 0.3922,
+      "step": 6796
+    },
+    {
+      "epoch": 0.49981616295315834,
+      "grad_norm": 1.2343254089355469,
+      "learning_rate": 0.00010493984664828466,
+      "loss": 0.3986,
+      "step": 6797
+    },
+    {
+      "epoch": 0.499889697771895,
+      "grad_norm": 0.6263077259063721,
+      "learning_rate": 0.00010491605940173541,
+      "loss": 0.3992,
+      "step": 6798
+    },
+    {
+      "epoch": 0.4999632325906317,
+      "grad_norm": 0.6469645500183105,
+      "learning_rate": 0.00010489227187634208,
+      "loss": 0.3902,
+      "step": 6799
+    },
+    {
+      "epoch": 0.5000367674093683,
+      "grad_norm": 0.6480942368507385,
+      "learning_rate": 0.00010486848407345391,
+      "loss": 0.4071,
+      "step": 6800
+    },
+    {
+      "epoch": 0.500110302228105,
+      "grad_norm": 0.5455988049507141,
+      "learning_rate": 0.00010484469599442019,
+      "loss": 0.3921,
+      "step": 6801
+    },
+    {
+      "epoch": 0.5001838370468417,
+      "grad_norm": 0.5757057666778564,
+      "learning_rate": 0.0001048209076405902,
+      "loss": 0.3869,
+      "step": 6802
+    },
+    {
+      "epoch": 0.5002573718655784,
+      "grad_norm": 0.5951406359672546,
+      "learning_rate": 0.00010479711901331326,
+      "loss": 0.4162,
+      "step": 6803
+    },
+    {
+      "epoch": 0.500330906684315,
+      "grad_norm": 0.77271968126297,
+      "learning_rate": 0.00010477333011393866,
+      "loss": 0.4089,
+      "step": 6804
+    },
+    {
+      "epoch": 0.5004044415030517,
+      "grad_norm": 0.7404239177703857,
+      "learning_rate": 0.00010474954094381571,
+      "loss": 0.3948,
+      "step": 6805
+    },
+    {
+      "epoch": 0.5004779763217884,
+      "grad_norm": 0.6844283938407898,
+      "learning_rate": 0.00010472575150429382,
+      "loss": 0.3916,
+      "step": 6806
+    },
+    {
+      "epoch": 0.500551511140525,
+      "grad_norm": 0.51955246925354,
+      "learning_rate": 0.00010470196179672229,
+      "loss": 0.3739,
+      "step": 6807
+    },
+    {
+      "epoch": 0.5006250459592617,
+      "grad_norm": 0.5101710557937622,
+      "learning_rate": 0.00010467817182245054,
+      "loss": 0.3663,
+      "step": 6808
+    },
+    {
+      "epoch": 0.5006985807779983,
+      "grad_norm": 0.6816007494926453,
+      "learning_rate": 0.00010465438158282796,
+      "loss": 0.3873,
+      "step": 6809
+    },
+    {
+      "epoch": 0.5007721155967351,
+      "grad_norm": 0.7284507155418396,
+      "learning_rate": 0.00010463059107920393,
+      "loss": 0.4108,
+      "step": 6810
+    },
+    {
+      "epoch": 0.5008456504154717,
+      "grad_norm": 0.710649311542511,
+      "learning_rate": 0.00010460680031292789,
+      "loss": 0.3636,
+      "step": 6811
+    },
+    {
+      "epoch": 0.5009191852342084,
+      "grad_norm": 0.6319148540496826,
+      "learning_rate": 0.00010458300928534928,
+      "loss": 0.369,
+      "step": 6812
+    },
+    {
+      "epoch": 0.500992720052945,
+      "grad_norm": 0.49306583404541016,
+      "learning_rate": 0.00010455921799781751,
+      "loss": 0.379,
+      "step": 6813
+    },
+    {
+      "epoch": 0.5010662548716818,
+      "grad_norm": 0.7567586898803711,
+      "learning_rate": 0.00010453542645168213,
+      "loss": 0.3974,
+      "step": 6814
+    },
+    {
+      "epoch": 0.5011397896904184,
+      "grad_norm": 0.8013209700584412,
+      "learning_rate": 0.00010451163464829254,
+      "loss": 0.3722,
+      "step": 6815
+    },
+    {
+      "epoch": 0.5012133245091551,
+      "grad_norm": 0.49445003271102905,
+      "learning_rate": 0.0001044878425889983,
+      "loss": 0.3886,
+      "step": 6816
+    },
+    {
+      "epoch": 0.5012868593278917,
+      "grad_norm": 0.7907334566116333,
+      "learning_rate": 0.00010446405027514889,
+      "loss": 0.3817,
+      "step": 6817
+    },
+    {
+      "epoch": 0.5013603941466285,
+      "grad_norm": 0.833968997001648,
+      "learning_rate": 0.00010444025770809382,
+      "loss": 0.3783,
+      "step": 6818
+    },
+    {
+      "epoch": 0.5014339289653651,
+      "grad_norm": 0.6043686270713806,
+      "learning_rate": 0.00010441646488918266,
+      "loss": 0.3963,
+      "step": 6819
+    },
+    {
+      "epoch": 0.5015074637841018,
+      "grad_norm": 0.6316748261451721,
+      "learning_rate": 0.00010439267181976495,
+      "loss": 0.3798,
+      "step": 6820
+    },
+    {
+      "epoch": 0.5015809986028384,
+      "grad_norm": 1.1332154273986816,
+      "learning_rate": 0.00010436887850119026,
+      "loss": 0.3762,
+      "step": 6821
+    },
+    {
+      "epoch": 0.5016545334215751,
+      "grad_norm": 0.6798984408378601,
+      "learning_rate": 0.00010434508493480817,
+      "loss": 0.3664,
+      "step": 6822
+    },
+    {
+      "epoch": 0.5017280682403118,
+      "grad_norm": 0.7793279886245728,
+      "learning_rate": 0.00010432129112196824,
+      "loss": 0.3651,
+      "step": 6823
+    },
+    {
+      "epoch": 0.5018016030590484,
+      "grad_norm": 1.2859275341033936,
+      "learning_rate": 0.00010429749706402016,
+      "loss": 0.369,
+      "step": 6824
+    },
+    {
+      "epoch": 0.5018751378777851,
+      "grad_norm": 0.5334396958351135,
+      "learning_rate": 0.00010427370276231352,
+      "loss": 0.3823,
+      "step": 6825
+    },
+    {
+      "epoch": 0.5019486726965218,
+      "grad_norm": 0.936091423034668,
+      "learning_rate": 0.00010424990821819791,
+      "loss": 0.3642,
+      "step": 6826
+    },
+    {
+      "epoch": 0.5020222075152585,
+      "grad_norm": 0.8561679720878601,
+      "learning_rate": 0.00010422611343302307,
+      "loss": 0.3785,
+      "step": 6827
+    },
+    {
+      "epoch": 0.5020957423339951,
+      "grad_norm": 0.8245511054992676,
+      "learning_rate": 0.00010420231840813856,
+      "loss": 0.389,
+      "step": 6828
+    },
+    {
+      "epoch": 0.5021692771527319,
+      "grad_norm": 1.1256879568099976,
+      "learning_rate": 0.00010417852314489416,
+      "loss": 0.4003,
+      "step": 6829
+    },
+    {
+      "epoch": 0.5022428119714685,
+      "grad_norm": 0.755230724811554,
+      "learning_rate": 0.0001041547276446395,
+      "loss": 0.3989,
+      "step": 6830
+    },
+    {
+      "epoch": 0.5023163467902052,
+      "grad_norm": 0.8816762566566467,
+      "learning_rate": 0.00010413093190872427,
+      "loss": 0.3898,
+      "step": 6831
+    },
+    {
+      "epoch": 0.5023898816089418,
+      "grad_norm": 1.4258679151535034,
+      "learning_rate": 0.00010410713593849828,
+      "loss": 0.4059,
+      "step": 6832
+    },
+    {
+      "epoch": 0.5024634164276786,
+      "grad_norm": 0.5876091718673706,
+      "learning_rate": 0.00010408333973531117,
+      "loss": 0.3775,
+      "step": 6833
+    },
+    {
+      "epoch": 0.5025369512464152,
+      "grad_norm": 1.4760808944702148,
+      "learning_rate": 0.00010405954330051275,
+      "loss": 0.388,
+      "step": 6834
+    },
+    {
+      "epoch": 0.5026104860651518,
+      "grad_norm": 0.8379570245742798,
+      "learning_rate": 0.00010403574663545273,
+      "loss": 0.3971,
+      "step": 6835
+    },
+    {
+      "epoch": 0.5026840208838885,
+      "grad_norm": 0.48766905069351196,
+      "learning_rate": 0.00010401194974148087,
+      "loss": 0.3715,
+      "step": 6836
+    },
+    {
+      "epoch": 0.5027575557026251,
+      "grad_norm": 1.1833909749984741,
+      "learning_rate": 0.00010398815261994703,
+      "loss": 0.4038,
+      "step": 6837
+    },
+    {
+      "epoch": 0.5028310905213619,
+      "grad_norm": 1.6237077713012695,
+      "learning_rate": 0.00010396435527220096,
+      "loss": 0.3952,
+      "step": 6838
+    },
+    {
+      "epoch": 0.5029046253400985,
+      "grad_norm": 1.4716830253601074,
+      "learning_rate": 0.00010394055769959248,
+      "loss": 0.415,
+      "step": 6839
+    },
+    {
+      "epoch": 0.5029781601588352,
+      "grad_norm": 2.293944835662842,
+      "learning_rate": 0.0001039167599034714,
+      "loss": 0.3841,
+      "step": 6840
+    },
+    {
+      "epoch": 0.5030516949775719,
+      "grad_norm": 3.113219738006592,
+      "learning_rate": 0.00010389296188518756,
+      "loss": 0.3858,
+      "step": 6841
+    },
+    {
+      "epoch": 0.5031252297963086,
+      "grad_norm": 0.895716667175293,
+      "learning_rate": 0.00010386916364609081,
+      "loss": 0.4027,
+      "step": 6842
+    },
+    {
+      "epoch": 0.5031987646150452,
+      "grad_norm": 1.4653480052947998,
+      "learning_rate": 0.00010384536518753103,
+      "loss": 0.3639,
+      "step": 6843
+    },
+    {
+      "epoch": 0.5032722994337819,
+      "grad_norm": 3.0138964653015137,
+      "learning_rate": 0.00010382156651085808,
+      "loss": 0.439,
+      "step": 6844
+    },
+    {
+      "epoch": 0.5033458342525186,
+      "grad_norm": 1.2317482233047485,
+      "learning_rate": 0.00010379776761742184,
+      "loss": 0.3812,
+      "step": 6845
+    },
+    {
+      "epoch": 0.5034193690712553,
+      "grad_norm": 1.7957843542099,
+      "learning_rate": 0.00010377396850857217,
+      "loss": 0.388,
+      "step": 6846
+    },
+    {
+      "epoch": 0.5034929038899919,
+      "grad_norm": 1.3577728271484375,
+      "learning_rate": 0.00010375016918565907,
+      "loss": 0.4236,
+      "step": 6847
+    },
+    {
+      "epoch": 0.5035664387087286,
+      "grad_norm": 1.5936510562896729,
+      "learning_rate": 0.00010372636965003242,
+      "loss": 0.4108,
+      "step": 6848
+    },
+    {
+      "epoch": 0.5036399735274653,
+      "grad_norm": 1.0904711484909058,
+      "learning_rate": 0.00010370256990304211,
+      "loss": 0.4385,
+      "step": 6849
+    },
+    {
+      "epoch": 0.5037135083462019,
+      "grad_norm": 2.855497360229492,
+      "learning_rate": 0.00010367876994603818,
+      "loss": 0.3811,
+      "step": 6850
+    },
+    {
+      "epoch": 0.5037870431649386,
+      "grad_norm": 2.2801060676574707,
+      "learning_rate": 0.0001036549697803705,
+      "loss": 0.3989,
+      "step": 6851
+    },
+    {
+      "epoch": 0.5038605779836752,
+      "grad_norm": 1.598266839981079,
+      "learning_rate": 0.00010363116940738909,
+      "loss": 0.4097,
+      "step": 6852
+    },
+    {
+      "epoch": 0.503934112802412,
+      "grad_norm": 2.0304811000823975,
+      "learning_rate": 0.00010360736882844393,
+      "loss": 0.4035,
+      "step": 6853
+    },
+    {
+      "epoch": 0.5040076476211486,
+      "grad_norm": 2.0821638107299805,
+      "learning_rate": 0.00010358356804488496,
+      "loss": 0.3942,
+      "step": 6854
+    },
+    {
+      "epoch": 0.5040811824398853,
+      "grad_norm": 1.3787494897842407,
+      "learning_rate": 0.00010355976705806228,
+      "loss": 0.4085,
+      "step": 6855
+    },
+    {
+      "epoch": 0.5041547172586219,
+      "grad_norm": 1.485762596130371,
+      "learning_rate": 0.00010353596586932583,
+      "loss": 0.3945,
+      "step": 6856
+    },
+    {
+      "epoch": 0.5042282520773587,
+      "grad_norm": 1.552289366722107,
+      "learning_rate": 0.00010351216448002568,
+      "loss": 0.4249,
+      "step": 6857
+    },
+    {
+      "epoch": 0.5043017868960953,
+      "grad_norm": 1.2129734754562378,
+      "learning_rate": 0.00010348836289151183,
+      "loss": 0.3954,
+      "step": 6858
+    },
+    {
+      "epoch": 0.504375321714832,
+      "grad_norm": 1.1258928775787354,
+      "learning_rate": 0.00010346456110513436,
+      "loss": 0.3939,
+      "step": 6859
+    },
+    {
+      "epoch": 0.5044488565335686,
+      "grad_norm": 0.7881874442100525,
+      "learning_rate": 0.00010344075912224334,
+      "loss": 0.4056,
+      "step": 6860
+    },
+    {
+      "epoch": 0.5045223913523054,
+      "grad_norm": 1.296977162361145,
+      "learning_rate": 0.00010341695694418882,
+      "loss": 0.4087,
+      "step": 6861
+    },
+    {
+      "epoch": 0.504595926171042,
+      "grad_norm": 0.8304060101509094,
+      "learning_rate": 0.00010339315457232093,
+      "loss": 0.3852,
+      "step": 6862
+    },
+    {
+      "epoch": 0.5046694609897786,
+      "grad_norm": 0.8590853810310364,
+      "learning_rate": 0.0001033693520079897,
+      "loss": 0.3773,
+      "step": 6863
+    },
+    {
+      "epoch": 0.5047429958085153,
+      "grad_norm": 0.9005637168884277,
+      "learning_rate": 0.00010334554925254525,
+      "loss": 0.4019,
+      "step": 6864
+    },
+    {
+      "epoch": 0.504816530627252,
+      "grad_norm": 1.020241618156433,
+      "learning_rate": 0.00010332174630733776,
+      "loss": 0.3609,
+      "step": 6865
+    },
+    {
+      "epoch": 0.5048900654459887,
+      "grad_norm": 0.7866693139076233,
+      "learning_rate": 0.00010329794317371728,
+      "loss": 0.3932,
+      "step": 6866
+    },
+    {
+      "epoch": 0.5049636002647253,
+      "grad_norm": 0.7302238941192627,
+      "learning_rate": 0.00010327413985303398,
+      "loss": 0.4384,
+      "step": 6867
+    },
+    {
+      "epoch": 0.505037135083462,
+      "grad_norm": 0.8089302182197571,
+      "learning_rate": 0.00010325033634663806,
+      "loss": 0.3794,
+      "step": 6868
+    },
+    {
+      "epoch": 0.5051106699021987,
+      "grad_norm": 0.7195939421653748,
+      "learning_rate": 0.00010322653265587959,
+      "loss": 0.4355,
+      "step": 6869
+    },
+    {
+      "epoch": 0.5051842047209354,
+      "grad_norm": 0.5651144981384277,
+      "learning_rate": 0.00010320272878210879,
+      "loss": 0.3595,
+      "step": 6870
+    },
+    {
+      "epoch": 0.505257739539672,
+      "grad_norm": 0.6186054348945618,
+      "learning_rate": 0.00010317892472667583,
+      "loss": 0.3869,
+      "step": 6871
+    },
+    {
+      "epoch": 0.5053312743584087,
+      "grad_norm": 0.5709956288337708,
+      "learning_rate": 0.0001031551204909309,
+      "loss": 0.3958,
+      "step": 6872
+    },
+    {
+      "epoch": 0.5054048091771454,
+      "grad_norm": 0.5329927206039429,
+      "learning_rate": 0.00010313131607622422,
+      "loss": 0.383,
+      "step": 6873
+    },
+    {
+      "epoch": 0.5054783439958821,
+      "grad_norm": 0.6535879969596863,
+      "learning_rate": 0.00010310751148390598,
+      "loss": 0.366,
+      "step": 6874
+    },
+    {
+      "epoch": 0.5055518788146187,
+      "grad_norm": 0.861283004283905,
+      "learning_rate": 0.00010308370671532644,
+      "loss": 0.4065,
+      "step": 6875
+    },
+    {
+      "epoch": 0.5056254136333554,
+      "grad_norm": 1.4684278964996338,
+      "learning_rate": 0.00010305990177183577,
+      "loss": 0.4256,
+      "step": 6876
+    },
+    {
+      "epoch": 0.5056989484520921,
+      "grad_norm": 0.644671618938446,
+      "learning_rate": 0.00010303609665478421,
+      "loss": 0.3791,
+      "step": 6877
+    },
+    {
+      "epoch": 0.5057724832708287,
+      "grad_norm": 1.3320558071136475,
+      "learning_rate": 0.0001030122913655221,
+      "loss": 0.386,
+      "step": 6878
+    },
+    {
+      "epoch": 0.5058460180895654,
+      "grad_norm": 0.8161589503288269,
+      "learning_rate": 0.00010298848590539963,
+      "loss": 0.3691,
+      "step": 6879
+    },
+    {
+      "epoch": 0.505919552908302,
+      "grad_norm": 1.030167818069458,
+      "learning_rate": 0.00010296468027576709,
+      "loss": 0.3764,
+      "step": 6880
+    },
+    {
+      "epoch": 0.5059930877270388,
+      "grad_norm": 1.1473650932312012,
+      "learning_rate": 0.00010294087447797473,
+      "loss": 0.3945,
+      "step": 6881
+    },
+    {
+      "epoch": 0.5060666225457754,
+      "grad_norm": 0.7000412344932556,
+      "learning_rate": 0.00010291706851337288,
+      "loss": 0.3592,
+      "step": 6882
+    },
+    {
+      "epoch": 0.5061401573645121,
+      "grad_norm": 1.4980660676956177,
+      "learning_rate": 0.00010289326238331183,
+      "loss": 0.4036,
+      "step": 6883
+    },
+    {
+      "epoch": 0.5062136921832487,
+      "grad_norm": 0.6714105606079102,
+      "learning_rate": 0.0001028694560891419,
+      "loss": 0.4112,
+      "step": 6884
+    },
+    {
+      "epoch": 0.5062872270019855,
+      "grad_norm": 1.7472611665725708,
+      "learning_rate": 0.00010284564963221337,
+      "loss": 0.3789,
+      "step": 6885
+    },
+    {
+      "epoch": 0.5063607618207221,
+      "grad_norm": 0.6572700142860413,
+      "learning_rate": 0.00010282184301387661,
+      "loss": 0.3521,
+      "step": 6886
+    },
+    {
+      "epoch": 0.5064342966394588,
+      "grad_norm": 1.9108136892318726,
+      "learning_rate": 0.0001027980362354819,
+      "loss": 0.4088,
+      "step": 6887
+    },
+    {
+      "epoch": 0.5065078314581954,
+      "grad_norm": 1.0420942306518555,
+      "learning_rate": 0.00010277422929837963,
+      "loss": 0.3642,
+      "step": 6888
+    },
+    {
+      "epoch": 0.5065813662769322,
+      "grad_norm": 2.278050422668457,
+      "learning_rate": 0.00010275042220392017,
+      "loss": 0.377,
+      "step": 6889
+    },
+    {
+      "epoch": 0.5066549010956688,
+      "grad_norm": 2.175820827484131,
+      "learning_rate": 0.00010272661495345383,
+      "loss": 0.375,
+      "step": 6890
+    },
+    {
+      "epoch": 0.5067284359144054,
+      "grad_norm": 1.0190900564193726,
+      "learning_rate": 0.00010270280754833106,
+      "loss": 0.3769,
+      "step": 6891
+    },
+    {
+      "epoch": 0.5068019707331421,
+      "grad_norm": 1.1128963232040405,
+      "learning_rate": 0.00010267899998990213,
+      "loss": 0.3854,
+      "step": 6892
+    },
+    {
+      "epoch": 0.5068755055518788,
+      "grad_norm": 1.2073783874511719,
+      "learning_rate": 0.00010265519227951754,
+      "loss": 0.395,
+      "step": 6893
+    },
+    {
+      "epoch": 0.5069490403706155,
+      "grad_norm": 0.618138313293457,
+      "learning_rate": 0.00010263138441852764,
+      "loss": 0.3628,
+      "step": 6894
+    },
+    {
+      "epoch": 0.5070225751893521,
+      "grad_norm": 1.2495464086532593,
+      "learning_rate": 0.00010260757640828283,
+      "loss": 0.3866,
+      "step": 6895
+    },
+    {
+      "epoch": 0.5070961100080889,
+      "grad_norm": 0.5454410910606384,
+      "learning_rate": 0.00010258376825013352,
+      "loss": 0.3681,
+      "step": 6896
+    },
+    {
+      "epoch": 0.5071696448268255,
+      "grad_norm": 1.448433518409729,
+      "learning_rate": 0.00010255995994543017,
+      "loss": 0.3859,
+      "step": 6897
+    },
+    {
+      "epoch": 0.5072431796455622,
+      "grad_norm": 0.525751531124115,
+      "learning_rate": 0.00010253615149552318,
+      "loss": 0.3576,
+      "step": 6898
+    },
+    {
+      "epoch": 0.5073167144642988,
+      "grad_norm": 1.3323363065719604,
+      "learning_rate": 0.00010251234290176301,
+      "loss": 0.3588,
+      "step": 6899
+    },
+    {
+      "epoch": 0.5073902492830356,
+      "grad_norm": 0.5495213866233826,
+      "learning_rate": 0.00010248853416550007,
+      "loss": 0.3875,
+      "step": 6900
+    },
+    {
+      "epoch": 0.5074637841017722,
+      "grad_norm": 1.4098907709121704,
+      "learning_rate": 0.00010246472528808487,
+      "loss": 0.3817,
+      "step": 6901
+    },
+    {
+      "epoch": 0.5075373189205089,
+      "grad_norm": 0.7353671789169312,
+      "learning_rate": 0.00010244091627086784,
+      "loss": 0.3819,
+      "step": 6902
+    },
+    {
+      "epoch": 0.5076108537392455,
+      "grad_norm": 0.8014718294143677,
+      "learning_rate": 0.0001024171071151995,
+      "loss": 0.3807,
+      "step": 6903
+    },
+    {
+      "epoch": 0.5076843885579823,
+      "grad_norm": 0.9635906219482422,
+      "learning_rate": 0.00010239329782243026,
+      "loss": 0.3701,
+      "step": 6904
+    },
+    {
+      "epoch": 0.5077579233767189,
+      "grad_norm": 0.5782277584075928,
+      "learning_rate": 0.00010236948839391062,
+      "loss": 0.365,
+      "step": 6905
+    },
+    {
+      "epoch": 0.5078314581954555,
+      "grad_norm": 1.0190390348434448,
+      "learning_rate": 0.00010234567883099113,
+      "loss": 0.3641,
+      "step": 6906
+    },
+    {
+      "epoch": 0.5079049930141922,
+      "grad_norm": 1.0076013803482056,
+      "learning_rate": 0.00010232186913502226,
+      "loss": 0.3939,
+      "step": 6907
+    },
+    {
+      "epoch": 0.5079785278329289,
+      "grad_norm": 0.7207874059677124,
+      "learning_rate": 0.0001022980593073545,
+      "loss": 0.3736,
+      "step": 6908
+    },
+    {
+      "epoch": 0.5080520626516656,
+      "grad_norm": 0.7707656621932983,
+      "learning_rate": 0.00010227424934933845,
+      "loss": 0.4151,
+      "step": 6909
+    },
+    {
+      "epoch": 0.5081255974704022,
+      "grad_norm": 0.6762974262237549,
+      "learning_rate": 0.00010225043926232449,
+      "loss": 0.3897,
+      "step": 6910
+    },
+    {
+      "epoch": 0.5081991322891389,
+      "grad_norm": 0.7238894701004028,
+      "learning_rate": 0.00010222662904766329,
+      "loss": 0.3577,
+      "step": 6911
+    },
+    {
+      "epoch": 0.5082726671078756,
+      "grad_norm": 0.5026848912239075,
+      "learning_rate": 0.00010220281870670536,
+      "loss": 0.3804,
+      "step": 6912
+    },
+    {
+      "epoch": 0.5083462019266123,
+      "grad_norm": 0.6862624287605286,
+      "learning_rate": 0.00010217900824080117,
+      "loss": 0.3493,
+      "step": 6913
+    },
+    {
+      "epoch": 0.5084197367453489,
+      "grad_norm": 0.5330853462219238,
+      "learning_rate": 0.0001021551976513014,
+      "loss": 0.4151,
+      "step": 6914
+    },
+    {
+      "epoch": 0.5084932715640856,
+      "grad_norm": 0.634003758430481,
+      "learning_rate": 0.00010213138693955649,
+      "loss": 0.3867,
+      "step": 6915
+    },
+    {
+      "epoch": 0.5085668063828223,
+      "grad_norm": 0.6365873217582703,
+      "learning_rate": 0.0001021075761069171,
+      "loss": 0.3872,
+      "step": 6916
+    },
+    {
+      "epoch": 0.508640341201559,
+      "grad_norm": 0.7798009514808655,
+      "learning_rate": 0.00010208376515473374,
+      "loss": 0.3966,
+      "step": 6917
+    },
+    {
+      "epoch": 0.5087138760202956,
+      "grad_norm": 0.6029713749885559,
+      "learning_rate": 0.00010205995408435702,
+      "loss": 0.3795,
+      "step": 6918
+    },
+    {
+      "epoch": 0.5087874108390322,
+      "grad_norm": 0.6239964365959167,
+      "learning_rate": 0.00010203614289713756,
+      "loss": 0.387,
+      "step": 6919
+    },
+    {
+      "epoch": 0.508860945657769,
+      "grad_norm": 0.858747124671936,
+      "learning_rate": 0.00010201233159442592,
+      "loss": 0.3742,
+      "step": 6920
+    },
+    {
+      "epoch": 0.5089344804765056,
+      "grad_norm": 0.974949061870575,
+      "learning_rate": 0.0001019885201775727,
+      "loss": 0.3688,
+      "step": 6921
+    },
+    {
+      "epoch": 0.5090080152952423,
+      "grad_norm": 0.800799548625946,
+      "learning_rate": 0.00010196470864792851,
+      "loss": 0.3743,
+      "step": 6922
+    },
+    {
+      "epoch": 0.5090815501139789,
+      "grad_norm": 0.5875540375709534,
+      "learning_rate": 0.00010194089700684397,
+      "loss": 0.3855,
+      "step": 6923
+    },
+    {
+      "epoch": 0.5091550849327157,
+      "grad_norm": 0.523476243019104,
+      "learning_rate": 0.00010191708525566972,
+      "loss": 0.4061,
+      "step": 6924
+    },
+    {
+      "epoch": 0.5092286197514523,
+      "grad_norm": 0.6478791236877441,
+      "learning_rate": 0.00010189327339575636,
+      "loss": 0.4059,
+      "step": 6925
+    },
+    {
+      "epoch": 0.509302154570189,
+      "grad_norm": 0.8200158476829529,
+      "learning_rate": 0.00010186946142845455,
+      "loss": 0.3671,
+      "step": 6926
+    },
+    {
+      "epoch": 0.5093756893889256,
+      "grad_norm": 0.9743598103523254,
+      "learning_rate": 0.00010184564935511491,
+      "loss": 0.3858,
+      "step": 6927
+    },
+    {
+      "epoch": 0.5094492242076624,
+      "grad_norm": 0.7946497201919556,
+      "learning_rate": 0.00010182183717708807,
+      "loss": 0.3725,
+      "step": 6928
+    },
+    {
+      "epoch": 0.509522759026399,
+      "grad_norm": 0.7004277110099792,
+      "learning_rate": 0.00010179802489572471,
+      "loss": 0.3851,
+      "step": 6929
+    },
+    {
+      "epoch": 0.5095962938451357,
+      "grad_norm": 0.6546820998191833,
+      "learning_rate": 0.00010177421251237549,
+      "loss": 0.3847,
+      "step": 6930
+    },
+    {
+      "epoch": 0.5096698286638723,
+      "grad_norm": 0.6432023644447327,
+      "learning_rate": 0.00010175040002839105,
+      "loss": 0.3671,
+      "step": 6931
+    },
+    {
+      "epoch": 0.5097433634826091,
+      "grad_norm": 0.4938010573387146,
+      "learning_rate": 0.00010172658744512211,
+      "loss": 0.3738,
+      "step": 6932
+    },
+    {
+      "epoch": 0.5098168983013457,
+      "grad_norm": 0.7180113792419434,
+      "learning_rate": 0.00010170277476391925,
+      "loss": 0.3872,
+      "step": 6933
+    },
+    {
+      "epoch": 0.5098904331200823,
+      "grad_norm": 0.7854035496711731,
+      "learning_rate": 0.00010167896198613325,
+      "loss": 0.3776,
+      "step": 6934
+    },
+    {
+      "epoch": 0.509963967938819,
+      "grad_norm": 0.7395696640014648,
+      "learning_rate": 0.00010165514911311476,
+      "loss": 0.3755,
+      "step": 6935
+    },
+    {
+      "epoch": 0.5100375027575557,
+      "grad_norm": 0.8163260817527771,
+      "learning_rate": 0.00010163133614621442,
+      "loss": 0.3979,
+      "step": 6936
+    },
+    {
+      "epoch": 0.5101110375762924,
+      "grad_norm": 0.7770224809646606,
+      "learning_rate": 0.00010160752308678301,
+      "loss": 0.3895,
+      "step": 6937
+    },
+    {
+      "epoch": 0.510184572395029,
+      "grad_norm": 0.6380110383033752,
+      "learning_rate": 0.00010158370993617119,
+      "loss": 0.3772,
+      "step": 6938
+    },
+    {
+      "epoch": 0.5102581072137657,
+      "grad_norm": 0.5224650502204895,
+      "learning_rate": 0.00010155989669572967,
+      "loss": 0.3868,
+      "step": 6939
+    },
+    {
+      "epoch": 0.5103316420325024,
+      "grad_norm": 1.052246332168579,
+      "learning_rate": 0.00010153608336680918,
+      "loss": 0.4106,
+      "step": 6940
+    },
+    {
+      "epoch": 0.5104051768512391,
+      "grad_norm": 0.7933809757232666,
+      "learning_rate": 0.00010151226995076038,
+      "loss": 0.3921,
+      "step": 6941
+    },
+    {
+      "epoch": 0.5104787116699757,
+      "grad_norm": 0.7982372641563416,
+      "learning_rate": 0.00010148845644893406,
+      "loss": 0.3828,
+      "step": 6942
+    },
+    {
+      "epoch": 0.5105522464887124,
+      "grad_norm": 0.531072735786438,
+      "learning_rate": 0.00010146464286268091,
+      "loss": 0.3636,
+      "step": 6943
+    },
+    {
+      "epoch": 0.5106257813074491,
+      "grad_norm": 0.8375490307807922,
+      "learning_rate": 0.00010144082919335167,
+      "loss": 0.3745,
+      "step": 6944
+    },
+    {
+      "epoch": 0.5106993161261858,
+      "grad_norm": 0.7307660579681396,
+      "learning_rate": 0.0001014170154422971,
+      "loss": 0.3723,
+      "step": 6945
+    },
+    {
+      "epoch": 0.5107728509449224,
+      "grad_norm": 0.6921009421348572,
+      "learning_rate": 0.00010139320161086788,
+      "loss": 0.3608,
+      "step": 6946
+    },
+    {
+      "epoch": 0.510846385763659,
+      "grad_norm": 0.5459523797035217,
+      "learning_rate": 0.00010136938770041482,
+      "loss": 0.3765,
+      "step": 6947
+    },
+    {
+      "epoch": 0.5109199205823958,
+      "grad_norm": 0.8584962487220764,
+      "learning_rate": 0.00010134557371228863,
+      "loss": 0.396,
+      "step": 6948
+    },
+    {
+      "epoch": 0.5109934554011324,
+      "grad_norm": 0.763245165348053,
+      "learning_rate": 0.00010132175964784008,
+      "loss": 0.3751,
+      "step": 6949
+    },
+    {
+      "epoch": 0.5110669902198691,
+      "grad_norm": 0.9482444524765015,
+      "learning_rate": 0.00010129794550841996,
+      "loss": 0.3667,
+      "step": 6950
+    },
+    {
+      "epoch": 0.5111405250386057,
+      "grad_norm": 1.0996134281158447,
+      "learning_rate": 0.00010127413129537895,
+      "loss": 0.3817,
+      "step": 6951
+    },
+    {
+      "epoch": 0.5112140598573425,
+      "grad_norm": 0.7675939798355103,
+      "learning_rate": 0.0001012503170100679,
+      "loss": 0.3739,
+      "step": 6952
+    },
+    {
+      "epoch": 0.5112875946760791,
+      "grad_norm": 1.4578994512557983,
+      "learning_rate": 0.00010122650265383755,
+      "loss": 0.4153,
+      "step": 6953
+    },
+    {
+      "epoch": 0.5113611294948158,
+      "grad_norm": 1.0244667530059814,
+      "learning_rate": 0.00010120268822803867,
+      "loss": 0.3723,
+      "step": 6954
+    },
+    {
+      "epoch": 0.5114346643135524,
+      "grad_norm": 1.4506229162216187,
+      "learning_rate": 0.00010117887373402206,
+      "loss": 0.3857,
+      "step": 6955
+    },
+    {
+      "epoch": 0.5115081991322892,
+      "grad_norm": 1.4079787731170654,
+      "learning_rate": 0.00010115505917313845,
+      "loss": 0.4014,
+      "step": 6956
+    },
+    {
+      "epoch": 0.5115817339510258,
+      "grad_norm": 1.1793792247772217,
+      "learning_rate": 0.0001011312445467387,
+      "loss": 0.376,
+      "step": 6957
+    },
+    {
+      "epoch": 0.5116552687697625,
+      "grad_norm": 1.172900676727295,
+      "learning_rate": 0.00010110742985617356,
+      "loss": 0.3619,
+      "step": 6958
+    },
+    {
+      "epoch": 0.5117288035884991,
+      "grad_norm": 0.4739299416542053,
+      "learning_rate": 0.00010108361510279381,
+      "loss": 0.393,
+      "step": 6959
+    },
+    {
+      "epoch": 0.5118023384072359,
+      "grad_norm": 0.9674410820007324,
+      "learning_rate": 0.00010105980028795028,
+      "loss": 0.3627,
+      "step": 6960
+    },
+    {
+      "epoch": 0.5118758732259725,
+      "grad_norm": 0.5612653493881226,
+      "learning_rate": 0.00010103598541299375,
+      "loss": 0.3833,
+      "step": 6961
+    },
+    {
+      "epoch": 0.5119494080447091,
+      "grad_norm": 0.9114152193069458,
+      "learning_rate": 0.00010101217047927505,
+      "loss": 0.3792,
+      "step": 6962
+    },
+    {
+      "epoch": 0.5120229428634459,
+      "grad_norm": 0.8241981267929077,
+      "learning_rate": 0.00010098835548814499,
+      "loss": 0.3762,
+      "step": 6963
+    },
+    {
+      "epoch": 0.5120964776821825,
+      "grad_norm": 0.7392498254776001,
+      "learning_rate": 0.00010096454044095431,
+      "loss": 0.3841,
+      "step": 6964
+    },
+    {
+      "epoch": 0.5121700125009192,
+      "grad_norm": 0.7240743637084961,
+      "learning_rate": 0.0001009407253390539,
+      "loss": 0.3704,
+      "step": 6965
+    },
+    {
+      "epoch": 0.5122435473196558,
+      "grad_norm": 0.6881362199783325,
+      "learning_rate": 0.00010091691018379457,
+      "loss": 0.3742,
+      "step": 6966
+    },
+    {
+      "epoch": 0.5123170821383926,
+      "grad_norm": 0.67800372838974,
+      "learning_rate": 0.00010089309497652713,
+      "loss": 0.3561,
+      "step": 6967
+    },
+    {
+      "epoch": 0.5123906169571292,
+      "grad_norm": 0.5648748278617859,
+      "learning_rate": 0.00010086927971860238,
+      "loss": 0.3594,
+      "step": 6968
+    },
+    {
+      "epoch": 0.5124641517758659,
+      "grad_norm": 0.7764260172843933,
+      "learning_rate": 0.00010084546441137118,
+      "loss": 0.3831,
+      "step": 6969
+    },
+    {
+      "epoch": 0.5125376865946025,
+      "grad_norm": 0.46042701601982117,
+      "learning_rate": 0.00010082164905618434,
+      "loss": 0.3446,
+      "step": 6970
+    },
+    {
+      "epoch": 0.5126112214133393,
+      "grad_norm": 1.200769066810608,
+      "learning_rate": 0.0001007978336543927,
+      "loss": 0.4014,
+      "step": 6971
+    },
+    {
+      "epoch": 0.5126847562320759,
+      "grad_norm": 0.8303195834159851,
+      "learning_rate": 0.00010077401820734708,
+      "loss": 0.3859,
+      "step": 6972
+    },
+    {
+      "epoch": 0.5127582910508126,
+      "grad_norm": 0.6092516183853149,
+      "learning_rate": 0.00010075020271639836,
+      "loss": 0.3625,
+      "step": 6973
+    },
+    {
+      "epoch": 0.5128318258695492,
+      "grad_norm": 0.5546369552612305,
+      "learning_rate": 0.00010072638718289733,
+      "loss": 0.3991,
+      "step": 6974
+    },
+    {
+      "epoch": 0.5129053606882858,
+      "grad_norm": 0.7419028878211975,
+      "learning_rate": 0.00010070257160819487,
+      "loss": 0.3688,
+      "step": 6975
+    },
+    {
+      "epoch": 0.5129788955070226,
+      "grad_norm": 0.7076804041862488,
+      "learning_rate": 0.00010067875599364181,
+      "loss": 0.3381,
+      "step": 6976
+    },
+    {
+      "epoch": 0.5130524303257592,
+      "grad_norm": 0.5212616324424744,
+      "learning_rate": 0.00010065494034058895,
+      "loss": 0.3919,
+      "step": 6977
+    },
+    {
+      "epoch": 0.5131259651444959,
+      "grad_norm": 0.739172101020813,
+      "learning_rate": 0.00010063112465038723,
+      "loss": 0.3675,
+      "step": 6978
+    },
+    {
+      "epoch": 0.5131994999632326,
+      "grad_norm": 0.8033172488212585,
+      "learning_rate": 0.00010060730892438743,
+      "loss": 0.3649,
+      "step": 6979
+    },
+    {
+      "epoch": 0.5132730347819693,
+      "grad_norm": 0.6724511981010437,
+      "learning_rate": 0.00010058349316394043,
+      "loss": 0.3701,
+      "step": 6980
+    },
+    {
+      "epoch": 0.5133465696007059,
+      "grad_norm": 0.5544342994689941,
+      "learning_rate": 0.00010055967737039709,
+      "loss": 0.3711,
+      "step": 6981
+    },
+    {
+      "epoch": 0.5134201044194426,
+      "grad_norm": 0.7689630389213562,
+      "learning_rate": 0.00010053586154510823,
+      "loss": 0.3754,
+      "step": 6982
+    },
+    {
+      "epoch": 0.5134936392381793,
+      "grad_norm": 0.6339815855026245,
+      "learning_rate": 0.00010051204568942477,
+      "loss": 0.3613,
+      "step": 6983
+    },
+    {
+      "epoch": 0.513567174056916,
+      "grad_norm": 0.7195934653282166,
+      "learning_rate": 0.00010048822980469755,
+      "loss": 0.3925,
+      "step": 6984
+    },
+    {
+      "epoch": 0.5136407088756526,
+      "grad_norm": 1.1619948148727417,
+      "learning_rate": 0.00010046441389227739,
+      "loss": 0.3869,
+      "step": 6985
+    },
+    {
+      "epoch": 0.5137142436943893,
+      "grad_norm": 0.8577642440795898,
+      "learning_rate": 0.00010044059795351519,
+      "loss": 0.3715,
+      "step": 6986
+    },
+    {
+      "epoch": 0.513787778513126,
+      "grad_norm": 0.47854748368263245,
+      "learning_rate": 0.00010041678198976179,
+      "loss": 0.3787,
+      "step": 6987
+    },
+    {
+      "epoch": 0.5138613133318627,
+      "grad_norm": 0.9683218598365784,
+      "learning_rate": 0.00010039296600236811,
+      "loss": 0.3657,
+      "step": 6988
+    },
+    {
+      "epoch": 0.5139348481505993,
+      "grad_norm": 1.1027969121932983,
+      "learning_rate": 0.00010036914999268498,
+      "loss": 0.364,
+      "step": 6989
+    },
+    {
+      "epoch": 0.5140083829693359,
+      "grad_norm": 0.8893772959709167,
+      "learning_rate": 0.00010034533396206327,
+      "loss": 0.3832,
+      "step": 6990
+    },
+    {
+      "epoch": 0.5140819177880727,
+      "grad_norm": 0.8725157380104065,
+      "learning_rate": 0.00010032151791185384,
+      "loss": 0.3949,
+      "step": 6991
+    },
+    {
+      "epoch": 0.5141554526068093,
+      "grad_norm": 0.6134024858474731,
+      "learning_rate": 0.00010029770184340756,
+      "loss": 0.3621,
+      "step": 6992
+    },
+    {
+      "epoch": 0.514228987425546,
+      "grad_norm": 0.7796329259872437,
+      "learning_rate": 0.00010027388575807535,
+      "loss": 0.4039,
+      "step": 6993
+    },
+    {
+      "epoch": 0.5143025222442826,
+      "grad_norm": 0.6064843535423279,
+      "learning_rate": 0.00010025006965720805,
+      "loss": 0.3865,
+      "step": 6994
+    },
+    {
+      "epoch": 0.5143760570630194,
+      "grad_norm": 0.5042343139648438,
+      "learning_rate": 0.00010022625354215651,
+      "loss": 0.3571,
+      "step": 6995
+    },
+    {
+      "epoch": 0.514449591881756,
+      "grad_norm": 0.5268373489379883,
+      "learning_rate": 0.00010020243741427166,
+      "loss": 0.3734,
+      "step": 6996
+    },
+    {
+      "epoch": 0.5145231267004927,
+      "grad_norm": 0.5343688130378723,
+      "learning_rate": 0.00010017862127490431,
+      "loss": 0.3521,
+      "step": 6997
+    },
+    {
+      "epoch": 0.5145966615192293,
+      "grad_norm": 0.5521165728569031,
+      "learning_rate": 0.00010015480512540541,
+      "loss": 0.3932,
+      "step": 6998
+    },
+    {
+      "epoch": 0.5146701963379661,
+      "grad_norm": 0.5797030925750732,
+      "learning_rate": 0.00010013098896712578,
+      "loss": 0.3545,
+      "step": 6999
+    },
+    {
+      "epoch": 0.5147437311567027,
+      "grad_norm": 0.5008895397186279,
+      "learning_rate": 0.0001001071728014163,
+      "loss": 0.3673,
+      "step": 7000
+    },
+    {
+      "epoch": 0.5148172659754394,
+      "grad_norm": 0.7673706412315369,
+      "learning_rate": 0.00010008335662962791,
+      "loss": 0.3636,
+      "step": 7001
+    },
+    {
+      "epoch": 0.514890800794176,
+      "grad_norm": 0.43098264932632446,
+      "learning_rate": 0.00010005954045311143,
+      "loss": 0.3635,
+      "step": 7002
+    },
+    {
+      "epoch": 0.5149643356129127,
+      "grad_norm": 0.8605313897132874,
+      "learning_rate": 0.00010003572427321775,
+      "loss": 0.3763,
+      "step": 7003
+    },
+    {
+      "epoch": 0.5150378704316494,
+      "grad_norm": 0.7491247057914734,
+      "learning_rate": 0.00010001190809129774,
+      "loss": 0.3717,
+      "step": 7004
+    },
+    {
+      "epoch": 0.515111405250386,
+      "grad_norm": 0.800662100315094,
+      "learning_rate": 9.99880919087023e-05,
+      "loss": 0.384,
+      "step": 7005
+    },
+    {
+      "epoch": 0.5151849400691227,
+      "grad_norm": 0.7647120952606201,
+      "learning_rate": 9.996427572678232e-05,
+      "loss": 0.3801,
+      "step": 7006
+    },
+    {
+      "epoch": 0.5152584748878594,
+      "grad_norm": 0.6799689531326294,
+      "learning_rate": 9.99404595468886e-05,
+      "loss": 0.4115,
+      "step": 7007
+    },
+    {
+      "epoch": 0.5153320097065961,
+      "grad_norm": 0.711214542388916,
+      "learning_rate": 9.991664337037212e-05,
+      "loss": 0.3756,
+      "step": 7008
+    },
+    {
+      "epoch": 0.5154055445253327,
+      "grad_norm": 0.6877127289772034,
+      "learning_rate": 9.98928271985837e-05,
+      "loss": 0.3489,
+      "step": 7009
+    },
+    {
+      "epoch": 0.5154790793440694,
+      "grad_norm": 0.6801254749298096,
+      "learning_rate": 9.986901103287423e-05,
+      "loss": 0.4021,
+      "step": 7010
+    },
+    {
+      "epoch": 0.5155526141628061,
+      "grad_norm": 0.6584367156028748,
+      "learning_rate": 9.984519487459461e-05,
+      "loss": 0.3753,
+      "step": 7011
+    },
+    {
+      "epoch": 0.5156261489815428,
+      "grad_norm": 0.8196418285369873,
+      "learning_rate": 9.982137872509571e-05,
+      "loss": 0.3583,
+      "step": 7012
+    },
+    {
+      "epoch": 0.5156996838002794,
+      "grad_norm": 0.8891412019729614,
+      "learning_rate": 9.979756258572835e-05,
+      "loss": 0.3895,
+      "step": 7013
+    },
+    {
+      "epoch": 0.5157732186190161,
+      "grad_norm": 1.650316596031189,
+      "learning_rate": 9.97737464578435e-05,
+      "loss": 0.3918,
+      "step": 7014
+    },
+    {
+      "epoch": 0.5158467534377528,
+      "grad_norm": 0.825536847114563,
+      "learning_rate": 9.974993034279198e-05,
+      "loss": 0.3959,
+      "step": 7015
+    },
+    {
+      "epoch": 0.5159202882564895,
+      "grad_norm": 1.8770288228988647,
+      "learning_rate": 9.972611424192465e-05,
+      "loss": 0.3679,
+      "step": 7016
+    },
+    {
+      "epoch": 0.5159938230752261,
+      "grad_norm": 1.1104389429092407,
+      "learning_rate": 9.970229815659246e-05,
+      "loss": 0.3628,
+      "step": 7017
+    },
+    {
+      "epoch": 0.5160673578939627,
+      "grad_norm": 1.8307615518569946,
+      "learning_rate": 9.96784820881462e-05,
+      "loss": 0.3719,
+      "step": 7018
+    },
+    {
+      "epoch": 0.5161408927126995,
+      "grad_norm": 1.5720877647399902,
+      "learning_rate": 9.965466603793679e-05,
+      "loss": 0.3636,
+      "step": 7019
+    },
+    {
+      "epoch": 0.5162144275314361,
+      "grad_norm": 1.7207627296447754,
+      "learning_rate": 9.963085000731504e-05,
+      "loss": 0.3672,
+      "step": 7020
+    },
+    {
+      "epoch": 0.5162879623501728,
+      "grad_norm": 1.2447718381881714,
+      "learning_rate": 9.960703399763191e-05,
+      "loss": 0.3768,
+      "step": 7021
+    },
+    {
+      "epoch": 0.5163614971689094,
+      "grad_norm": 1.5809507369995117,
+      "learning_rate": 9.958321801023822e-05,
+      "loss": 0.3677,
+      "step": 7022
+    },
+    {
+      "epoch": 0.5164350319876462,
+      "grad_norm": 1.1527754068374634,
+      "learning_rate": 9.955940204648482e-05,
+      "loss": 0.3724,
+      "step": 7023
+    },
+    {
+      "epoch": 0.5165085668063828,
+      "grad_norm": 1.8783315420150757,
+      "learning_rate": 9.953558610772263e-05,
+      "loss": 0.3814,
+      "step": 7024
+    },
+    {
+      "epoch": 0.5165821016251195,
+      "grad_norm": 1.6166900396347046,
+      "learning_rate": 9.951177019530249e-05,
+      "loss": 0.3445,
+      "step": 7025
+    },
+    {
+      "epoch": 0.5166556364438561,
+      "grad_norm": 1.2559159994125366,
+      "learning_rate": 9.948795431057523e-05,
+      "loss": 0.3389,
+      "step": 7026
+    },
+    {
+      "epoch": 0.5167291712625929,
+      "grad_norm": 1.1278574466705322,
+      "learning_rate": 9.946413845489177e-05,
+      "loss": 0.398,
+      "step": 7027
+    },
+    {
+      "epoch": 0.5168027060813295,
+      "grad_norm": 1.756243348121643,
+      "learning_rate": 9.944032262960296e-05,
+      "loss": 0.3918,
+      "step": 7028
+    },
+    {
+      "epoch": 0.5168762409000662,
+      "grad_norm": 1.387344241142273,
+      "learning_rate": 9.941650683605962e-05,
+      "loss": 0.3698,
+      "step": 7029
+    },
+    {
+      "epoch": 0.5169497757188029,
+      "grad_norm": 1.9213221073150635,
+      "learning_rate": 9.93926910756126e-05,
+      "loss": 0.3556,
+      "step": 7030
+    },
+    {
+      "epoch": 0.5170233105375395,
+      "grad_norm": 1.8317824602127075,
+      "learning_rate": 9.936887534961281e-05,
+      "loss": 0.3774,
+      "step": 7031
+    },
+    {
+      "epoch": 0.5170968453562762,
+      "grad_norm": 1.3969117403030396,
+      "learning_rate": 9.934505965941108e-05,
+      "loss": 0.3521,
+      "step": 7032
+    },
+    {
+      "epoch": 0.5171703801750128,
+      "grad_norm": 1.1102147102355957,
+      "learning_rate": 9.932124400635823e-05,
+      "loss": 0.3806,
+      "step": 7033
+    },
+    {
+      "epoch": 0.5172439149937496,
+      "grad_norm": 1.625108003616333,
+      "learning_rate": 9.929742839180516e-05,
+      "loss": 0.3905,
+      "step": 7034
+    },
+    {
+      "epoch": 0.5173174498124862,
+      "grad_norm": 1.1010870933532715,
+      "learning_rate": 9.92736128171027e-05,
+      "loss": 0.3636,
+      "step": 7035
+    },
+    {
+      "epoch": 0.5173909846312229,
+      "grad_norm": 2.436784505844116,
+      "learning_rate": 9.924979728360165e-05,
+      "loss": 0.3688,
+      "step": 7036
+    },
+    {
+      "epoch": 0.5174645194499595,
+      "grad_norm": 2.242478370666504,
+      "learning_rate": 9.922598179265291e-05,
+      "loss": 0.3905,
+      "step": 7037
+    },
+    {
+      "epoch": 0.5175380542686963,
+      "grad_norm": 1.1737369298934937,
+      "learning_rate": 9.920216634560732e-05,
+      "loss": 0.378,
+      "step": 7038
+    },
+    {
+      "epoch": 0.5176115890874329,
+      "grad_norm": 1.3717150688171387,
+      "learning_rate": 9.917835094381565e-05,
+      "loss": 0.3762,
+      "step": 7039
+    },
+    {
+      "epoch": 0.5176851239061696,
+      "grad_norm": 1.0193463563919067,
+      "learning_rate": 9.915453558862885e-05,
+      "loss": 0.383,
+      "step": 7040
+    },
+    {
+      "epoch": 0.5177586587249062,
+      "grad_norm": 0.9043282270431519,
+      "learning_rate": 9.913072028139764e-05,
+      "loss": 0.3802,
+      "step": 7041
+    },
+    {
+      "epoch": 0.517832193543643,
+      "grad_norm": 1.4757564067840576,
+      "learning_rate": 9.910690502347292e-05,
+      "loss": 0.3633,
+      "step": 7042
+    },
+    {
+      "epoch": 0.5179057283623796,
+      "grad_norm": 0.6556277871131897,
+      "learning_rate": 9.908308981620546e-05,
+      "loss": 0.3812,
+      "step": 7043
+    },
+    {
+      "epoch": 0.5179792631811163,
+      "grad_norm": 1.4838405847549438,
+      "learning_rate": 9.905927466094613e-05,
+      "loss": 0.3959,
+      "step": 7044
+    },
+    {
+      "epoch": 0.5180527979998529,
+      "grad_norm": 0.6432194113731384,
+      "learning_rate": 9.903545955904571e-05,
+      "loss": 0.3944,
+      "step": 7045
+    },
+    {
+      "epoch": 0.5181263328185896,
+      "grad_norm": 1.6307108402252197,
+      "learning_rate": 9.901164451185505e-05,
+      "loss": 0.38,
+      "step": 7046
+    },
+    {
+      "epoch": 0.5181998676373263,
+      "grad_norm": 0.6090010404586792,
+      "learning_rate": 9.898782952072497e-05,
+      "loss": 0.3975,
+      "step": 7047
+    },
+    {
+      "epoch": 0.5182734024560629,
+      "grad_norm": 1.6167353391647339,
+      "learning_rate": 9.896401458700627e-05,
+      "loss": 0.38,
+      "step": 7048
+    },
+    {
+      "epoch": 0.5183469372747996,
+      "grad_norm": 0.6842823028564453,
+      "learning_rate": 9.894019971204973e-05,
+      "loss": 0.3816,
+      "step": 7049
+    },
+    {
+      "epoch": 0.5184204720935363,
+      "grad_norm": 2.341948986053467,
+      "learning_rate": 9.89163848972062e-05,
+      "loss": 0.3765,
+      "step": 7050
+    },
+    {
+      "epoch": 0.518494006912273,
+      "grad_norm": 1.5422735214233398,
+      "learning_rate": 9.889257014382647e-05,
+      "loss": 0.3621,
+      "step": 7051
+    },
+    {
+      "epoch": 0.5185675417310096,
+      "grad_norm": 2.307889938354492,
+      "learning_rate": 9.886875545326135e-05,
+      "loss": 0.3684,
+      "step": 7052
+    },
+    {
+      "epoch": 0.5186410765497463,
+      "grad_norm": 2.4424490928649902,
+      "learning_rate": 9.884494082686157e-05,
+      "loss": 0.3603,
+      "step": 7053
+    },
+    {
+      "epoch": 0.518714611368483,
+      "grad_norm": 0.8767408132553101,
+      "learning_rate": 9.882112626597799e-05,
+      "loss": 0.3769,
+      "step": 7054
+    },
+    {
+      "epoch": 0.5187881461872197,
+      "grad_norm": 2.0469651222229004,
+      "learning_rate": 9.879731177196137e-05,
+      "loss": 0.3852,
+      "step": 7055
+    },
+    {
+      "epoch": 0.5188616810059563,
+      "grad_norm": 1.468308687210083,
+      "learning_rate": 9.877349734616246e-05,
+      "loss": 0.3449,
+      "step": 7056
+    },
+    {
+      "epoch": 0.518935215824693,
+      "grad_norm": 1.7857061624526978,
+      "learning_rate": 9.87496829899321e-05,
+      "loss": 0.3706,
+      "step": 7057
+    },
+    {
+      "epoch": 0.5190087506434297,
+      "grad_norm": 1.6522903442382812,
+      "learning_rate": 9.872586870462106e-05,
+      "loss": 0.3446,
+      "step": 7058
+    },
+    {
+      "epoch": 0.5190822854621663,
+      "grad_norm": 1.7246990203857422,
+      "learning_rate": 9.870205449158006e-05,
+      "loss": 0.3755,
+      "step": 7059
+    },
+    {
+      "epoch": 0.519155820280903,
+      "grad_norm": 1.6447720527648926,
+      "learning_rate": 9.867824035215992e-05,
+      "loss": 0.3624,
+      "step": 7060
+    },
+    {
+      "epoch": 0.5192293550996396,
+      "grad_norm": 1.4712893962860107,
+      "learning_rate": 9.865442628771138e-05,
+      "loss": 0.379,
+      "step": 7061
+    },
+    {
+      "epoch": 0.5193028899183764,
+      "grad_norm": 1.3213081359863281,
+      "learning_rate": 9.863061229958519e-05,
+      "loss": 0.396,
+      "step": 7062
+    },
+    {
+      "epoch": 0.519376424737113,
+      "grad_norm": 1.247594952583313,
+      "learning_rate": 9.860679838913215e-05,
+      "loss": 0.3844,
+      "step": 7063
+    },
+    {
+      "epoch": 0.5194499595558497,
+      "grad_norm": 1.023764967918396,
+      "learning_rate": 9.858298455770296e-05,
+      "loss": 0.3746,
+      "step": 7064
+    },
+    {
+      "epoch": 0.5195234943745863,
+      "grad_norm": 1.4305988550186157,
+      "learning_rate": 9.855917080664837e-05,
+      "loss": 0.3638,
+      "step": 7065
+    },
+    {
+      "epoch": 0.5195970291933231,
+      "grad_norm": 1.1928246021270752,
+      "learning_rate": 9.853535713731911e-05,
+      "loss": 0.3977,
+      "step": 7066
+    },
+    {
+      "epoch": 0.5196705640120597,
+      "grad_norm": 1.3083432912826538,
+      "learning_rate": 9.851154355106597e-05,
+      "loss": 0.3575,
+      "step": 7067
+    },
+    {
+      "epoch": 0.5197440988307964,
+      "grad_norm": 1.1033700704574585,
+      "learning_rate": 9.848773004923964e-05,
+      "loss": 0.3758,
+      "step": 7068
+    },
+    {
+      "epoch": 0.519817633649533,
+      "grad_norm": 1.6682946681976318,
+      "learning_rate": 9.846391663319085e-05,
+      "loss": 0.3543,
+      "step": 7069
+    },
+    {
+      "epoch": 0.5198911684682698,
+      "grad_norm": 1.0266906023025513,
+      "learning_rate": 9.844010330427034e-05,
+      "loss": 0.4074,
+      "step": 7070
+    },
+    {
+      "epoch": 0.5199647032870064,
+      "grad_norm": 1.6637169122695923,
+      "learning_rate": 9.841629006382883e-05,
+      "loss": 0.3722,
+      "step": 7071
+    },
+    {
+      "epoch": 0.5200382381057431,
+      "grad_norm": 1.2682172060012817,
+      "learning_rate": 9.8392476913217e-05,
+      "loss": 0.377,
+      "step": 7072
+    },
+    {
+      "epoch": 0.5201117729244797,
+      "grad_norm": 1.5692411661148071,
+      "learning_rate": 9.836866385378557e-05,
+      "loss": 0.385,
+      "step": 7073
+    },
+    {
+      "epoch": 0.5201853077432164,
+      "grad_norm": 1.3697524070739746,
+      "learning_rate": 9.834485088688526e-05,
+      "loss": 0.3666,
+      "step": 7074
+    },
+    {
+      "epoch": 0.5202588425619531,
+      "grad_norm": 1.7978178262710571,
+      "learning_rate": 9.832103801386679e-05,
+      "loss": 0.3809,
+      "step": 7075
+    },
+    {
+      "epoch": 0.5203323773806897,
+      "grad_norm": 1.7165476083755493,
+      "learning_rate": 9.829722523608076e-05,
+      "loss": 0.3814,
+      "step": 7076
+    },
+    {
+      "epoch": 0.5204059121994264,
+      "grad_norm": 1.1565489768981934,
+      "learning_rate": 9.827341255487794e-05,
+      "loss": 0.3813,
+      "step": 7077
+    },
+    {
+      "epoch": 0.5204794470181631,
+      "grad_norm": 1.0121455192565918,
+      "learning_rate": 9.824959997160898e-05,
+      "loss": 0.3592,
+      "step": 7078
+    },
+    {
+      "epoch": 0.5205529818368998,
+      "grad_norm": 1.6832953691482544,
+      "learning_rate": 9.822578748762454e-05,
+      "loss": 0.4055,
+      "step": 7079
+    },
+    {
+      "epoch": 0.5206265166556364,
+      "grad_norm": 1.242855191230774,
+      "learning_rate": 9.820197510427531e-05,
+      "loss": 0.3545,
+      "step": 7080
+    },
+    {
+      "epoch": 0.5207000514743731,
+      "grad_norm": 1.8137340545654297,
+      "learning_rate": 9.817816282291195e-05,
+      "loss": 0.3856,
+      "step": 7081
+    },
+    {
+      "epoch": 0.5207735862931098,
+      "grad_norm": 2.2343618869781494,
+      "learning_rate": 9.815435064488511e-05,
+      "loss": 0.3825,
+      "step": 7082
+    },
+    {
+      "epoch": 0.5208471211118465,
+      "grad_norm": 1.2916728258132935,
+      "learning_rate": 9.813053857154547e-05,
+      "loss": 0.3547,
+      "step": 7083
+    },
+    {
+      "epoch": 0.5209206559305831,
+      "grad_norm": 1.4363796710968018,
+      "learning_rate": 9.810672660424366e-05,
+      "loss": 0.3732,
+      "step": 7084
+    },
+    {
+      "epoch": 0.5209941907493199,
+      "grad_norm": 1.4408701658248901,
+      "learning_rate": 9.808291474433028e-05,
+      "loss": 0.3936,
+      "step": 7085
+    },
+    {
+      "epoch": 0.5210677255680565,
+      "grad_norm": 1.1660000085830688,
+      "learning_rate": 9.805910299315603e-05,
+      "loss": 0.3865,
+      "step": 7086
+    },
+    {
+      "epoch": 0.5211412603867931,
+      "grad_norm": 1.5336663722991943,
+      "learning_rate": 9.803529135207153e-05,
+      "loss": 0.3839,
+      "step": 7087
+    },
+    {
+      "epoch": 0.5212147952055298,
+      "grad_norm": 1.1277133226394653,
+      "learning_rate": 9.801147982242735e-05,
+      "loss": 0.3718,
+      "step": 7088
+    },
+    {
+      "epoch": 0.5212883300242664,
+      "grad_norm": 1.871336817741394,
+      "learning_rate": 9.79876684055741e-05,
+      "loss": 0.3703,
+      "step": 7089
+    },
+    {
+      "epoch": 0.5213618648430032,
+      "grad_norm": 1.4198256731033325,
+      "learning_rate": 9.796385710286246e-05,
+      "loss": 0.3498,
+      "step": 7090
+    },
+    {
+      "epoch": 0.5214353996617398,
+      "grad_norm": 1.4731569290161133,
+      "learning_rate": 9.794004591564299e-05,
+      "loss": 0.3622,
+      "step": 7091
+    },
+    {
+      "epoch": 0.5215089344804765,
+      "grad_norm": 1.1097257137298584,
+      "learning_rate": 9.791623484526627e-05,
+      "loss": 0.3705,
+      "step": 7092
+    },
+    {
+      "epoch": 0.5215824692992131,
+      "grad_norm": 1.4600423574447632,
+      "learning_rate": 9.789242389308292e-05,
+      "loss": 0.3691,
+      "step": 7093
+    },
+    {
+      "epoch": 0.5216560041179499,
+      "grad_norm": 1.578157901763916,
+      "learning_rate": 9.786861306044352e-05,
+      "loss": 0.3852,
+      "step": 7094
+    },
+    {
+      "epoch": 0.5217295389366865,
+      "grad_norm": 1.7427477836608887,
+      "learning_rate": 9.784480234869861e-05,
+      "loss": 0.3813,
+      "step": 7095
+    },
+    {
+      "epoch": 0.5218030737554232,
+      "grad_norm": 1.6590003967285156,
+      "learning_rate": 9.782099175919882e-05,
+      "loss": 0.3866,
+      "step": 7096
+    },
+    {
+      "epoch": 0.5218766085741599,
+      "grad_norm": 1.0811042785644531,
+      "learning_rate": 9.779718129329467e-05,
+      "loss": 0.3731,
+      "step": 7097
+    },
+    {
+      "epoch": 0.5219501433928966,
+      "grad_norm": 0.9615621566772461,
+      "learning_rate": 9.777337095233675e-05,
+      "loss": 0.3616,
+      "step": 7098
+    },
+    {
+      "epoch": 0.5220236782116332,
+      "grad_norm": 1.484755516052246,
+      "learning_rate": 9.774956073767552e-05,
+      "loss": 0.362,
+      "step": 7099
+    },
+    {
+      "epoch": 0.5220972130303699,
+      "grad_norm": 0.9072230458259583,
+      "learning_rate": 9.77257506506616e-05,
+      "loss": 0.3665,
+      "step": 7100
+    },
+    {
+      "epoch": 0.5221707478491066,
+      "grad_norm": 1.9435968399047852,
+      "learning_rate": 9.770194069264552e-05,
+      "loss": 0.3778,
+      "step": 7101
+    },
+    {
+      "epoch": 0.5222442826678432,
+      "grad_norm": 1.7645657062530518,
+      "learning_rate": 9.767813086497776e-05,
+      "loss": 0.3777,
+      "step": 7102
+    },
+    {
+      "epoch": 0.5223178174865799,
+      "grad_norm": 1.1558144092559814,
+      "learning_rate": 9.765432116900889e-05,
+      "loss": 0.3692,
+      "step": 7103
+    },
+    {
+      "epoch": 0.5223913523053165,
+      "grad_norm": 1.3134407997131348,
+      "learning_rate": 9.763051160608939e-05,
+      "loss": 0.3744,
+      "step": 7104
+    },
+    {
+      "epoch": 0.5224648871240533,
+      "grad_norm": 1.1638838052749634,
+      "learning_rate": 9.760670217756977e-05,
+      "loss": 0.3796,
+      "step": 7105
+    },
+    {
+      "epoch": 0.5225384219427899,
+      "grad_norm": 0.9433916211128235,
+      "learning_rate": 9.758289288480054e-05,
+      "loss": 0.403,
+      "step": 7106
+    },
+    {
+      "epoch": 0.5226119567615266,
+      "grad_norm": 1.4700725078582764,
+      "learning_rate": 9.755908372913217e-05,
+      "loss": 0.3534,
+      "step": 7107
+    },
+    {
+      "epoch": 0.5226854915802632,
+      "grad_norm": 1.1662036180496216,
+      "learning_rate": 9.753527471191513e-05,
+      "loss": 0.3634,
+      "step": 7108
+    },
+    {
+      "epoch": 0.522759026399,
+      "grad_norm": 2.0272412300109863,
+      "learning_rate": 9.751146583449993e-05,
+      "loss": 0.3873,
+      "step": 7109
+    },
+    {
+      "epoch": 0.5228325612177366,
+      "grad_norm": 1.806928277015686,
+      "learning_rate": 9.748765709823704e-05,
+      "loss": 0.3666,
+      "step": 7110
+    },
+    {
+      "epoch": 0.5229060960364733,
+      "grad_norm": 1.0740827322006226,
+      "learning_rate": 9.746384850447686e-05,
+      "loss": 0.3681,
+      "step": 7111
+    },
+    {
+      "epoch": 0.5229796308552099,
+      "grad_norm": 1.2031607627868652,
+      "learning_rate": 9.744004005456987e-05,
+      "loss": 0.347,
+      "step": 7112
+    },
+    {
+      "epoch": 0.5230531656739467,
+      "grad_norm": 1.346186637878418,
+      "learning_rate": 9.741623174986651e-05,
+      "loss": 0.3553,
+      "step": 7113
+    },
+    {
+      "epoch": 0.5231267004926833,
+      "grad_norm": 1.2415810823440552,
+      "learning_rate": 9.739242359171721e-05,
+      "loss": 0.3593,
+      "step": 7114
+    },
+    {
+      "epoch": 0.5232002353114199,
+      "grad_norm": 1.96031653881073,
+      "learning_rate": 9.736861558147237e-05,
+      "loss": 0.3691,
+      "step": 7115
+    },
+    {
+      "epoch": 0.5232737701301566,
+      "grad_norm": 1.6827462911605835,
+      "learning_rate": 9.734480772048248e-05,
+      "loss": 0.3549,
+      "step": 7116
+    },
+    {
+      "epoch": 0.5233473049488933,
+      "grad_norm": 1.0201301574707031,
+      "learning_rate": 9.732100001009788e-05,
+      "loss": 0.3557,
+      "step": 7117
+    },
+    {
+      "epoch": 0.52342083976763,
+      "grad_norm": 0.9984351396560669,
+      "learning_rate": 9.729719245166897e-05,
+      "loss": 0.3831,
+      "step": 7118
+    },
+    {
+      "epoch": 0.5234943745863666,
+      "grad_norm": 1.148034691810608,
+      "learning_rate": 9.727338504654618e-05,
+      "loss": 0.3602,
+      "step": 7119
+    },
+    {
+      "epoch": 0.5235679094051033,
+      "grad_norm": 0.7471445798873901,
+      "learning_rate": 9.724957779607985e-05,
+      "loss": 0.3614,
+      "step": 7120
+    },
+    {
+      "epoch": 0.52364144422384,
+      "grad_norm": 1.574520468711853,
+      "learning_rate": 9.722577070162037e-05,
+      "loss": 0.3402,
+      "step": 7121
+    },
+    {
+      "epoch": 0.5237149790425767,
+      "grad_norm": 0.9627656936645508,
+      "learning_rate": 9.720196376451814e-05,
+      "loss": 0.3579,
+      "step": 7122
+    },
+    {
+      "epoch": 0.5237885138613133,
+      "grad_norm": 1.7723140716552734,
+      "learning_rate": 9.717815698612345e-05,
+      "loss": 0.3373,
+      "step": 7123
+    },
+    {
+      "epoch": 0.52386204868005,
+      "grad_norm": 1.968131422996521,
+      "learning_rate": 9.715435036778668e-05,
+      "loss": 0.3621,
+      "step": 7124
+    },
+    {
+      "epoch": 0.5239355834987867,
+      "grad_norm": 0.7472513914108276,
+      "learning_rate": 9.713054391085813e-05,
+      "loss": 0.3598,
+      "step": 7125
+    },
+    {
+      "epoch": 0.5240091183175234,
+      "grad_norm": 1.5131677389144897,
+      "learning_rate": 9.710673761668819e-05,
+      "loss": 0.3797,
+      "step": 7126
+    },
+    {
+      "epoch": 0.52408265313626,
+      "grad_norm": 0.8889824151992798,
+      "learning_rate": 9.708293148662713e-05,
+      "loss": 0.3737,
+      "step": 7127
+    },
+    {
+      "epoch": 0.5241561879549967,
+      "grad_norm": 0.9433708786964417,
+      "learning_rate": 9.705912552202528e-05,
+      "loss": 0.3834,
+      "step": 7128
+    },
+    {
+      "epoch": 0.5242297227737334,
+      "grad_norm": 1.114631175994873,
+      "learning_rate": 9.703531972423294e-05,
+      "loss": 0.3688,
+      "step": 7129
+    },
+    {
+      "epoch": 0.52430325759247,
+      "grad_norm": 0.7529508471488953,
+      "learning_rate": 9.70115140946004e-05,
+      "loss": 0.3611,
+      "step": 7130
+    },
+    {
+      "epoch": 0.5243767924112067,
+      "grad_norm": 0.8704289793968201,
+      "learning_rate": 9.69877086344779e-05,
+      "loss": 0.3451,
+      "step": 7131
+    },
+    {
+      "epoch": 0.5244503272299433,
+      "grad_norm": 0.719288170337677,
+      "learning_rate": 9.696390334521578e-05,
+      "loss": 0.3909,
+      "step": 7132
+    },
+    {
+      "epoch": 0.5245238620486801,
+      "grad_norm": 0.7216335535049438,
+      "learning_rate": 9.694009822816425e-05,
+      "loss": 0.3821,
+      "step": 7133
+    },
+    {
+      "epoch": 0.5245973968674167,
+      "grad_norm": 0.82071453332901,
+      "learning_rate": 9.691629328467362e-05,
+      "loss": 0.3628,
+      "step": 7134
+    },
+    {
+      "epoch": 0.5246709316861534,
+      "grad_norm": 0.6755279302597046,
+      "learning_rate": 9.689248851609404e-05,
+      "loss": 0.394,
+      "step": 7135
+    },
+    {
+      "epoch": 0.52474446650489,
+      "grad_norm": 0.862583577632904,
+      "learning_rate": 9.68686839237758e-05,
+      "loss": 0.3886,
+      "step": 7136
+    },
+    {
+      "epoch": 0.5248180013236268,
+      "grad_norm": 0.9288032650947571,
+      "learning_rate": 9.684487950906911e-05,
+      "loss": 0.3561,
+      "step": 7137
+    },
+    {
+      "epoch": 0.5248915361423634,
+      "grad_norm": 1.0538673400878906,
+      "learning_rate": 9.68210752733242e-05,
+      "loss": 0.3555,
+      "step": 7138
+    },
+    {
+      "epoch": 0.5249650709611001,
+      "grad_norm": 1.6210078001022339,
+      "learning_rate": 9.679727121789123e-05,
+      "loss": 0.3923,
+      "step": 7139
+    },
+    {
+      "epoch": 0.5250386057798367,
+      "grad_norm": 0.9252362251281738,
+      "learning_rate": 9.677346734412045e-05,
+      "loss": 0.3616,
+      "step": 7140
+    },
+    {
+      "epoch": 0.5251121405985735,
+      "grad_norm": 1.5386673212051392,
+      "learning_rate": 9.674966365336197e-05,
+      "loss": 0.3599,
+      "step": 7141
+    },
+    {
+      "epoch": 0.5251856754173101,
+      "grad_norm": 1.178550124168396,
+      "learning_rate": 9.672586014696602e-05,
+      "loss": 0.3384,
+      "step": 7142
+    },
+    {
+      "epoch": 0.5252592102360467,
+      "grad_norm": 1.6600446701049805,
+      "learning_rate": 9.670205682628274e-05,
+      "loss": 0.3682,
+      "step": 7143
+    },
+    {
+      "epoch": 0.5253327450547834,
+      "grad_norm": 1.3558202981948853,
+      "learning_rate": 9.667825369266225e-05,
+      "loss": 0.3595,
+      "step": 7144
+    },
+    {
+      "epoch": 0.5254062798735201,
+      "grad_norm": 1.6894885301589966,
+      "learning_rate": 9.665445074745477e-05,
+      "loss": 0.3598,
+      "step": 7145
+    },
+    {
+      "epoch": 0.5254798146922568,
+      "grad_norm": 1.4224244356155396,
+      "learning_rate": 9.663064799201035e-05,
+      "loss": 0.3984,
+      "step": 7146
+    },
+    {
+      "epoch": 0.5255533495109934,
+      "grad_norm": 1.5021629333496094,
+      "learning_rate": 9.660684542767913e-05,
+      "loss": 0.3859,
+      "step": 7147
+    },
+    {
+      "epoch": 0.5256268843297301,
+      "grad_norm": 1.256057620048523,
+      "learning_rate": 9.65830430558112e-05,
+      "loss": 0.3934,
+      "step": 7148
+    },
+    {
+      "epoch": 0.5257004191484668,
+      "grad_norm": 1.229222059249878,
+      "learning_rate": 9.655924087775669e-05,
+      "loss": 0.3681,
+      "step": 7149
+    },
+    {
+      "epoch": 0.5257739539672035,
+      "grad_norm": 1.1071738004684448,
+      "learning_rate": 9.653543889486566e-05,
+      "loss": 0.349,
+      "step": 7150
+    },
+    {
+      "epoch": 0.5258474887859401,
+      "grad_norm": 1.3442269563674927,
+      "learning_rate": 9.651163710848818e-05,
+      "loss": 0.3449,
+      "step": 7151
+    },
+    {
+      "epoch": 0.5259210236046769,
+      "grad_norm": 1.1347600221633911,
+      "learning_rate": 9.648783551997436e-05,
+      "loss": 0.3609,
+      "step": 7152
+    },
+    {
+      "epoch": 0.5259945584234135,
+      "grad_norm": 1.7108619213104248,
+      "learning_rate": 9.64640341306742e-05,
+      "loss": 0.3848,
+      "step": 7153
+    },
+    {
+      "epoch": 0.5260680932421502,
+      "grad_norm": 1.4794178009033203,
+      "learning_rate": 9.644023294193773e-05,
+      "loss": 0.387,
+      "step": 7154
+    },
+    {
+      "epoch": 0.5261416280608868,
+      "grad_norm": 1.14536452293396,
+      "learning_rate": 9.641643195511503e-05,
+      "loss": 0.3823,
+      "step": 7155
+    },
+    {
+      "epoch": 0.5262151628796236,
+      "grad_norm": 1.2871266603469849,
+      "learning_rate": 9.639263117155611e-05,
+      "loss": 0.3813,
+      "step": 7156
+    },
+    {
+      "epoch": 0.5262886976983602,
+      "grad_norm": 1.4418669939041138,
+      "learning_rate": 9.636883059261096e-05,
+      "loss": 0.3603,
+      "step": 7157
+    },
+    {
+      "epoch": 0.5263622325170968,
+      "grad_norm": 1.4685966968536377,
+      "learning_rate": 9.634503021962953e-05,
+      "loss": 0.3565,
+      "step": 7158
+    },
+    {
+      "epoch": 0.5264357673358335,
+      "grad_norm": 1.644921064376831,
+      "learning_rate": 9.632123005396186e-05,
+      "loss": 0.3868,
+      "step": 7159
+    },
+    {
+      "epoch": 0.5265093021545701,
+      "grad_norm": 1.2406879663467407,
+      "learning_rate": 9.62974300969579e-05,
+      "loss": 0.3834,
+      "step": 7160
+    },
+    {
+      "epoch": 0.5265828369733069,
+      "grad_norm": 1.2134079933166504,
+      "learning_rate": 9.627363034996761e-05,
+      "loss": 0.3562,
+      "step": 7161
+    },
+    {
+      "epoch": 0.5266563717920435,
+      "grad_norm": 1.1041450500488281,
+      "learning_rate": 9.624983081434094e-05,
+      "loss": 0.4096,
+      "step": 7162
+    },
+    {
+      "epoch": 0.5267299066107802,
+      "grad_norm": 1.4466402530670166,
+      "learning_rate": 9.622603149142784e-05,
+      "loss": 0.3554,
+      "step": 7163
+    },
+    {
+      "epoch": 0.5268034414295168,
+      "grad_norm": 1.2004505395889282,
+      "learning_rate": 9.62022323825782e-05,
+      "loss": 0.3522,
+      "step": 7164
+    },
+    {
+      "epoch": 0.5268769762482536,
+      "grad_norm": 1.7623111009597778,
+      "learning_rate": 9.617843348914195e-05,
+      "loss": 0.3687,
+      "step": 7165
+    },
+    {
+      "epoch": 0.5269505110669902,
+      "grad_norm": 1.4929182529449463,
+      "learning_rate": 9.615463481246899e-05,
+      "loss": 0.3817,
+      "step": 7166
+    },
+    {
+      "epoch": 0.5270240458857269,
+      "grad_norm": 1.3063092231750488,
+      "learning_rate": 9.613083635390919e-05,
+      "loss": 0.3494,
+      "step": 7167
+    },
+    {
+      "epoch": 0.5270975807044636,
+      "grad_norm": 1.1612385511398315,
+      "learning_rate": 9.610703811481245e-05,
+      "loss": 0.3728,
+      "step": 7168
+    },
+    {
+      "epoch": 0.5271711155232003,
+      "grad_norm": 1.2041301727294922,
+      "learning_rate": 9.608324009652863e-05,
+      "loss": 0.3649,
+      "step": 7169
+    },
+    {
+      "epoch": 0.5272446503419369,
+      "grad_norm": 0.9849921464920044,
+      "learning_rate": 9.605944230040756e-05,
+      "loss": 0.341,
+      "step": 7170
+    },
+    {
+      "epoch": 0.5273181851606735,
+      "grad_norm": 1.484831690788269,
+      "learning_rate": 9.603564472779906e-05,
+      "loss": 0.3649,
+      "step": 7171
+    },
+    {
+      "epoch": 0.5273917199794103,
+      "grad_norm": 1.4748659133911133,
+      "learning_rate": 9.6011847380053e-05,
+      "loss": 0.3827,
+      "step": 7172
+    },
+    {
+      "epoch": 0.5274652547981469,
+      "grad_norm": 1.3545255661010742,
+      "learning_rate": 9.598805025851914e-05,
+      "loss": 0.386,
+      "step": 7173
+    },
+    {
+      "epoch": 0.5275387896168836,
+      "grad_norm": 1.2431703805923462,
+      "learning_rate": 9.59642533645473e-05,
+      "loss": 0.3531,
+      "step": 7174
+    },
+    {
+      "epoch": 0.5276123244356202,
+      "grad_norm": 1.259001612663269,
+      "learning_rate": 9.594045669948728e-05,
+      "loss": 0.386,
+      "step": 7175
+    },
+    {
+      "epoch": 0.527685859254357,
+      "grad_norm": 0.9403466582298279,
+      "learning_rate": 9.591666026468884e-05,
+      "loss": 0.3933,
+      "step": 7176
+    },
+    {
+      "epoch": 0.5277593940730936,
+      "grad_norm": 1.5941017866134644,
+      "learning_rate": 9.589286406150173e-05,
+      "loss": 0.3646,
+      "step": 7177
+    },
+    {
+      "epoch": 0.5278329288918303,
+      "grad_norm": 0.9871537685394287,
+      "learning_rate": 9.586906809127571e-05,
+      "loss": 0.3598,
+      "step": 7178
+    },
+    {
+      "epoch": 0.5279064637105669,
+      "grad_norm": 1.932822346687317,
+      "learning_rate": 9.584527235536053e-05,
+      "loss": 0.3651,
+      "step": 7179
+    },
+    {
+      "epoch": 0.5279799985293037,
+      "grad_norm": 1.693627119064331,
+      "learning_rate": 9.58214768551059e-05,
+      "loss": 0.3578,
+      "step": 7180
+    },
+    {
+      "epoch": 0.5280535333480403,
+      "grad_norm": 1.0931581258773804,
+      "learning_rate": 9.579768159186146e-05,
+      "loss": 0.3803,
+      "step": 7181
+    },
+    {
+      "epoch": 0.528127068166777,
+      "grad_norm": 1.099389910697937,
+      "learning_rate": 9.577388656697698e-05,
+      "loss": 0.3348,
+      "step": 7182
+    },
+    {
+      "epoch": 0.5282006029855136,
+      "grad_norm": 1.309509515762329,
+      "learning_rate": 9.575009178180213e-05,
+      "loss": 0.35,
+      "step": 7183
+    },
+    {
+      "epoch": 0.5282741378042504,
+      "grad_norm": 1.1594316959381104,
+      "learning_rate": 9.572629723768652e-05,
+      "loss": 0.35,
+      "step": 7184
+    },
+    {
+      "epoch": 0.528347672622987,
+      "grad_norm": 1.3354488611221313,
+      "learning_rate": 9.570250293597985e-05,
+      "loss": 0.3685,
+      "step": 7185
+    },
+    {
+      "epoch": 0.5284212074417236,
+      "grad_norm": 1.3345105648040771,
+      "learning_rate": 9.567870887803177e-05,
+      "loss": 0.3685,
+      "step": 7186
+    },
+    {
+      "epoch": 0.5284947422604603,
+      "grad_norm": 1.060811996459961,
+      "learning_rate": 9.565491506519185e-05,
+      "loss": 0.3775,
+      "step": 7187
+    },
+    {
+      "epoch": 0.528568277079197,
+      "grad_norm": 0.8207142353057861,
+      "learning_rate": 9.563112149880976e-05,
+      "loss": 0.3625,
+      "step": 7188
+    },
+    {
+      "epoch": 0.5286418118979337,
+      "grad_norm": 1.48578941822052,
+      "learning_rate": 9.560732818023508e-05,
+      "loss": 0.354,
+      "step": 7189
+    },
+    {
+      "epoch": 0.5287153467166703,
+      "grad_norm": 0.8230457305908203,
+      "learning_rate": 9.558353511081735e-05,
+      "loss": 0.3942,
+      "step": 7190
+    },
+    {
+      "epoch": 0.528788881535407,
+      "grad_norm": 1.791642665863037,
+      "learning_rate": 9.555974229190619e-05,
+      "loss": 0.3589,
+      "step": 7191
+    },
+    {
+      "epoch": 0.5288624163541437,
+      "grad_norm": 1.1229302883148193,
+      "learning_rate": 9.553594972485114e-05,
+      "loss": 0.3793,
+      "step": 7192
+    },
+    {
+      "epoch": 0.5289359511728804,
+      "grad_norm": 2.4383935928344727,
+      "learning_rate": 9.551215741100173e-05,
+      "loss": 0.3482,
+      "step": 7193
+    },
+    {
+      "epoch": 0.529009485991617,
+      "grad_norm": 2.2023987770080566,
+      "learning_rate": 9.548836535170747e-05,
+      "loss": 0.3765,
+      "step": 7194
+    },
+    {
+      "epoch": 0.5290830208103537,
+      "grad_norm": 0.8852183818817139,
+      "learning_rate": 9.546457354831791e-05,
+      "loss": 0.3755,
+      "step": 7195
+    },
+    {
+      "epoch": 0.5291565556290904,
+      "grad_norm": 1.5238890647888184,
+      "learning_rate": 9.54407820021825e-05,
+      "loss": 0.3915,
+      "step": 7196
+    },
+    {
+      "epoch": 0.5292300904478271,
+      "grad_norm": 1.0972931385040283,
+      "learning_rate": 9.541699071465076e-05,
+      "loss": 0.3632,
+      "step": 7197
+    },
+    {
+      "epoch": 0.5293036252665637,
+      "grad_norm": 1.3075436353683472,
+      "learning_rate": 9.539319968707215e-05,
+      "loss": 0.3975,
+      "step": 7198
+    },
+    {
+      "epoch": 0.5293771600853003,
+      "grad_norm": 1.3109169006347656,
+      "learning_rate": 9.53694089207961e-05,
+      "loss": 0.3548,
+      "step": 7199
+    },
+    {
+      "epoch": 0.5294506949040371,
+      "grad_norm": 1.1419289112091064,
+      "learning_rate": 9.534561841717206e-05,
+      "loss": 0.3879,
+      "step": 7200
+    },
+    {
+      "epoch": 0.5295242297227737,
+      "grad_norm": 1.4740480184555054,
+      "learning_rate": 9.532182817754947e-05,
+      "loss": 0.3771,
+      "step": 7201
+    },
+    {
+      "epoch": 0.5295977645415104,
+      "grad_norm": 1.0457448959350586,
+      "learning_rate": 9.529803820327772e-05,
+      "loss": 0.3738,
+      "step": 7202
+    },
+    {
+      "epoch": 0.529671299360247,
+      "grad_norm": 1.230332374572754,
+      "learning_rate": 9.527424849570619e-05,
+      "loss": 0.3724,
+      "step": 7203
+    },
+    {
+      "epoch": 0.5297448341789838,
+      "grad_norm": 1.1266512870788574,
+      "learning_rate": 9.52504590561843e-05,
+      "loss": 0.3715,
+      "step": 7204
+    },
+    {
+      "epoch": 0.5298183689977204,
+      "grad_norm": 1.2192163467407227,
+      "learning_rate": 9.522666988606138e-05,
+      "loss": 0.3547,
+      "step": 7205
+    },
+    {
+      "epoch": 0.5298919038164571,
+      "grad_norm": 1.1803522109985352,
+      "learning_rate": 9.520288098668679e-05,
+      "loss": 0.3599,
+      "step": 7206
+    },
+    {
+      "epoch": 0.5299654386351937,
+      "grad_norm": 1.2486202716827393,
+      "learning_rate": 9.517909235940981e-05,
+      "loss": 0.3519,
+      "step": 7207
+    },
+    {
+      "epoch": 0.5300389734539305,
+      "grad_norm": 1.2200851440429688,
+      "learning_rate": 9.515530400557983e-05,
+      "loss": 0.3652,
+      "step": 7208
+    },
+    {
+      "epoch": 0.5301125082726671,
+      "grad_norm": 0.762481689453125,
+      "learning_rate": 9.513151592654612e-05,
+      "loss": 0.3614,
+      "step": 7209
+    },
+    {
+      "epoch": 0.5301860430914038,
+      "grad_norm": 1.1605385541915894,
+      "learning_rate": 9.510772812365795e-05,
+      "loss": 0.3628,
+      "step": 7210
+    },
+    {
+      "epoch": 0.5302595779101404,
+      "grad_norm": 0.7638424634933472,
+      "learning_rate": 9.508394059826461e-05,
+      "loss": 0.3536,
+      "step": 7211
+    },
+    {
+      "epoch": 0.5303331127288772,
+      "grad_norm": 1.31907320022583,
+      "learning_rate": 9.506015335171536e-05,
+      "loss": 0.3864,
+      "step": 7212
+    },
+    {
+      "epoch": 0.5304066475476138,
+      "grad_norm": 0.7960651516914368,
+      "learning_rate": 9.50363663853594e-05,
+      "loss": 0.3846,
+      "step": 7213
+    },
+    {
+      "epoch": 0.5304801823663504,
+      "grad_norm": 1.4132665395736694,
+      "learning_rate": 9.501257970054601e-05,
+      "loss": 0.3899,
+      "step": 7214
+    },
+    {
+      "epoch": 0.5305537171850871,
+      "grad_norm": 0.9531996846199036,
+      "learning_rate": 9.498879329862436e-05,
+      "loss": 0.3534,
+      "step": 7215
+    },
+    {
+      "epoch": 0.5306272520038238,
+      "grad_norm": 2.0461742877960205,
+      "learning_rate": 9.496500718094363e-05,
+      "loss": 0.3409,
+      "step": 7216
+    },
+    {
+      "epoch": 0.5307007868225605,
+      "grad_norm": 1.5764192342758179,
+      "learning_rate": 9.4941221348853e-05,
+      "loss": 0.4236,
+      "step": 7217
+    },
+    {
+      "epoch": 0.5307743216412971,
+      "grad_norm": 1.3118692636489868,
+      "learning_rate": 9.491743580370162e-05,
+      "loss": 0.3803,
+      "step": 7218
+    },
+    {
+      "epoch": 0.5308478564600339,
+      "grad_norm": 1.0329914093017578,
+      "learning_rate": 9.489365054683868e-05,
+      "loss": 0.3708,
+      "step": 7219
+    },
+    {
+      "epoch": 0.5309213912787705,
+      "grad_norm": 1.4288182258605957,
+      "learning_rate": 9.486986557961324e-05,
+      "loss": 0.4015,
+      "step": 7220
+    },
+    {
+      "epoch": 0.5309949260975072,
+      "grad_norm": 1.150070309638977,
+      "learning_rate": 9.484608090337446e-05,
+      "loss": 0.3661,
+      "step": 7221
+    },
+    {
+      "epoch": 0.5310684609162438,
+      "grad_norm": 1.8445055484771729,
+      "learning_rate": 9.482229651947139e-05,
+      "loss": 0.3706,
+      "step": 7222
+    },
+    {
+      "epoch": 0.5311419957349806,
+      "grad_norm": 1.4581702947616577,
+      "learning_rate": 9.479851242925313e-05,
+      "loss": 0.3728,
+      "step": 7223
+    },
+    {
+      "epoch": 0.5312155305537172,
+      "grad_norm": 1.2194639444351196,
+      "learning_rate": 9.477472863406874e-05,
+      "loss": 0.3645,
+      "step": 7224
+    },
+    {
+      "epoch": 0.5312890653724539,
+      "grad_norm": 1.3643345832824707,
+      "learning_rate": 9.475094513526726e-05,
+      "loss": 0.3722,
+      "step": 7225
+    },
+    {
+      "epoch": 0.5313626001911905,
+      "grad_norm": 1.1241660118103027,
+      "learning_rate": 9.47271619341977e-05,
+      "loss": 0.4163,
+      "step": 7226
+    },
+    {
+      "epoch": 0.5314361350099271,
+      "grad_norm": 1.348734974861145,
+      "learning_rate": 9.470337903220912e-05,
+      "loss": 0.3734,
+      "step": 7227
+    },
+    {
+      "epoch": 0.5315096698286639,
+      "grad_norm": 1.5426536798477173,
+      "learning_rate": 9.467959643065044e-05,
+      "loss": 0.3763,
+      "step": 7228
+    },
+    {
+      "epoch": 0.5315832046474005,
+      "grad_norm": 0.8953385949134827,
+      "learning_rate": 9.465581413087068e-05,
+      "loss": 0.3886,
+      "step": 7229
+    },
+    {
+      "epoch": 0.5316567394661372,
+      "grad_norm": 1.524532437324524,
+      "learning_rate": 9.463203213421875e-05,
+      "loss": 0.3954,
+      "step": 7230
+    },
+    {
+      "epoch": 0.5317302742848738,
+      "grad_norm": 1.1919137239456177,
+      "learning_rate": 9.460825044204365e-05,
+      "loss": 0.3679,
+      "step": 7231
+    },
+    {
+      "epoch": 0.5318038091036106,
+      "grad_norm": 1.414925217628479,
+      "learning_rate": 9.458446905569427e-05,
+      "loss": 0.3401,
+      "step": 7232
+    },
+    {
+      "epoch": 0.5318773439223472,
+      "grad_norm": 1.4006606340408325,
+      "learning_rate": 9.456068797651949e-05,
+      "loss": 0.3505,
+      "step": 7233
+    },
+    {
+      "epoch": 0.5319508787410839,
+      "grad_norm": 1.1967071294784546,
+      "learning_rate": 9.453690720586827e-05,
+      "loss": 0.3776,
+      "step": 7234
+    },
+    {
+      "epoch": 0.5320244135598206,
+      "grad_norm": 0.990182101726532,
+      "learning_rate": 9.451312674508942e-05,
+      "loss": 0.3983,
+      "step": 7235
+    },
+    {
+      "epoch": 0.5320979483785573,
+      "grad_norm": 1.7103748321533203,
+      "learning_rate": 9.44893465955318e-05,
+      "loss": 0.3677,
+      "step": 7236
+    },
+    {
+      "epoch": 0.5321714831972939,
+      "grad_norm": 1.4977954626083374,
+      "learning_rate": 9.446556675854428e-05,
+      "loss": 0.3633,
+      "step": 7237
+    },
+    {
+      "epoch": 0.5322450180160306,
+      "grad_norm": 1.425459623336792,
+      "learning_rate": 9.444178723547565e-05,
+      "loss": 0.3912,
+      "step": 7238
+    },
+    {
+      "epoch": 0.5323185528347673,
+      "grad_norm": 1.4520543813705444,
+      "learning_rate": 9.441800802767473e-05,
+      "loss": 0.3603,
+      "step": 7239
+    },
+    {
+      "epoch": 0.532392087653504,
+      "grad_norm": 1.1796832084655762,
+      "learning_rate": 9.439422913649025e-05,
+      "loss": 0.3429,
+      "step": 7240
+    },
+    {
+      "epoch": 0.5324656224722406,
+      "grad_norm": 1.0059031248092651,
+      "learning_rate": 9.437045056327103e-05,
+      "loss": 0.3629,
+      "step": 7241
+    },
+    {
+      "epoch": 0.5325391572909772,
+      "grad_norm": 1.1463630199432373,
+      "learning_rate": 9.434667230936579e-05,
+      "loss": 0.3777,
+      "step": 7242
+    },
+    {
+      "epoch": 0.532612692109714,
+      "grad_norm": 0.764229416847229,
+      "learning_rate": 9.432289437612325e-05,
+      "loss": 0.4024,
+      "step": 7243
+    },
+    {
+      "epoch": 0.5326862269284506,
+      "grad_norm": 1.0943878889083862,
+      "learning_rate": 9.429911676489215e-05,
+      "loss": 0.3606,
+      "step": 7244
+    },
+    {
+      "epoch": 0.5327597617471873,
+      "grad_norm": 0.5458812713623047,
+      "learning_rate": 9.427533947702116e-05,
+      "loss": 0.3639,
+      "step": 7245
+    },
+    {
+      "epoch": 0.5328332965659239,
+      "grad_norm": 1.0052632093429565,
+      "learning_rate": 9.425156251385894e-05,
+      "loss": 0.3857,
+      "step": 7246
+    },
+    {
+      "epoch": 0.5329068313846607,
+      "grad_norm": 0.4864506125450134,
+      "learning_rate": 9.422778587675418e-05,
+      "loss": 0.3982,
+      "step": 7247
+    },
+    {
+      "epoch": 0.5329803662033973,
+      "grad_norm": 1.086087942123413,
+      "learning_rate": 9.420400956705549e-05,
+      "loss": 0.3688,
+      "step": 7248
+    },
+    {
+      "epoch": 0.533053901022134,
+      "grad_norm": 0.7141873240470886,
+      "learning_rate": 9.418023358611148e-05,
+      "loss": 0.3863,
+      "step": 7249
+    },
+    {
+      "epoch": 0.5331274358408706,
+      "grad_norm": 0.822238028049469,
+      "learning_rate": 9.415645793527081e-05,
+      "loss": 0.3715,
+      "step": 7250
+    },
+    {
+      "epoch": 0.5332009706596074,
+      "grad_norm": 1.093701958656311,
+      "learning_rate": 9.413268261588198e-05,
+      "loss": 0.4032,
+      "step": 7251
+    },
+    {
+      "epoch": 0.533274505478344,
+      "grad_norm": 0.5392915606498718,
+      "learning_rate": 9.41089076292936e-05,
+      "loss": 0.3923,
+      "step": 7252
+    },
+    {
+      "epoch": 0.5333480402970807,
+      "grad_norm": 1.1889288425445557,
+      "learning_rate": 9.408513297685416e-05,
+      "loss": 0.3703,
+      "step": 7253
+    },
+    {
+      "epoch": 0.5334215751158173,
+      "grad_norm": 0.4814428389072418,
+      "learning_rate": 9.406135865991224e-05,
+      "loss": 0.3753,
+      "step": 7254
+    },
+    {
+      "epoch": 0.533495109934554,
+      "grad_norm": 1.0472478866577148,
+      "learning_rate": 9.403758467981634e-05,
+      "loss": 0.4094,
+      "step": 7255
+    },
+    {
+      "epoch": 0.5335686447532907,
+      "grad_norm": 0.9217185974121094,
+      "learning_rate": 9.40138110379149e-05,
+      "loss": 0.3672,
+      "step": 7256
+    },
+    {
+      "epoch": 0.5336421795720273,
+      "grad_norm": 0.5079224109649658,
+      "learning_rate": 9.399003773555643e-05,
+      "loss": 0.364,
+      "step": 7257
+    },
+    {
+      "epoch": 0.533715714390764,
+      "grad_norm": 0.5679404139518738,
+      "learning_rate": 9.396626477408937e-05,
+      "loss": 0.4006,
+      "step": 7258
+    },
+    {
+      "epoch": 0.5337892492095007,
+      "grad_norm": 0.5881639719009399,
+      "learning_rate": 9.394249215486211e-05,
+      "loss": 0.3615,
+      "step": 7259
+    },
+    {
+      "epoch": 0.5338627840282374,
+      "grad_norm": 0.9697460532188416,
+      "learning_rate": 9.391871987922313e-05,
+      "loss": 0.3778,
+      "step": 7260
+    },
+    {
+      "epoch": 0.533936318846974,
+      "grad_norm": 0.548347532749176,
+      "learning_rate": 9.389494794852077e-05,
+      "loss": 0.3806,
+      "step": 7261
+    },
+    {
+      "epoch": 0.5340098536657107,
+      "grad_norm": 0.8722951412200928,
+      "learning_rate": 9.387117636410341e-05,
+      "loss": 0.3938,
+      "step": 7262
+    },
+    {
+      "epoch": 0.5340833884844474,
+      "grad_norm": 1.0023561716079712,
+      "learning_rate": 9.384740512731936e-05,
+      "loss": 0.3633,
+      "step": 7263
+    },
+    {
+      "epoch": 0.5341569233031841,
+      "grad_norm": 0.6502938270568848,
+      "learning_rate": 9.3823634239517e-05,
+      "loss": 0.3551,
+      "step": 7264
+    },
+    {
+      "epoch": 0.5342304581219207,
+      "grad_norm": 1.2074143886566162,
+      "learning_rate": 9.379986370204463e-05,
+      "loss": 0.3622,
+      "step": 7265
+    },
+    {
+      "epoch": 0.5343039929406574,
+      "grad_norm": 0.7446590662002563,
+      "learning_rate": 9.377609351625052e-05,
+      "loss": 0.369,
+      "step": 7266
+    },
+    {
+      "epoch": 0.5343775277593941,
+      "grad_norm": 0.9448925852775574,
+      "learning_rate": 9.375232368348295e-05,
+      "loss": 0.3904,
+      "step": 7267
+    },
+    {
+      "epoch": 0.5344510625781308,
+      "grad_norm": 1.1593705415725708,
+      "learning_rate": 9.372855420509021e-05,
+      "loss": 0.3497,
+      "step": 7268
+    },
+    {
+      "epoch": 0.5345245973968674,
+      "grad_norm": 0.590197741985321,
+      "learning_rate": 9.370478508242044e-05,
+      "loss": 0.3413,
+      "step": 7269
+    },
+    {
+      "epoch": 0.534598132215604,
+      "grad_norm": 1.0485050678253174,
+      "learning_rate": 9.368101631682194e-05,
+      "loss": 0.3479,
+      "step": 7270
+    },
+    {
+      "epoch": 0.5346716670343408,
+      "grad_norm": 0.62042236328125,
+      "learning_rate": 9.365724790964288e-05,
+      "loss": 0.3623,
+      "step": 7271
+    },
+    {
+      "epoch": 0.5347452018530774,
+      "grad_norm": 0.6363711953163147,
+      "learning_rate": 9.363347986223136e-05,
+      "loss": 0.3684,
+      "step": 7272
+    },
+    {
+      "epoch": 0.5348187366718141,
+      "grad_norm": 0.517540693283081,
+      "learning_rate": 9.360971217593562e-05,
+      "loss": 0.3486,
+      "step": 7273
+    },
+    {
+      "epoch": 0.5348922714905507,
+      "grad_norm": 0.6075787544250488,
+      "learning_rate": 9.358594485210376e-05,
+      "loss": 0.3406,
+      "step": 7274
+    },
+    {
+      "epoch": 0.5349658063092875,
+      "grad_norm": 0.6840835809707642,
+      "learning_rate": 9.356217789208387e-05,
+      "loss": 0.3494,
+      "step": 7275
+    },
+    {
+      "epoch": 0.5350393411280241,
+      "grad_norm": 0.46150657534599304,
+      "learning_rate": 9.353841129722402e-05,
+      "loss": 0.3711,
+      "step": 7276
+    },
+    {
+      "epoch": 0.5351128759467608,
+      "grad_norm": 0.8826652765274048,
+      "learning_rate": 9.351464506887231e-05,
+      "loss": 0.3758,
+      "step": 7277
+    },
+    {
+      "epoch": 0.5351864107654974,
+      "grad_norm": 0.8460638523101807,
+      "learning_rate": 9.349087920837679e-05,
+      "loss": 0.3566,
+      "step": 7278
+    },
+    {
+      "epoch": 0.5352599455842342,
+      "grad_norm": 0.5551605224609375,
+      "learning_rate": 9.346711371708544e-05,
+      "loss": 0.378,
+      "step": 7279
+    },
+    {
+      "epoch": 0.5353334804029708,
+      "grad_norm": 1.1509603261947632,
+      "learning_rate": 9.344334859634631e-05,
+      "loss": 0.3487,
+      "step": 7280
+    },
+    {
+      "epoch": 0.5354070152217075,
+      "grad_norm": 0.6365744471549988,
+      "learning_rate": 9.341958384750739e-05,
+      "loss": 0.3668,
+      "step": 7281
+    },
+    {
+      "epoch": 0.5354805500404441,
+      "grad_norm": 0.65008544921875,
+      "learning_rate": 9.339581947191658e-05,
+      "loss": 0.3608,
+      "step": 7282
+    },
+    {
+      "epoch": 0.5355540848591808,
+      "grad_norm": 0.5985718369483948,
+      "learning_rate": 9.337205547092188e-05,
+      "loss": 0.3712,
+      "step": 7283
+    },
+    {
+      "epoch": 0.5356276196779175,
+      "grad_norm": 0.5714049339294434,
+      "learning_rate": 9.334829184587119e-05,
+      "loss": 0.3619,
+      "step": 7284
+    },
+    {
+      "epoch": 0.5357011544966541,
+      "grad_norm": 0.9306445717811584,
+      "learning_rate": 9.332452859811243e-05,
+      "loss": 0.3326,
+      "step": 7285
+    },
+    {
+      "epoch": 0.5357746893153909,
+      "grad_norm": 0.619705319404602,
+      "learning_rate": 9.330076572899341e-05,
+      "loss": 0.3456,
+      "step": 7286
+    },
+    {
+      "epoch": 0.5358482241341275,
+      "grad_norm": 0.493845671415329,
+      "learning_rate": 9.327700323986205e-05,
+      "loss": 0.3517,
+      "step": 7287
+    },
+    {
+      "epoch": 0.5359217589528642,
+      "grad_norm": 0.6189579963684082,
+      "learning_rate": 9.325324113206616e-05,
+      "loss": 0.343,
+      "step": 7288
+    },
+    {
+      "epoch": 0.5359952937716008,
+      "grad_norm": 0.4842481315135956,
+      "learning_rate": 9.322947940695353e-05,
+      "loss": 0.3408,
+      "step": 7289
+    },
+    {
+      "epoch": 0.5360688285903376,
+      "grad_norm": 0.7503200769424438,
+      "learning_rate": 9.3205718065872e-05,
+      "loss": 0.3455,
+      "step": 7290
+    },
+    {
+      "epoch": 0.5361423634090742,
+      "grad_norm": 0.6126765012741089,
+      "learning_rate": 9.31819571101693e-05,
+      "loss": 0.3487,
+      "step": 7291
+    },
+    {
+      "epoch": 0.5362158982278109,
+      "grad_norm": 0.6578033566474915,
+      "learning_rate": 9.315819654119318e-05,
+      "loss": 0.3663,
+      "step": 7292
+    },
+    {
+      "epoch": 0.5362894330465475,
+      "grad_norm": 0.8285261988639832,
+      "learning_rate": 9.313443636029139e-05,
+      "loss": 0.3705,
+      "step": 7293
+    },
+    {
+      "epoch": 0.5363629678652843,
+      "grad_norm": 0.4959491491317749,
+      "learning_rate": 9.311067656881162e-05,
+      "loss": 0.3375,
+      "step": 7294
+    },
+    {
+      "epoch": 0.5364365026840209,
+      "grad_norm": 0.504363477230072,
+      "learning_rate": 9.308691716810151e-05,
+      "loss": 0.3528,
+      "step": 7295
+    },
+    {
+      "epoch": 0.5365100375027576,
+      "grad_norm": 0.5385110378265381,
+      "learning_rate": 9.306315815950878e-05,
+      "loss": 0.3366,
+      "step": 7296
+    },
+    {
+      "epoch": 0.5365835723214942,
+      "grad_norm": 0.5559543967247009,
+      "learning_rate": 9.303939954438106e-05,
+      "loss": 0.3562,
+      "step": 7297
+    },
+    {
+      "epoch": 0.5366571071402308,
+      "grad_norm": 0.8725659847259521,
+      "learning_rate": 9.301564132406593e-05,
+      "loss": 0.353,
+      "step": 7298
+    },
+    {
+      "epoch": 0.5367306419589676,
+      "grad_norm": 0.5457035899162292,
+      "learning_rate": 9.299188349991097e-05,
+      "loss": 0.3471,
+      "step": 7299
+    },
+    {
+      "epoch": 0.5368041767777042,
+      "grad_norm": 0.6554101705551147,
+      "learning_rate": 9.29681260732638e-05,
+      "loss": 0.3413,
+      "step": 7300
+    },
+    {
+      "epoch": 0.5368777115964409,
+      "grad_norm": 0.8648754358291626,
+      "learning_rate": 9.294436904547194e-05,
+      "loss": 0.3302,
+      "step": 7301
+    },
+    {
+      "epoch": 0.5369512464151776,
+      "grad_norm": 0.7007479667663574,
+      "learning_rate": 9.29206124178829e-05,
+      "loss": 0.3553,
+      "step": 7302
+    },
+    {
+      "epoch": 0.5370247812339143,
+      "grad_norm": 0.5611249208450317,
+      "learning_rate": 9.289685619184419e-05,
+      "loss": 0.3606,
+      "step": 7303
+    },
+    {
+      "epoch": 0.5370983160526509,
+      "grad_norm": 0.7371490001678467,
+      "learning_rate": 9.287310036870332e-05,
+      "loss": 0.3794,
+      "step": 7304
+    },
+    {
+      "epoch": 0.5371718508713876,
+      "grad_norm": 0.6376768350601196,
+      "learning_rate": 9.284934494980767e-05,
+      "loss": 0.3804,
+      "step": 7305
+    },
+    {
+      "epoch": 0.5372453856901243,
+      "grad_norm": 0.47845056653022766,
+      "learning_rate": 9.282558993650476e-05,
+      "loss": 0.3374,
+      "step": 7306
+    },
+    {
+      "epoch": 0.537318920508861,
+      "grad_norm": 0.6841081976890564,
+      "learning_rate": 9.280183533014197e-05,
+      "loss": 0.3647,
+      "step": 7307
+    },
+    {
+      "epoch": 0.5373924553275976,
+      "grad_norm": 0.8855305314064026,
+      "learning_rate": 9.277808113206663e-05,
+      "loss": 0.3417,
+      "step": 7308
+    },
+    {
+      "epoch": 0.5374659901463343,
+      "grad_norm": 0.5561086535453796,
+      "learning_rate": 9.27543273436262e-05,
+      "loss": 0.317,
+      "step": 7309
+    },
+    {
+      "epoch": 0.537539524965071,
+      "grad_norm": 0.8709150552749634,
+      "learning_rate": 9.273057396616796e-05,
+      "loss": 0.3317,
+      "step": 7310
+    },
+    {
+      "epoch": 0.5376130597838076,
+      "grad_norm": 1.2016156911849976,
+      "learning_rate": 9.270682100103922e-05,
+      "loss": 0.3475,
+      "step": 7311
+    },
+    {
+      "epoch": 0.5376865946025443,
+      "grad_norm": 0.6650933027267456,
+      "learning_rate": 9.268306844958729e-05,
+      "loss": 0.3222,
+      "step": 7312
+    },
+    {
+      "epoch": 0.5377601294212809,
+      "grad_norm": 0.5878218412399292,
+      "learning_rate": 9.265931631315944e-05,
+      "loss": 0.3617,
+      "step": 7313
+    },
+    {
+      "epoch": 0.5378336642400177,
+      "grad_norm": 0.5996586680412292,
+      "learning_rate": 9.263556459310292e-05,
+      "loss": 0.3253,
+      "step": 7314
+    },
+    {
+      "epoch": 0.5379071990587543,
+      "grad_norm": 0.520183265209198,
+      "learning_rate": 9.261181329076493e-05,
+      "loss": 0.3162,
+      "step": 7315
+    },
+    {
+      "epoch": 0.537980733877491,
+      "grad_norm": 0.6897767782211304,
+      "learning_rate": 9.258806240749271e-05,
+      "loss": 0.3468,
+      "step": 7316
+    },
+    {
+      "epoch": 0.5380542686962276,
+      "grad_norm": 0.4894483685493469,
+      "learning_rate": 9.256431194463342e-05,
+      "loss": 0.3254,
+      "step": 7317
+    },
+    {
+      "epoch": 0.5381278035149644,
+      "grad_norm": 0.6325922012329102,
+      "learning_rate": 9.254056190353418e-05,
+      "loss": 0.3435,
+      "step": 7318
+    },
+    {
+      "epoch": 0.538201338333701,
+      "grad_norm": 0.6058053374290466,
+      "learning_rate": 9.251681228554215e-05,
+      "loss": 0.3295,
+      "step": 7319
+    },
+    {
+      "epoch": 0.5382748731524377,
+      "grad_norm": 0.5743230581283569,
+      "learning_rate": 9.249306309200446e-05,
+      "loss": 0.3181,
+      "step": 7320
+    },
+    {
+      "epoch": 0.5383484079711743,
+      "grad_norm": 0.7654228210449219,
+      "learning_rate": 9.246931432426812e-05,
+      "loss": 0.3118,
+      "step": 7321
+    },
+    {
+      "epoch": 0.5384219427899111,
+      "grad_norm": 0.6255936026573181,
+      "learning_rate": 9.244556598368021e-05,
+      "loss": 0.3228,
+      "step": 7322
+    },
+    {
+      "epoch": 0.5384954776086477,
+      "grad_norm": 0.5538206696510315,
+      "learning_rate": 9.242181807158778e-05,
+      "loss": 0.324,
+      "step": 7323
+    },
+    {
+      "epoch": 0.5385690124273844,
+      "grad_norm": 0.4639492928981781,
+      "learning_rate": 9.239807058933782e-05,
+      "loss": 0.3163,
+      "step": 7324
+    },
+    {
+      "epoch": 0.538642547246121,
+      "grad_norm": 0.5495204329490662,
+      "learning_rate": 9.237432353827731e-05,
+      "loss": 0.3452,
+      "step": 7325
+    },
+    {
+      "epoch": 0.5387160820648577,
+      "grad_norm": 0.5126709938049316,
+      "learning_rate": 9.235057691975324e-05,
+      "loss": 0.3138,
+      "step": 7326
+    },
+    {
+      "epoch": 0.5387896168835944,
+      "grad_norm": 0.6468183398246765,
+      "learning_rate": 9.232683073511252e-05,
+      "loss": 0.3531,
+      "step": 7327
+    },
+    {
+      "epoch": 0.538863151702331,
+      "grad_norm": 0.6298972964286804,
+      "learning_rate": 9.230308498570204e-05,
+      "loss": 0.3219,
+      "step": 7328
+    },
+    {
+      "epoch": 0.5389366865210677,
+      "grad_norm": 0.5026318430900574,
+      "learning_rate": 9.227933967286872e-05,
+      "loss": 0.3383,
+      "step": 7329
+    },
+    {
+      "epoch": 0.5390102213398044,
+      "grad_norm": 0.5768738985061646,
+      "learning_rate": 9.225559479795938e-05,
+      "loss": 0.3304,
+      "step": 7330
+    },
+    {
+      "epoch": 0.5390837561585411,
+      "grad_norm": 0.8386024832725525,
+      "learning_rate": 9.223185036232089e-05,
+      "loss": 0.3304,
+      "step": 7331
+    },
+    {
+      "epoch": 0.5391572909772777,
+      "grad_norm": 0.6812724471092224,
+      "learning_rate": 9.220810636730007e-05,
+      "loss": 0.3258,
+      "step": 7332
+    },
+    {
+      "epoch": 0.5392308257960144,
+      "grad_norm": 0.5454076528549194,
+      "learning_rate": 9.218436281424365e-05,
+      "loss": 0.3447,
+      "step": 7333
+    },
+    {
+      "epoch": 0.5393043606147511,
+      "grad_norm": 0.6886448264122009,
+      "learning_rate": 9.216061970449844e-05,
+      "loss": 0.3558,
+      "step": 7334
+    },
+    {
+      "epoch": 0.5393778954334878,
+      "grad_norm": 0.5618711113929749,
+      "learning_rate": 9.213687703941112e-05,
+      "loss": 0.3078,
+      "step": 7335
+    },
+    {
+      "epoch": 0.5394514302522244,
+      "grad_norm": 0.5206992626190186,
+      "learning_rate": 9.211313482032845e-05,
+      "loss": 0.326,
+      "step": 7336
+    },
+    {
+      "epoch": 0.5395249650709611,
+      "grad_norm": 0.8631874322891235,
+      "learning_rate": 9.20893930485971e-05,
+      "loss": 0.3195,
+      "step": 7337
+    },
+    {
+      "epoch": 0.5395984998896978,
+      "grad_norm": 0.6066030263900757,
+      "learning_rate": 9.20656517255637e-05,
+      "loss": 0.3213,
+      "step": 7338
+    },
+    {
+      "epoch": 0.5396720347084344,
+      "grad_norm": 0.5455069541931152,
+      "learning_rate": 9.204191085257495e-05,
+      "loss": 0.3174,
+      "step": 7339
+    },
+    {
+      "epoch": 0.5397455695271711,
+      "grad_norm": 0.7003557085990906,
+      "learning_rate": 9.20181704309774e-05,
+      "loss": 0.3436,
+      "step": 7340
+    },
+    {
+      "epoch": 0.5398191043459077,
+      "grad_norm": 0.5184950828552246,
+      "learning_rate": 9.199443046211763e-05,
+      "loss": 0.3277,
+      "step": 7341
+    },
+    {
+      "epoch": 0.5398926391646445,
+      "grad_norm": 0.5340814590454102,
+      "learning_rate": 9.197069094734222e-05,
+      "loss": 0.3593,
+      "step": 7342
+    },
+    {
+      "epoch": 0.5399661739833811,
+      "grad_norm": 0.8321855664253235,
+      "learning_rate": 9.19469518879977e-05,
+      "loss": 0.3389,
+      "step": 7343
+    },
+    {
+      "epoch": 0.5400397088021178,
+      "grad_norm": 0.7443217635154724,
+      "learning_rate": 9.192321328543058e-05,
+      "loss": 0.3515,
+      "step": 7344
+    },
+    {
+      "epoch": 0.5401132436208544,
+      "grad_norm": 0.5696628093719482,
+      "learning_rate": 9.18994751409873e-05,
+      "loss": 0.337,
+      "step": 7345
+    },
+    {
+      "epoch": 0.5401867784395912,
+      "grad_norm": 0.8080364465713501,
+      "learning_rate": 9.187573745601436e-05,
+      "loss": 0.3404,
+      "step": 7346
+    },
+    {
+      "epoch": 0.5402603132583278,
+      "grad_norm": 0.5429478883743286,
+      "learning_rate": 9.185200023185815e-05,
+      "loss": 0.348,
+      "step": 7347
+    },
+    {
+      "epoch": 0.5403338480770645,
+      "grad_norm": 0.6123213768005371,
+      "learning_rate": 9.182826346986508e-05,
+      "loss": 0.3527,
+      "step": 7348
+    },
+    {
+      "epoch": 0.5404073828958011,
+      "grad_norm": 0.9159016013145447,
+      "learning_rate": 9.180452717138154e-05,
+      "loss": 0.3599,
+      "step": 7349
+    },
+    {
+      "epoch": 0.5404809177145379,
+      "grad_norm": 0.6125258803367615,
+      "learning_rate": 9.178079133775388e-05,
+      "loss": 0.3452,
+      "step": 7350
+    },
+    {
+      "epoch": 0.5405544525332745,
+      "grad_norm": 0.5466569066047668,
+      "learning_rate": 9.175705597032838e-05,
+      "loss": 0.3306,
+      "step": 7351
+    },
+    {
+      "epoch": 0.5406279873520112,
+      "grad_norm": 0.775576114654541,
+      "learning_rate": 9.173332107045139e-05,
+      "loss": 0.342,
+      "step": 7352
+    },
+    {
+      "epoch": 0.5407015221707479,
+      "grad_norm": 0.57684326171875,
+      "learning_rate": 9.170958663946915e-05,
+      "loss": 0.3313,
+      "step": 7353
+    },
+    {
+      "epoch": 0.5407750569894845,
+      "grad_norm": 0.6339473724365234,
+      "learning_rate": 9.168585267872789e-05,
+      "loss": 0.3272,
+      "step": 7354
+    },
+    {
+      "epoch": 0.5408485918082212,
+      "grad_norm": 0.529930591583252,
+      "learning_rate": 9.166211918957385e-05,
+      "loss": 0.3377,
+      "step": 7355
+    },
+    {
+      "epoch": 0.5409221266269578,
+      "grad_norm": 0.5180321335792542,
+      "learning_rate": 9.163838617335324e-05,
+      "loss": 0.3292,
+      "step": 7356
+    },
+    {
+      "epoch": 0.5409956614456946,
+      "grad_norm": 0.7516595721244812,
+      "learning_rate": 9.161465363141216e-05,
+      "loss": 0.3353,
+      "step": 7357
+    },
+    {
+      "epoch": 0.5410691962644312,
+      "grad_norm": 0.8671605587005615,
+      "learning_rate": 9.159092156509677e-05,
+      "loss": 0.3061,
+      "step": 7358
+    },
+    {
+      "epoch": 0.5411427310831679,
+      "grad_norm": 0.4681026041507721,
+      "learning_rate": 9.156718997575319e-05,
+      "loss": 0.3338,
+      "step": 7359
+    },
+    {
+      "epoch": 0.5412162659019045,
+      "grad_norm": 1.2579902410507202,
+      "learning_rate": 9.154345886472749e-05,
+      "loss": 0.3358,
+      "step": 7360
+    },
+    {
+      "epoch": 0.5412898007206413,
+      "grad_norm": 1.0311795473098755,
+      "learning_rate": 9.151972823336572e-05,
+      "loss": 0.3536,
+      "step": 7361
+    },
+    {
+      "epoch": 0.5413633355393779,
+      "grad_norm": 0.5110471844673157,
+      "learning_rate": 9.149599808301393e-05,
+      "loss": 0.3844,
+      "step": 7362
+    },
+    {
+      "epoch": 0.5414368703581146,
+      "grad_norm": 1.255849838256836,
+      "learning_rate": 9.14722684150181e-05,
+      "loss": 0.3381,
+      "step": 7363
+    },
+    {
+      "epoch": 0.5415104051768512,
+      "grad_norm": 0.558333694934845,
+      "learning_rate": 9.14485392307242e-05,
+      "loss": 0.3412,
+      "step": 7364
+    },
+    {
+      "epoch": 0.541583939995588,
+      "grad_norm": 1.1093008518218994,
+      "learning_rate": 9.142481053147818e-05,
+      "loss": 0.3451,
+      "step": 7365
+    },
+    {
+      "epoch": 0.5416574748143246,
+      "grad_norm": 1.077789068222046,
+      "learning_rate": 9.140108231862597e-05,
+      "loss": 0.3409,
+      "step": 7366
+    },
+    {
+      "epoch": 0.5417310096330612,
+      "grad_norm": 0.5293436050415039,
+      "learning_rate": 9.137735459351345e-05,
+      "loss": 0.3704,
+      "step": 7367
+    },
+    {
+      "epoch": 0.5418045444517979,
+      "grad_norm": 0.8026546239852905,
+      "learning_rate": 9.135362735748645e-05,
+      "loss": 0.3747,
+      "step": 7368
+    },
+    {
+      "epoch": 0.5418780792705346,
+      "grad_norm": 0.5831196904182434,
+      "learning_rate": 9.132990061189086e-05,
+      "loss": 0.3382,
+      "step": 7369
+    },
+    {
+      "epoch": 0.5419516140892713,
+      "grad_norm": 0.5809759497642517,
+      "learning_rate": 9.130617435807244e-05,
+      "loss": 0.3397,
+      "step": 7370
+    },
+    {
+      "epoch": 0.5420251489080079,
+      "grad_norm": 0.7794941663742065,
+      "learning_rate": 9.128244859737696e-05,
+      "loss": 0.3608,
+      "step": 7371
+    },
+    {
+      "epoch": 0.5420986837267446,
+      "grad_norm": 0.498051255941391,
+      "learning_rate": 9.125872333115022e-05,
+      "loss": 0.3593,
+      "step": 7372
+    },
+    {
+      "epoch": 0.5421722185454813,
+      "grad_norm": 0.7688331007957458,
+      "learning_rate": 9.123499856073794e-05,
+      "loss": 0.3326,
+      "step": 7373
+    },
+    {
+      "epoch": 0.542245753364218,
+      "grad_norm": 0.623866081237793,
+      "learning_rate": 9.121127428748575e-05,
+      "loss": 0.343,
+      "step": 7374
+    },
+    {
+      "epoch": 0.5423192881829546,
+      "grad_norm": 0.4999053478240967,
+      "learning_rate": 9.118755051273938e-05,
+      "loss": 0.3159,
+      "step": 7375
+    },
+    {
+      "epoch": 0.5423928230016913,
+      "grad_norm": 0.9360897541046143,
+      "learning_rate": 9.116382723784446e-05,
+      "loss": 0.3499,
+      "step": 7376
+    },
+    {
+      "epoch": 0.542466357820428,
+      "grad_norm": 0.8763942122459412,
+      "learning_rate": 9.114010446414655e-05,
+      "loss": 0.349,
+      "step": 7377
+    },
+    {
+      "epoch": 0.5425398926391647,
+      "grad_norm": 0.43548014760017395,
+      "learning_rate": 9.111638219299128e-05,
+      "loss": 0.332,
+      "step": 7378
+    },
+    {
+      "epoch": 0.5426134274579013,
+      "grad_norm": 0.9622340202331543,
+      "learning_rate": 9.109266042572421e-05,
+      "loss": 0.3198,
+      "step": 7379
+    },
+    {
+      "epoch": 0.542686962276638,
+      "grad_norm": 0.6820002198219299,
+      "learning_rate": 9.106893916369083e-05,
+      "loss": 0.3332,
+      "step": 7380
+    },
+    {
+      "epoch": 0.5427604970953747,
+      "grad_norm": 0.6874619126319885,
+      "learning_rate": 9.104521840823663e-05,
+      "loss": 0.3638,
+      "step": 7381
+    },
+    {
+      "epoch": 0.5428340319141113,
+      "grad_norm": 0.80240797996521,
+      "learning_rate": 9.102149816070711e-05,
+      "loss": 0.334,
+      "step": 7382
+    },
+    {
+      "epoch": 0.542907566732848,
+      "grad_norm": 0.6058374047279358,
+      "learning_rate": 9.099777842244767e-05,
+      "loss": 0.3699,
+      "step": 7383
+    },
+    {
+      "epoch": 0.5429811015515846,
+      "grad_norm": 0.6843531131744385,
+      "learning_rate": 9.097405919480374e-05,
+      "loss": 0.3477,
+      "step": 7384
+    },
+    {
+      "epoch": 0.5430546363703214,
+      "grad_norm": 0.5609549283981323,
+      "learning_rate": 9.095034047912072e-05,
+      "loss": 0.3374,
+      "step": 7385
+    },
+    {
+      "epoch": 0.543128171189058,
+      "grad_norm": 0.6262587308883667,
+      "learning_rate": 9.092662227674392e-05,
+      "loss": 0.3582,
+      "step": 7386
+    },
+    {
+      "epoch": 0.5432017060077947,
+      "grad_norm": 0.6108248233795166,
+      "learning_rate": 9.090290458901867e-05,
+      "loss": 0.3283,
+      "step": 7387
+    },
+    {
+      "epoch": 0.5432752408265313,
+      "grad_norm": 0.516929030418396,
+      "learning_rate": 9.087918741729029e-05,
+      "loss": 0.3604,
+      "step": 7388
+    },
+    {
+      "epoch": 0.5433487756452681,
+      "grad_norm": 0.5920730829238892,
+      "learning_rate": 9.085547076290402e-05,
+      "loss": 0.3716,
+      "step": 7389
+    },
+    {
+      "epoch": 0.5434223104640047,
+      "grad_norm": 0.5795868635177612,
+      "learning_rate": 9.08317546272051e-05,
+      "loss": 0.3542,
+      "step": 7390
+    },
+    {
+      "epoch": 0.5434958452827414,
+      "grad_norm": 0.6081956028938293,
+      "learning_rate": 9.080803901153875e-05,
+      "loss": 0.3668,
+      "step": 7391
+    },
+    {
+      "epoch": 0.543569380101478,
+      "grad_norm": 0.6697195172309875,
+      "learning_rate": 9.078432391725011e-05,
+      "loss": 0.3597,
+      "step": 7392
+    },
+    {
+      "epoch": 0.5436429149202148,
+      "grad_norm": 0.6832908391952515,
+      "learning_rate": 9.076060934568436e-05,
+      "loss": 0.3765,
+      "step": 7393
+    },
+    {
+      "epoch": 0.5437164497389514,
+      "grad_norm": 0.6505957841873169,
+      "learning_rate": 9.073689529818656e-05,
+      "loss": 0.3635,
+      "step": 7394
+    },
+    {
+      "epoch": 0.543789984557688,
+      "grad_norm": 0.5337830185890198,
+      "learning_rate": 9.071318177610186e-05,
+      "loss": 0.3686,
+      "step": 7395
+    },
+    {
+      "epoch": 0.5438635193764247,
+      "grad_norm": 0.7568143606185913,
+      "learning_rate": 9.06894687807753e-05,
+      "loss": 0.3589,
+      "step": 7396
+    },
+    {
+      "epoch": 0.5439370541951614,
+      "grad_norm": 0.4652504026889801,
+      "learning_rate": 9.066575631355187e-05,
+      "loss": 0.3562,
+      "step": 7397
+    },
+    {
+      "epoch": 0.5440105890138981,
+      "grad_norm": 0.5363380908966064,
+      "learning_rate": 9.064204437577662e-05,
+      "loss": 0.3347,
+      "step": 7398
+    },
+    {
+      "epoch": 0.5440841238326347,
+      "grad_norm": 0.4871891736984253,
+      "learning_rate": 9.061833296879447e-05,
+      "loss": 0.387,
+      "step": 7399
+    },
+    {
+      "epoch": 0.5441576586513714,
+      "grad_norm": 0.6430131196975708,
+      "learning_rate": 9.059462209395036e-05,
+      "loss": 0.3547,
+      "step": 7400
+    },
+    {
+      "epoch": 0.5442311934701081,
+      "grad_norm": 0.7638689875602722,
+      "learning_rate": 9.057091175258924e-05,
+      "loss": 0.3603,
+      "step": 7401
+    },
+    {
+      "epoch": 0.5443047282888448,
+      "grad_norm": 0.4861777722835541,
+      "learning_rate": 9.054720194605599e-05,
+      "loss": 0.3744,
+      "step": 7402
+    },
+    {
+      "epoch": 0.5443782631075814,
+      "grad_norm": 0.8207539916038513,
+      "learning_rate": 9.052349267569538e-05,
+      "loss": 0.3608,
+      "step": 7403
+    },
+    {
+      "epoch": 0.5444517979263181,
+      "grad_norm": 0.8623976707458496,
+      "learning_rate": 9.049978394285225e-05,
+      "loss": 0.3597,
+      "step": 7404
+    },
+    {
+      "epoch": 0.5445253327450548,
+      "grad_norm": 0.7026865482330322,
+      "learning_rate": 9.047607574887141e-05,
+      "loss": 0.3729,
+      "step": 7405
+    },
+    {
+      "epoch": 0.5445988675637915,
+      "grad_norm": 0.6041516661643982,
+      "learning_rate": 9.045236809509762e-05,
+      "loss": 0.3261,
+      "step": 7406
+    },
+    {
+      "epoch": 0.5446724023825281,
+      "grad_norm": 0.5178608298301697,
+      "learning_rate": 9.042866098287555e-05,
+      "loss": 0.3654,
+      "step": 7407
+    },
+    {
+      "epoch": 0.5447459372012649,
+      "grad_norm": 0.5849728584289551,
+      "learning_rate": 9.040495441354998e-05,
+      "loss": 0.3456,
+      "step": 7408
+    },
+    {
+      "epoch": 0.5448194720200015,
+      "grad_norm": 0.5089055895805359,
+      "learning_rate": 9.03812483884655e-05,
+      "loss": 0.3445,
+      "step": 7409
+    },
+    {
+      "epoch": 0.5448930068387381,
+      "grad_norm": 0.5273230671882629,
+      "learning_rate": 9.035754290896675e-05,
+      "loss": 0.336,
+      "step": 7410
+    },
+    {
+      "epoch": 0.5449665416574748,
+      "grad_norm": 0.48057371377944946,
+      "learning_rate": 9.033383797639836e-05,
+      "loss": 0.3575,
+      "step": 7411
+    },
+    {
+      "epoch": 0.5450400764762114,
+      "grad_norm": 0.524737536907196,
+      "learning_rate": 9.031013359210489e-05,
+      "loss": 0.3324,
+      "step": 7412
+    },
+    {
+      "epoch": 0.5451136112949482,
+      "grad_norm": 0.6255861520767212,
+      "learning_rate": 9.028642975743082e-05,
+      "loss": 0.346,
+      "step": 7413
+    },
+    {
+      "epoch": 0.5451871461136848,
+      "grad_norm": 0.7175849676132202,
+      "learning_rate": 9.026272647372078e-05,
+      "loss": 0.3893,
+      "step": 7414
+    },
+    {
+      "epoch": 0.5452606809324215,
+      "grad_norm": 0.5943023562431335,
+      "learning_rate": 9.023902374231913e-05,
+      "loss": 0.3901,
+      "step": 7415
+    },
+    {
+      "epoch": 0.5453342157511581,
+      "grad_norm": 1.2508211135864258,
+      "learning_rate": 9.021532156457037e-05,
+      "loss": 0.3395,
+      "step": 7416
+    },
+    {
+      "epoch": 0.5454077505698949,
+      "grad_norm": 0.46691203117370605,
+      "learning_rate": 9.019161994181886e-05,
+      "loss": 0.3637,
+      "step": 7417
+    },
+    {
+      "epoch": 0.5454812853886315,
+      "grad_norm": 1.1883732080459595,
+      "learning_rate": 9.016791887540905e-05,
+      "loss": 0.3419,
+      "step": 7418
+    },
+    {
+      "epoch": 0.5455548202073682,
+      "grad_norm": 0.8361116647720337,
+      "learning_rate": 9.014421836668526e-05,
+      "loss": 0.3473,
+      "step": 7419
+    },
+    {
+      "epoch": 0.5456283550261048,
+      "grad_norm": 0.5974255204200745,
+      "learning_rate": 9.012051841699177e-05,
+      "loss": 0.3647,
+      "step": 7420
+    },
+    {
+      "epoch": 0.5457018898448416,
+      "grad_norm": 1.1558040380477905,
+      "learning_rate": 9.009681902767293e-05,
+      "loss": 0.3432,
+      "step": 7421
+    },
+    {
+      "epoch": 0.5457754246635782,
+      "grad_norm": 0.590263843536377,
+      "learning_rate": 9.007312020007297e-05,
+      "loss": 0.3413,
+      "step": 7422
+    },
+    {
+      "epoch": 0.5458489594823148,
+      "grad_norm": 0.784144401550293,
+      "learning_rate": 9.004942193553609e-05,
+      "loss": 0.3492,
+      "step": 7423
+    },
+    {
+      "epoch": 0.5459224943010516,
+      "grad_norm": 0.9970535635948181,
+      "learning_rate": 9.00257242354065e-05,
+      "loss": 0.3544,
+      "step": 7424
+    },
+    {
+      "epoch": 0.5459960291197882,
+      "grad_norm": 0.8652710914611816,
+      "learning_rate": 9.000202710102837e-05,
+      "loss": 0.3721,
+      "step": 7425
+    },
+    {
+      "epoch": 0.5460695639385249,
+      "grad_norm": 0.8079366087913513,
+      "learning_rate": 8.997833053374585e-05,
+      "loss": 0.3687,
+      "step": 7426
+    },
+    {
+      "epoch": 0.5461430987572615,
+      "grad_norm": 0.7265594005584717,
+      "learning_rate": 8.995463453490293e-05,
+      "loss": 0.332,
+      "step": 7427
+    },
+    {
+      "epoch": 0.5462166335759983,
+      "grad_norm": 0.5137746930122375,
+      "learning_rate": 8.993093910584378e-05,
+      "loss": 0.3657,
+      "step": 7428
+    },
+    {
+      "epoch": 0.5462901683947349,
+      "grad_norm": 0.9316918253898621,
+      "learning_rate": 8.990724424791239e-05,
+      "loss": 0.3523,
+      "step": 7429
+    },
+    {
+      "epoch": 0.5463637032134716,
+      "grad_norm": 0.5803287625312805,
+      "learning_rate": 8.988354996245273e-05,
+      "loss": 0.3723,
+      "step": 7430
+    },
+    {
+      "epoch": 0.5464372380322082,
+      "grad_norm": 1.2795037031173706,
+      "learning_rate": 8.985985625080882e-05,
+      "loss": 0.358,
+      "step": 7431
+    },
+    {
+      "epoch": 0.546510772850945,
+      "grad_norm": 0.8711779117584229,
+      "learning_rate": 8.983616311432456e-05,
+      "loss": 0.3887,
+      "step": 7432
+    },
+    {
+      "epoch": 0.5465843076696816,
+      "grad_norm": 0.9886442422866821,
+      "learning_rate": 8.981247055434384e-05,
+      "loss": 0.3604,
+      "step": 7433
+    },
+    {
+      "epoch": 0.5466578424884183,
+      "grad_norm": 1.6526824235916138,
+      "learning_rate": 8.978877857221054e-05,
+      "loss": 0.354,
+      "step": 7434
+    },
+    {
+      "epoch": 0.5467313773071549,
+      "grad_norm": 0.5535598993301392,
+      "learning_rate": 8.976508716926851e-05,
+      "loss": 0.3585,
+      "step": 7435
+    },
+    {
+      "epoch": 0.5468049121258917,
+      "grad_norm": 0.9633367657661438,
+      "learning_rate": 8.974139634686152e-05,
+      "loss": 0.3679,
+      "step": 7436
+    },
+    {
+      "epoch": 0.5468784469446283,
+      "grad_norm": 0.5443990230560303,
+      "learning_rate": 8.97177061063334e-05,
+      "loss": 0.3565,
+      "step": 7437
+    },
+    {
+      "epoch": 0.5469519817633649,
+      "grad_norm": 0.7609564661979675,
+      "learning_rate": 8.969401644902781e-05,
+      "loss": 0.3334,
+      "step": 7438
+    },
+    {
+      "epoch": 0.5470255165821016,
+      "grad_norm": 1.0502573251724243,
+      "learning_rate": 8.967032737628849e-05,
+      "loss": 0.37,
+      "step": 7439
+    },
+    {
+      "epoch": 0.5470990514008383,
+      "grad_norm": 0.6107509732246399,
+      "learning_rate": 8.964663888945908e-05,
+      "loss": 0.363,
+      "step": 7440
+    },
+    {
+      "epoch": 0.547172586219575,
+      "grad_norm": 0.6485869288444519,
+      "learning_rate": 8.962295098988325e-05,
+      "loss": 0.3715,
+      "step": 7441
+    },
+    {
+      "epoch": 0.5472461210383116,
+      "grad_norm": 0.9964205622673035,
+      "learning_rate": 8.95992636789046e-05,
+      "loss": 0.3761,
+      "step": 7442
+    },
+    {
+      "epoch": 0.5473196558570483,
+      "grad_norm": 0.6603268980979919,
+      "learning_rate": 8.957557695786668e-05,
+      "loss": 0.3518,
+      "step": 7443
+    },
+    {
+      "epoch": 0.547393190675785,
+      "grad_norm": 0.6426653861999512,
+      "learning_rate": 8.955189082811306e-05,
+      "loss": 0.376,
+      "step": 7444
+    },
+    {
+      "epoch": 0.5474667254945217,
+      "grad_norm": 0.5891479849815369,
+      "learning_rate": 8.952820529098719e-05,
+      "loss": 0.3545,
+      "step": 7445
+    },
+    {
+      "epoch": 0.5475402603132583,
+      "grad_norm": 0.5193463563919067,
+      "learning_rate": 8.950452034783256e-05,
+      "loss": 0.3613,
+      "step": 7446
+    },
+    {
+      "epoch": 0.547613795131995,
+      "grad_norm": 0.8937029242515564,
+      "learning_rate": 8.948083599999264e-05,
+      "loss": 0.3549,
+      "step": 7447
+    },
+    {
+      "epoch": 0.5476873299507317,
+      "grad_norm": 1.122784972190857,
+      "learning_rate": 8.945715224881079e-05,
+      "loss": 0.3461,
+      "step": 7448
+    },
+    {
+      "epoch": 0.5477608647694684,
+      "grad_norm": 0.49236220121383667,
+      "learning_rate": 8.943346909563043e-05,
+      "loss": 0.3708,
+      "step": 7449
+    },
+    {
+      "epoch": 0.547834399588205,
+      "grad_norm": 1.1984084844589233,
+      "learning_rate": 8.94097865417948e-05,
+      "loss": 0.3838,
+      "step": 7450
+    },
+    {
+      "epoch": 0.5479079344069416,
+      "grad_norm": 0.5854957103729248,
+      "learning_rate": 8.938610458864727e-05,
+      "loss": 0.3909,
+      "step": 7451
+    },
+    {
+      "epoch": 0.5479814692256784,
+      "grad_norm": 0.9999114871025085,
+      "learning_rate": 8.93624232375311e-05,
+      "loss": 0.3568,
+      "step": 7452
+    },
+    {
+      "epoch": 0.548055004044415,
+      "grad_norm": 1.0169800519943237,
+      "learning_rate": 8.933874248978949e-05,
+      "loss": 0.3605,
+      "step": 7453
+    },
+    {
+      "epoch": 0.5481285388631517,
+      "grad_norm": 0.5176861882209778,
+      "learning_rate": 8.931506234676567e-05,
+      "loss": 0.355,
+      "step": 7454
+    },
+    {
+      "epoch": 0.5482020736818883,
+      "grad_norm": 0.7803283333778381,
+      "learning_rate": 8.92913828098028e-05,
+      "loss": 0.352,
+      "step": 7455
+    },
+    {
+      "epoch": 0.5482756085006251,
+      "grad_norm": 1.18159818649292,
+      "learning_rate": 8.926770388024396e-05,
+      "loss": 0.3872,
+      "step": 7456
+    },
+    {
+      "epoch": 0.5483491433193617,
+      "grad_norm": 0.47773995995521545,
+      "learning_rate": 8.924402555943233e-05,
+      "loss": 0.347,
+      "step": 7457
+    },
+    {
+      "epoch": 0.5484226781380984,
+      "grad_norm": 1.1259874105453491,
+      "learning_rate": 8.922034784871091e-05,
+      "loss": 0.3602,
+      "step": 7458
+    },
+    {
+      "epoch": 0.548496212956835,
+      "grad_norm": 0.8386411070823669,
+      "learning_rate": 8.919667074942271e-05,
+      "loss": 0.3341,
+      "step": 7459
+    },
+    {
+      "epoch": 0.5485697477755718,
+      "grad_norm": 0.6605470776557922,
+      "learning_rate": 8.917299426291076e-05,
+      "loss": 0.33,
+      "step": 7460
+    },
+    {
+      "epoch": 0.5486432825943084,
+      "grad_norm": 1.0124870538711548,
+      "learning_rate": 8.914931839051803e-05,
+      "loss": 0.3689,
+      "step": 7461
+    },
+    {
+      "epoch": 0.5487168174130451,
+      "grad_norm": 0.4925127327442169,
+      "learning_rate": 8.91256431335874e-05,
+      "loss": 0.3841,
+      "step": 7462
+    },
+    {
+      "epoch": 0.5487903522317817,
+      "grad_norm": 0.7580205798149109,
+      "learning_rate": 8.910196849346172e-05,
+      "loss": 0.3607,
+      "step": 7463
+    },
+    {
+      "epoch": 0.5488638870505185,
+      "grad_norm": 0.4683283865451813,
+      "learning_rate": 8.907829447148393e-05,
+      "loss": 0.342,
+      "step": 7464
+    },
+    {
+      "epoch": 0.5489374218692551,
+      "grad_norm": 0.9019880890846252,
+      "learning_rate": 8.905462106899681e-05,
+      "loss": 0.3736,
+      "step": 7465
+    },
+    {
+      "epoch": 0.5490109566879917,
+      "grad_norm": 1.1143869161605835,
+      "learning_rate": 8.90309482873431e-05,
+      "loss": 0.3625,
+      "step": 7466
+    },
+    {
+      "epoch": 0.5490844915067284,
+      "grad_norm": 0.4371732473373413,
+      "learning_rate": 8.900727612786562e-05,
+      "loss": 0.356,
+      "step": 7467
+    },
+    {
+      "epoch": 0.5491580263254651,
+      "grad_norm": 1.2707005739212036,
+      "learning_rate": 8.898360459190703e-05,
+      "loss": 0.3373,
+      "step": 7468
+    },
+    {
+      "epoch": 0.5492315611442018,
+      "grad_norm": 0.7819347977638245,
+      "learning_rate": 8.895993368080997e-05,
+      "loss": 0.369,
+      "step": 7469
+    },
+    {
+      "epoch": 0.5493050959629384,
+      "grad_norm": 0.8135472536087036,
+      "learning_rate": 8.893626339591717e-05,
+      "loss": 0.3577,
+      "step": 7470
+    },
+    {
+      "epoch": 0.5493786307816751,
+      "grad_norm": 1.2324817180633545,
+      "learning_rate": 8.891259373857117e-05,
+      "loss": 0.3635,
+      "step": 7471
+    },
+    {
+      "epoch": 0.5494521656004118,
+      "grad_norm": 0.5538389086723328,
+      "learning_rate": 8.888892471011459e-05,
+      "loss": 0.3553,
+      "step": 7472
+    },
+    {
+      "epoch": 0.5495257004191485,
+      "grad_norm": 1.1725081205368042,
+      "learning_rate": 8.886525631188987e-05,
+      "loss": 0.3534,
+      "step": 7473
+    },
+    {
+      "epoch": 0.5495992352378851,
+      "grad_norm": 0.5791106820106506,
+      "learning_rate": 8.884158854523959e-05,
+      "loss": 0.367,
+      "step": 7474
+    },
+    {
+      "epoch": 0.5496727700566219,
+      "grad_norm": 1.4300869703292847,
+      "learning_rate": 8.881792141150618e-05,
+      "loss": 0.3741,
+      "step": 7475
+    },
+    {
+      "epoch": 0.5497463048753585,
+      "grad_norm": 0.668545663356781,
+      "learning_rate": 8.879425491203205e-05,
+      "loss": 0.4341,
+      "step": 7476
+    },
+    {
+      "epoch": 0.5498198396940952,
+      "grad_norm": 0.9724771976470947,
+      "learning_rate": 8.877058904815963e-05,
+      "loss": 0.3675,
+      "step": 7477
+    },
+    {
+      "epoch": 0.5498933745128318,
+      "grad_norm": 0.667543351650238,
+      "learning_rate": 8.874692382123126e-05,
+      "loss": 0.4104,
+      "step": 7478
+    },
+    {
+      "epoch": 0.5499669093315684,
+      "grad_norm": 0.6810120344161987,
+      "learning_rate": 8.872325923258921e-05,
+      "loss": 0.3564,
+      "step": 7479
+    },
+    {
+      "epoch": 0.5500404441503052,
+      "grad_norm": 0.8047006130218506,
+      "learning_rate": 8.869959528357582e-05,
+      "loss": 0.3537,
+      "step": 7480
+    },
+    {
+      "epoch": 0.5501139789690418,
+      "grad_norm": 0.5758931040763855,
+      "learning_rate": 8.867593197553331e-05,
+      "loss": 0.34,
+      "step": 7481
+    },
+    {
+      "epoch": 0.5501875137877785,
+      "grad_norm": 0.7134798169136047,
+      "learning_rate": 8.865226930980389e-05,
+      "loss": 0.3587,
+      "step": 7482
+    },
+    {
+      "epoch": 0.5502610486065151,
+      "grad_norm": 0.645654559135437,
+      "learning_rate": 8.862860728772973e-05,
+      "loss": 0.3743,
+      "step": 7483
+    },
+    {
+      "epoch": 0.5503345834252519,
+      "grad_norm": 0.7507840394973755,
+      "learning_rate": 8.860494591065298e-05,
+      "loss": 0.3511,
+      "step": 7484
+    },
+    {
+      "epoch": 0.5504081182439885,
+      "grad_norm": 1.050045371055603,
+      "learning_rate": 8.858128517991572e-05,
+      "loss": 0.3734,
+      "step": 7485
+    },
+    {
+      "epoch": 0.5504816530627252,
+      "grad_norm": 0.8864396214485168,
+      "learning_rate": 8.855762509685999e-05,
+      "loss": 0.3893,
+      "step": 7486
+    },
+    {
+      "epoch": 0.5505551878814618,
+      "grad_norm": 0.5032808184623718,
+      "learning_rate": 8.853396566282786e-05,
+      "loss": 0.3662,
+      "step": 7487
+    },
+    {
+      "epoch": 0.5506287227001986,
+      "grad_norm": 0.7008019089698792,
+      "learning_rate": 8.851030687916129e-05,
+      "loss": 0.3346,
+      "step": 7488
+    },
+    {
+      "epoch": 0.5507022575189352,
+      "grad_norm": 0.8720541596412659,
+      "learning_rate": 8.848664874720223e-05,
+      "loss": 0.3499,
+      "step": 7489
+    },
+    {
+      "epoch": 0.5507757923376719,
+      "grad_norm": 0.645379364490509,
+      "learning_rate": 8.846299126829262e-05,
+      "loss": 0.3795,
+      "step": 7490
+    },
+    {
+      "epoch": 0.5508493271564086,
+      "grad_norm": 0.7609565854072571,
+      "learning_rate": 8.843933444377431e-05,
+      "loss": 0.3784,
+      "step": 7491
+    },
+    {
+      "epoch": 0.5509228619751453,
+      "grad_norm": 0.9389234781265259,
+      "learning_rate": 8.841567827498914e-05,
+      "loss": 0.3391,
+      "step": 7492
+    },
+    {
+      "epoch": 0.5509963967938819,
+      "grad_norm": 0.6441373229026794,
+      "learning_rate": 8.839202276327895e-05,
+      "loss": 0.3666,
+      "step": 7493
+    },
+    {
+      "epoch": 0.5510699316126185,
+      "grad_norm": 0.5822252035140991,
+      "learning_rate": 8.836836790998547e-05,
+      "loss": 0.3321,
+      "step": 7494
+    },
+    {
+      "epoch": 0.5511434664313553,
+      "grad_norm": 0.4692037105560303,
+      "learning_rate": 8.834471371645042e-05,
+      "loss": 0.3364,
+      "step": 7495
+    },
+    {
+      "epoch": 0.5512170012500919,
+      "grad_norm": 0.6292122602462769,
+      "learning_rate": 8.832106018401556e-05,
+      "loss": 0.3554,
+      "step": 7496
+    },
+    {
+      "epoch": 0.5512905360688286,
+      "grad_norm": 0.6770101189613342,
+      "learning_rate": 8.829740731402245e-05,
+      "loss": 0.3702,
+      "step": 7497
+    },
+    {
+      "epoch": 0.5513640708875652,
+      "grad_norm": 0.7823976278305054,
+      "learning_rate": 8.827375510781276e-05,
+      "loss": 0.3526,
+      "step": 7498
+    },
+    {
+      "epoch": 0.551437605706302,
+      "grad_norm": 0.6034998297691345,
+      "learning_rate": 8.825010356672803e-05,
+      "loss": 0.35,
+      "step": 7499
+    },
+    {
+      "epoch": 0.5515111405250386,
+      "grad_norm": 0.5570348501205444,
+      "learning_rate": 8.822645269210986e-05,
+      "loss": 0.3562,
+      "step": 7500
+    },
+    {
+      "epoch": 0.5515846753437753,
+      "grad_norm": 0.6464783549308777,
+      "learning_rate": 8.820280248529971e-05,
+      "loss": 0.3603,
+      "step": 7501
+    },
+    {
+      "epoch": 0.5516582101625119,
+      "grad_norm": 0.5838094353675842,
+      "learning_rate": 8.817915294763903e-05,
+      "loss": 0.3547,
+      "step": 7502
+    },
+    {
+      "epoch": 0.5517317449812487,
+      "grad_norm": 0.788020133972168,
+      "learning_rate": 8.815550408046929e-05,
+      "loss": 0.3456,
+      "step": 7503
+    },
+    {
+      "epoch": 0.5518052797999853,
+      "grad_norm": 0.5409361720085144,
+      "learning_rate": 8.813185588513186e-05,
+      "loss": 0.3567,
+      "step": 7504
+    },
+    {
+      "epoch": 0.551878814618722,
+      "grad_norm": 0.7336738705635071,
+      "learning_rate": 8.810820836296806e-05,
+      "loss": 0.3647,
+      "step": 7505
+    },
+    {
+      "epoch": 0.5519523494374586,
+      "grad_norm": 0.7396862506866455,
+      "learning_rate": 8.808456151531927e-05,
+      "loss": 0.3483,
+      "step": 7506
+    },
+    {
+      "epoch": 0.5520258842561953,
+      "grad_norm": 0.534946858882904,
+      "learning_rate": 8.806091534352674e-05,
+      "loss": 0.3853,
+      "step": 7507
+    },
+    {
+      "epoch": 0.552099419074932,
+      "grad_norm": 0.660476565361023,
+      "learning_rate": 8.803726984893166e-05,
+      "loss": 0.3964,
+      "step": 7508
+    },
+    {
+      "epoch": 0.5521729538936686,
+      "grad_norm": 0.6349261403083801,
+      "learning_rate": 8.801362503287524e-05,
+      "loss": 0.3739,
+      "step": 7509
+    },
+    {
+      "epoch": 0.5522464887124053,
+      "grad_norm": 0.5176954865455627,
+      "learning_rate": 8.798998089669869e-05,
+      "loss": 0.3895,
+      "step": 7510
+    },
+    {
+      "epoch": 0.552320023531142,
+      "grad_norm": 0.6088402271270752,
+      "learning_rate": 8.796633744174309e-05,
+      "loss": 0.3455,
+      "step": 7511
+    },
+    {
+      "epoch": 0.5523935583498787,
+      "grad_norm": 0.7256307601928711,
+      "learning_rate": 8.794269466934953e-05,
+      "loss": 0.3703,
+      "step": 7512
+    },
+    {
+      "epoch": 0.5524670931686153,
+      "grad_norm": 0.5978339314460754,
+      "learning_rate": 8.791905258085906e-05,
+      "loss": 0.3798,
+      "step": 7513
+    },
+    {
+      "epoch": 0.552540627987352,
+      "grad_norm": 0.5557090640068054,
+      "learning_rate": 8.789541117761267e-05,
+      "loss": 0.3395,
+      "step": 7514
+    },
+    {
+      "epoch": 0.5526141628060887,
+      "grad_norm": 0.8600425720214844,
+      "learning_rate": 8.78717704609513e-05,
+      "loss": 0.3777,
+      "step": 7515
+    },
+    {
+      "epoch": 0.5526876976248254,
+      "grad_norm": 0.45477381348609924,
+      "learning_rate": 8.784813043221595e-05,
+      "loss": 0.3782,
+      "step": 7516
+    },
+    {
+      "epoch": 0.552761232443562,
+      "grad_norm": 0.8955132961273193,
+      "learning_rate": 8.782449109274746e-05,
+      "loss": 0.3788,
+      "step": 7517
+    },
+    {
+      "epoch": 0.5528347672622987,
+      "grad_norm": 0.6400097012519836,
+      "learning_rate": 8.780085244388666e-05,
+      "loss": 0.3929,
+      "step": 7518
+    },
+    {
+      "epoch": 0.5529083020810354,
+      "grad_norm": 0.7344452142715454,
+      "learning_rate": 8.777721448697443e-05,
+      "loss": 0.3933,
+      "step": 7519
+    },
+    {
+      "epoch": 0.5529818368997721,
+      "grad_norm": 1.0859845876693726,
+      "learning_rate": 8.775357722335149e-05,
+      "loss": 0.3713,
+      "step": 7520
+    },
+    {
+      "epoch": 0.5530553717185087,
+      "grad_norm": 0.7930981516838074,
+      "learning_rate": 8.772994065435855e-05,
+      "loss": 0.364,
+      "step": 7521
+    },
+    {
+      "epoch": 0.5531289065372453,
+      "grad_norm": 0.6149868369102478,
+      "learning_rate": 8.770630478133631e-05,
+      "loss": 0.3581,
+      "step": 7522
+    },
+    {
+      "epoch": 0.5532024413559821,
+      "grad_norm": 0.8597185015678406,
+      "learning_rate": 8.768266960562546e-05,
+      "loss": 0.3778,
+      "step": 7523
+    },
+    {
+      "epoch": 0.5532759761747187,
+      "grad_norm": 0.4876434803009033,
+      "learning_rate": 8.765903512856658e-05,
+      "loss": 0.3765,
+      "step": 7524
+    },
+    {
+      "epoch": 0.5533495109934554,
+      "grad_norm": 0.8039893507957458,
+      "learning_rate": 8.763540135150024e-05,
+      "loss": 0.3687,
+      "step": 7525
+    },
+    {
+      "epoch": 0.553423045812192,
+      "grad_norm": 0.6121858954429626,
+      "learning_rate": 8.7611768275767e-05,
+      "loss": 0.3637,
+      "step": 7526
+    },
+    {
+      "epoch": 0.5534965806309288,
+      "grad_norm": 0.7177013158798218,
+      "learning_rate": 8.758813590270734e-05,
+      "loss": 0.3676,
+      "step": 7527
+    },
+    {
+      "epoch": 0.5535701154496654,
+      "grad_norm": 1.1407068967819214,
+      "learning_rate": 8.756450423366169e-05,
+      "loss": 0.3548,
+      "step": 7528
+    },
+    {
+      "epoch": 0.5536436502684021,
+      "grad_norm": 0.6295998096466064,
+      "learning_rate": 8.754087326997052e-05,
+      "loss": 0.3671,
+      "step": 7529
+    },
+    {
+      "epoch": 0.5537171850871387,
+      "grad_norm": 0.9496809840202332,
+      "learning_rate": 8.751724301297413e-05,
+      "loss": 0.355,
+      "step": 7530
+    },
+    {
+      "epoch": 0.5537907199058755,
+      "grad_norm": 1.0650856494903564,
+      "learning_rate": 8.749361346401294e-05,
+      "loss": 0.3423,
+      "step": 7531
+    },
+    {
+      "epoch": 0.5538642547246121,
+      "grad_norm": 0.7818108201026917,
+      "learning_rate": 8.746998462442714e-05,
+      "loss": 0.3929,
+      "step": 7532
+    },
+    {
+      "epoch": 0.5539377895433488,
+      "grad_norm": 1.0893265008926392,
+      "learning_rate": 8.744635649555706e-05,
+      "loss": 0.3392,
+      "step": 7533
+    },
+    {
+      "epoch": 0.5540113243620854,
+      "grad_norm": 0.6262742877006531,
+      "learning_rate": 8.74227290787429e-05,
+      "loss": 0.3581,
+      "step": 7534
+    },
+    {
+      "epoch": 0.5540848591808221,
+      "grad_norm": 0.8348832726478577,
+      "learning_rate": 8.739910237532478e-05,
+      "loss": 0.3766,
+      "step": 7535
+    },
+    {
+      "epoch": 0.5541583939995588,
+      "grad_norm": 0.5223789215087891,
+      "learning_rate": 8.73754763866429e-05,
+      "loss": 0.3582,
+      "step": 7536
+    },
+    {
+      "epoch": 0.5542319288182954,
+      "grad_norm": 0.7410044074058533,
+      "learning_rate": 8.735185111403732e-05,
+      "loss": 0.3705,
+      "step": 7537
+    },
+    {
+      "epoch": 0.5543054636370321,
+      "grad_norm": 0.7146099209785461,
+      "learning_rate": 8.732822655884807e-05,
+      "loss": 0.3565,
+      "step": 7538
+    },
+    {
+      "epoch": 0.5543789984557688,
+      "grad_norm": 0.67421555519104,
+      "learning_rate": 8.73046027224152e-05,
+      "loss": 0.3599,
+      "step": 7539
+    },
+    {
+      "epoch": 0.5544525332745055,
+      "grad_norm": 0.47412458062171936,
+      "learning_rate": 8.728097960607869e-05,
+      "loss": 0.3409,
+      "step": 7540
+    },
+    {
+      "epoch": 0.5545260680932421,
+      "grad_norm": 0.6019271612167358,
+      "learning_rate": 8.725735721117838e-05,
+      "loss": 0.3949,
+      "step": 7541
+    },
+    {
+      "epoch": 0.5545996029119789,
+      "grad_norm": 0.5527322292327881,
+      "learning_rate": 8.723373553905429e-05,
+      "loss": 0.3629,
+      "step": 7542
+    },
+    {
+      "epoch": 0.5546731377307155,
+      "grad_norm": 0.6559256911277771,
+      "learning_rate": 8.721011459104615e-05,
+      "loss": 0.3335,
+      "step": 7543
+    },
+    {
+      "epoch": 0.5547466725494522,
+      "grad_norm": 0.9451995491981506,
+      "learning_rate": 8.718649436849382e-05,
+      "loss": 0.3754,
+      "step": 7544
+    },
+    {
+      "epoch": 0.5548202073681888,
+      "grad_norm": 0.49379533529281616,
+      "learning_rate": 8.716287487273702e-05,
+      "loss": 0.3783,
+      "step": 7545
+    },
+    {
+      "epoch": 0.5548937421869256,
+      "grad_norm": 0.889846682548523,
+      "learning_rate": 8.713925610511554e-05,
+      "loss": 0.4078,
+      "step": 7546
+    },
+    {
+      "epoch": 0.5549672770056622,
+      "grad_norm": 1.1718668937683105,
+      "learning_rate": 8.711563806696902e-05,
+      "loss": 0.3499,
+      "step": 7547
+    },
+    {
+      "epoch": 0.5550408118243989,
+      "grad_norm": 0.6523907780647278,
+      "learning_rate": 8.709202075963708e-05,
+      "loss": 0.3662,
+      "step": 7548
+    },
+    {
+      "epoch": 0.5551143466431355,
+      "grad_norm": 0.8201984763145447,
+      "learning_rate": 8.706840418445937e-05,
+      "loss": 0.3583,
+      "step": 7549
+    },
+    {
+      "epoch": 0.5551878814618721,
+      "grad_norm": 0.9957278370857239,
+      "learning_rate": 8.704478834277543e-05,
+      "loss": 0.3497,
+      "step": 7550
+    },
+    {
+      "epoch": 0.5552614162806089,
+      "grad_norm": 0.5295627117156982,
+      "learning_rate": 8.702117323592473e-05,
+      "loss": 0.4004,
+      "step": 7551
+    },
+    {
+      "epoch": 0.5553349510993455,
+      "grad_norm": 0.7554019689559937,
+      "learning_rate": 8.699755886524681e-05,
+      "loss": 0.3897,
+      "step": 7552
+    },
+    {
+      "epoch": 0.5554084859180822,
+      "grad_norm": 0.8301599621772766,
+      "learning_rate": 8.697394523208108e-05,
+      "loss": 0.3579,
+      "step": 7553
+    },
+    {
+      "epoch": 0.5554820207368188,
+      "grad_norm": 0.4926668405532837,
+      "learning_rate": 8.695033233776694e-05,
+      "loss": 0.38,
+      "step": 7554
+    },
+    {
+      "epoch": 0.5555555555555556,
+      "grad_norm": 1.2396469116210938,
+      "learning_rate": 8.692672018364368e-05,
+      "loss": 0.4085,
+      "step": 7555
+    },
+    {
+      "epoch": 0.5556290903742922,
+      "grad_norm": 1.0825728178024292,
+      "learning_rate": 8.690310877105066e-05,
+      "loss": 0.3615,
+      "step": 7556
+    },
+    {
+      "epoch": 0.5557026251930289,
+      "grad_norm": 0.5735538005828857,
+      "learning_rate": 8.687949810132714e-05,
+      "loss": 0.3521,
+      "step": 7557
+    },
+    {
+      "epoch": 0.5557761600117656,
+      "grad_norm": 1.1619657278060913,
+      "learning_rate": 8.685588817581231e-05,
+      "loss": 0.356,
+      "step": 7558
+    },
+    {
+      "epoch": 0.5558496948305023,
+      "grad_norm": 0.5733131170272827,
+      "learning_rate": 8.683227899584541e-05,
+      "loss": 0.3652,
+      "step": 7559
+    },
+    {
+      "epoch": 0.5559232296492389,
+      "grad_norm": 0.9129968285560608,
+      "learning_rate": 8.680867056276551e-05,
+      "loss": 0.3761,
+      "step": 7560
+    },
+    {
+      "epoch": 0.5559967644679756,
+      "grad_norm": 1.0912611484527588,
+      "learning_rate": 8.678506287791174e-05,
+      "loss": 0.3405,
+      "step": 7561
+    },
+    {
+      "epoch": 0.5560702992867123,
+      "grad_norm": 0.8040217757225037,
+      "learning_rate": 8.676145594262317e-05,
+      "loss": 0.3535,
+      "step": 7562
+    },
+    {
+      "epoch": 0.5561438341054489,
+      "grad_norm": 0.5530813336372375,
+      "learning_rate": 8.673784975823878e-05,
+      "loss": 0.3627,
+      "step": 7563
+    },
+    {
+      "epoch": 0.5562173689241856,
+      "grad_norm": 0.7472689747810364,
+      "learning_rate": 8.671424432609753e-05,
+      "loss": 0.3564,
+      "step": 7564
+    },
+    {
+      "epoch": 0.5562909037429222,
+      "grad_norm": 0.5645768642425537,
+      "learning_rate": 8.66906396475384e-05,
+      "loss": 0.3738,
+      "step": 7565
+    },
+    {
+      "epoch": 0.556364438561659,
+      "grad_norm": 0.6752889752388,
+      "learning_rate": 8.666703572390025e-05,
+      "loss": 0.3552,
+      "step": 7566
+    },
+    {
+      "epoch": 0.5564379733803956,
+      "grad_norm": 0.8487634658813477,
+      "learning_rate": 8.664343255652188e-05,
+      "loss": 0.3479,
+      "step": 7567
+    },
+    {
+      "epoch": 0.5565115081991323,
+      "grad_norm": 0.6432210803031921,
+      "learning_rate": 8.661983014674207e-05,
+      "loss": 0.3421,
+      "step": 7568
+    },
+    {
+      "epoch": 0.5565850430178689,
+      "grad_norm": 0.652496337890625,
+      "learning_rate": 8.659622849589966e-05,
+      "loss": 0.3456,
+      "step": 7569
+    },
+    {
+      "epoch": 0.5566585778366057,
+      "grad_norm": 0.9832175374031067,
+      "learning_rate": 8.657262760533331e-05,
+      "loss": 0.3413,
+      "step": 7570
+    },
+    {
+      "epoch": 0.5567321126553423,
+      "grad_norm": 0.6801847815513611,
+      "learning_rate": 8.654902747638167e-05,
+      "loss": 0.3547,
+      "step": 7571
+    },
+    {
+      "epoch": 0.556805647474079,
+      "grad_norm": 0.5266603827476501,
+      "learning_rate": 8.65254281103834e-05,
+      "loss": 0.3495,
+      "step": 7572
+    },
+    {
+      "epoch": 0.5568791822928156,
+      "grad_norm": 0.6645089387893677,
+      "learning_rate": 8.650182950867708e-05,
+      "loss": 0.3498,
+      "step": 7573
+    },
+    {
+      "epoch": 0.5569527171115524,
+      "grad_norm": 0.47905290126800537,
+      "learning_rate": 8.647823167260121e-05,
+      "loss": 0.3516,
+      "step": 7574
+    },
+    {
+      "epoch": 0.557026251930289,
+      "grad_norm": 0.5652728080749512,
+      "learning_rate": 8.645463460349433e-05,
+      "loss": 0.3777,
+      "step": 7575
+    },
+    {
+      "epoch": 0.5570997867490257,
+      "grad_norm": 0.5573289394378662,
+      "learning_rate": 8.643103830269488e-05,
+      "loss": 0.3466,
+      "step": 7576
+    },
+    {
+      "epoch": 0.5571733215677623,
+      "grad_norm": 0.5104591846466064,
+      "learning_rate": 8.640744277154122e-05,
+      "loss": 0.4002,
+      "step": 7577
+    },
+    {
+      "epoch": 0.557246856386499,
+      "grad_norm": 0.5952361822128296,
+      "learning_rate": 8.63838480113718e-05,
+      "loss": 0.3461,
+      "step": 7578
+    },
+    {
+      "epoch": 0.5573203912052357,
+      "grad_norm": 0.5801440477371216,
+      "learning_rate": 8.636025402352486e-05,
+      "loss": 0.371,
+      "step": 7579
+    },
+    {
+      "epoch": 0.5573939260239723,
+      "grad_norm": 0.5187621116638184,
+      "learning_rate": 8.633666080933873e-05,
+      "loss": 0.3505,
+      "step": 7580
+    },
+    {
+      "epoch": 0.557467460842709,
+      "grad_norm": 0.7647689580917358,
+      "learning_rate": 8.631306837015158e-05,
+      "loss": 0.3506,
+      "step": 7581
+    },
+    {
+      "epoch": 0.5575409956614457,
+      "grad_norm": 0.6651950478553772,
+      "learning_rate": 8.628947670730166e-05,
+      "loss": 0.346,
+      "step": 7582
+    },
+    {
+      "epoch": 0.5576145304801824,
+      "grad_norm": 0.6234857439994812,
+      "learning_rate": 8.626588582212709e-05,
+      "loss": 0.4122,
+      "step": 7583
+    },
+    {
+      "epoch": 0.557688065298919,
+      "grad_norm": 0.9035173654556274,
+      "learning_rate": 8.624229571596595e-05,
+      "loss": 0.3591,
+      "step": 7584
+    },
+    {
+      "epoch": 0.5577616001176557,
+      "grad_norm": 0.9854403138160706,
+      "learning_rate": 8.621870639015634e-05,
+      "loss": 0.3504,
+      "step": 7585
+    },
+    {
+      "epoch": 0.5578351349363924,
+      "grad_norm": 0.6457065343856812,
+      "learning_rate": 8.619511784603624e-05,
+      "loss": 0.3557,
+      "step": 7586
+    },
+    {
+      "epoch": 0.5579086697551291,
+      "grad_norm": 0.7090465426445007,
+      "learning_rate": 8.61715300849436e-05,
+      "loss": 0.3636,
+      "step": 7587
+    },
+    {
+      "epoch": 0.5579822045738657,
+      "grad_norm": 1.127787709236145,
+      "learning_rate": 8.614794310821641e-05,
+      "loss": 0.3559,
+      "step": 7588
+    },
+    {
+      "epoch": 0.5580557393926024,
+      "grad_norm": 0.47920238971710205,
+      "learning_rate": 8.612435691719251e-05,
+      "loss": 0.3359,
+      "step": 7589
+    },
+    {
+      "epoch": 0.5581292742113391,
+      "grad_norm": 0.8926876783370972,
+      "learning_rate": 8.610077151320972e-05,
+      "loss": 0.3656,
+      "step": 7590
+    },
+    {
+      "epoch": 0.5582028090300757,
+      "grad_norm": 1.0168640613555908,
+      "learning_rate": 8.60771868976058e-05,
+      "loss": 0.365,
+      "step": 7591
+    },
+    {
+      "epoch": 0.5582763438488124,
+      "grad_norm": 0.4473351240158081,
+      "learning_rate": 8.605360307171856e-05,
+      "loss": 0.3756,
+      "step": 7592
+    },
+    {
+      "epoch": 0.558349878667549,
+      "grad_norm": 1.466624140739441,
+      "learning_rate": 8.603002003688569e-05,
+      "loss": 0.3854,
+      "step": 7593
+    },
+    {
+      "epoch": 0.5584234134862858,
+      "grad_norm": 0.47537651658058167,
+      "learning_rate": 8.600643779444478e-05,
+      "loss": 0.37,
+      "step": 7594
+    },
+    {
+      "epoch": 0.5584969483050224,
+      "grad_norm": 1.436928391456604,
+      "learning_rate": 8.598285634573352e-05,
+      "loss": 0.3864,
+      "step": 7595
+    },
+    {
+      "epoch": 0.5585704831237591,
+      "grad_norm": 0.713569700717926,
+      "learning_rate": 8.595927569208943e-05,
+      "loss": 0.3427,
+      "step": 7596
+    },
+    {
+      "epoch": 0.5586440179424957,
+      "grad_norm": 0.6965517401695251,
+      "learning_rate": 8.593569583485004e-05,
+      "loss": 0.3666,
+      "step": 7597
+    },
+    {
+      "epoch": 0.5587175527612325,
+      "grad_norm": 1.2142906188964844,
+      "learning_rate": 8.591211677535283e-05,
+      "loss": 0.3803,
+      "step": 7598
+    },
+    {
+      "epoch": 0.5587910875799691,
+      "grad_norm": 0.6942252516746521,
+      "learning_rate": 8.588853851493523e-05,
+      "loss": 0.3685,
+      "step": 7599
+    },
+    {
+      "epoch": 0.5588646223987058,
+      "grad_norm": 0.6356267333030701,
+      "learning_rate": 8.586496105493458e-05,
+      "loss": 0.3913,
+      "step": 7600
+    },
+    {
+      "epoch": 0.5589381572174424,
+      "grad_norm": 0.8700408339500427,
+      "learning_rate": 8.584138439668833e-05,
+      "loss": 0.3895,
+      "step": 7601
+    },
+    {
+      "epoch": 0.5590116920361792,
+      "grad_norm": 0.6161202192306519,
+      "learning_rate": 8.581780854153365e-05,
+      "loss": 0.4039,
+      "step": 7602
+    },
+    {
+      "epoch": 0.5590852268549158,
+      "grad_norm": 0.8475890755653381,
+      "learning_rate": 8.579423349080785e-05,
+      "loss": 0.3744,
+      "step": 7603
+    },
+    {
+      "epoch": 0.5591587616736525,
+      "grad_norm": 0.969566285610199,
+      "learning_rate": 8.57706592458481e-05,
+      "loss": 0.3822,
+      "step": 7604
+    },
+    {
+      "epoch": 0.5592322964923891,
+      "grad_norm": 0.4917827546596527,
+      "learning_rate": 8.574708580799159e-05,
+      "loss": 0.3704,
+      "step": 7605
+    },
+    {
+      "epoch": 0.5593058313111258,
+      "grad_norm": 0.6723912954330444,
+      "learning_rate": 8.57235131785754e-05,
+      "loss": 0.367,
+      "step": 7606
+    },
+    {
+      "epoch": 0.5593793661298625,
+      "grad_norm": 0.530598521232605,
+      "learning_rate": 8.569994135893662e-05,
+      "loss": 0.3512,
+      "step": 7607
+    },
+    {
+      "epoch": 0.5594529009485991,
+      "grad_norm": 0.6779893040657043,
+      "learning_rate": 8.567637035041225e-05,
+      "loss": 0.3479,
+      "step": 7608
+    },
+    {
+      "epoch": 0.5595264357673359,
+      "grad_norm": 0.5277788043022156,
+      "learning_rate": 8.56528001543393e-05,
+      "loss": 0.3634,
+      "step": 7609
+    },
+    {
+      "epoch": 0.5595999705860725,
+      "grad_norm": 0.635797381401062,
+      "learning_rate": 8.562923077205463e-05,
+      "loss": 0.3647,
+      "step": 7610
+    },
+    {
+      "epoch": 0.5596735054048092,
+      "grad_norm": 0.7049433588981628,
+      "learning_rate": 8.560566220489517e-05,
+      "loss": 0.3511,
+      "step": 7611
+    },
+    {
+      "epoch": 0.5597470402235458,
+      "grad_norm": 0.6268551349639893,
+      "learning_rate": 8.558209445419771e-05,
+      "loss": 0.3476,
+      "step": 7612
+    },
+    {
+      "epoch": 0.5598205750422826,
+      "grad_norm": 0.4966408312320709,
+      "learning_rate": 8.555852752129915e-05,
+      "loss": 0.3601,
+      "step": 7613
+    },
+    {
+      "epoch": 0.5598941098610192,
+      "grad_norm": 0.48965105414390564,
+      "learning_rate": 8.553496140753609e-05,
+      "loss": 0.3511,
+      "step": 7614
+    },
+    {
+      "epoch": 0.5599676446797559,
+      "grad_norm": 0.5252439379692078,
+      "learning_rate": 8.55113961142453e-05,
+      "loss": 0.3609,
+      "step": 7615
+    },
+    {
+      "epoch": 0.5600411794984925,
+      "grad_norm": 0.6358673572540283,
+      "learning_rate": 8.54878316427634e-05,
+      "loss": 0.3386,
+      "step": 7616
+    },
+    {
+      "epoch": 0.5601147143172293,
+      "grad_norm": 0.4905989170074463,
+      "learning_rate": 8.5464267994427e-05,
+      "loss": 0.3638,
+      "step": 7617
+    },
+    {
+      "epoch": 0.5601882491359659,
+      "grad_norm": 0.6205123662948608,
+      "learning_rate": 8.544070517057267e-05,
+      "loss": 0.3648,
+      "step": 7618
+    },
+    {
+      "epoch": 0.5602617839547025,
+      "grad_norm": 0.48021870851516724,
+      "learning_rate": 8.541714317253692e-05,
+      "loss": 0.3556,
+      "step": 7619
+    },
+    {
+      "epoch": 0.5603353187734392,
+      "grad_norm": 0.7535849809646606,
+      "learning_rate": 8.539358200165617e-05,
+      "loss": 0.3709,
+      "step": 7620
+    },
+    {
+      "epoch": 0.5604088535921758,
+      "grad_norm": 0.5359964370727539,
+      "learning_rate": 8.537002165926689e-05,
+      "loss": 0.3391,
+      "step": 7621
+    },
+    {
+      "epoch": 0.5604823884109126,
+      "grad_norm": 1.1589163541793823,
+      "learning_rate": 8.53464621467054e-05,
+      "loss": 0.3841,
+      "step": 7622
+    },
+    {
+      "epoch": 0.5605559232296492,
+      "grad_norm": 0.48106229305267334,
+      "learning_rate": 8.532290346530804e-05,
+      "loss": 0.3465,
+      "step": 7623
+    },
+    {
+      "epoch": 0.5606294580483859,
+      "grad_norm": 1.0559041500091553,
+      "learning_rate": 8.529934561641113e-05,
+      "loss": 0.3385,
+      "step": 7624
+    },
+    {
+      "epoch": 0.5607029928671226,
+      "grad_norm": 1.1596314907073975,
+      "learning_rate": 8.527578860135083e-05,
+      "loss": 0.3331,
+      "step": 7625
+    },
+    {
+      "epoch": 0.5607765276858593,
+      "grad_norm": 0.6625240445137024,
+      "learning_rate": 8.525223242146334e-05,
+      "loss": 0.3652,
+      "step": 7626
+    },
+    {
+      "epoch": 0.5608500625045959,
+      "grad_norm": 0.8552000522613525,
+      "learning_rate": 8.522867707808475e-05,
+      "loss": 0.3592,
+      "step": 7627
+    },
+    {
+      "epoch": 0.5609235973233326,
+      "grad_norm": 0.7236054539680481,
+      "learning_rate": 8.520512257255122e-05,
+      "loss": 0.3628,
+      "step": 7628
+    },
+    {
+      "epoch": 0.5609971321420693,
+      "grad_norm": 0.6355282664299011,
+      "learning_rate": 8.518156890619876e-05,
+      "loss": 0.3688,
+      "step": 7629
+    },
+    {
+      "epoch": 0.561070666960806,
+      "grad_norm": 1.2864683866500854,
+      "learning_rate": 8.515801608036332e-05,
+      "loss": 0.3859,
+      "step": 7630
+    },
+    {
+      "epoch": 0.5611442017795426,
+      "grad_norm": 0.6515757441520691,
+      "learning_rate": 8.513446409638089e-05,
+      "loss": 0.3527,
+      "step": 7631
+    },
+    {
+      "epoch": 0.5612177365982793,
+      "grad_norm": 1.7298463582992554,
+      "learning_rate": 8.511091295558735e-05,
+      "loss": 0.3414,
+      "step": 7632
+    },
+    {
+      "epoch": 0.561291271417016,
+      "grad_norm": 1.0760852098464966,
+      "learning_rate": 8.508736265931852e-05,
+      "loss": 0.3675,
+      "step": 7633
+    },
+    {
+      "epoch": 0.5613648062357526,
+      "grad_norm": 1.742011547088623,
+      "learning_rate": 8.506381320891023e-05,
+      "loss": 0.3289,
+      "step": 7634
+    },
+    {
+      "epoch": 0.5614383410544893,
+      "grad_norm": 2.2245733737945557,
+      "learning_rate": 8.50402646056982e-05,
+      "loss": 0.3449,
+      "step": 7635
+    },
+    {
+      "epoch": 0.5615118758732259,
+      "grad_norm": 0.9526073932647705,
+      "learning_rate": 8.501671685101822e-05,
+      "loss": 0.336,
+      "step": 7636
+    },
+    {
+      "epoch": 0.5615854106919627,
+      "grad_norm": 1.3758060932159424,
+      "learning_rate": 8.49931699462058e-05,
+      "loss": 0.3562,
+      "step": 7637
+    },
+    {
+      "epoch": 0.5616589455106993,
+      "grad_norm": 1.201037049293518,
+      "learning_rate": 8.496962389259664e-05,
+      "loss": 0.3434,
+      "step": 7638
+    },
+    {
+      "epoch": 0.561732480329436,
+      "grad_norm": 0.610124945640564,
+      "learning_rate": 8.494607869152627e-05,
+      "loss": 0.3513,
+      "step": 7639
+    },
+    {
+      "epoch": 0.5618060151481726,
+      "grad_norm": 1.3186372518539429,
+      "learning_rate": 8.49225343443302e-05,
+      "loss": 0.3336,
+      "step": 7640
+    },
+    {
+      "epoch": 0.5618795499669094,
+      "grad_norm": 0.5040399432182312,
+      "learning_rate": 8.489899085234391e-05,
+      "loss": 0.3563,
+      "step": 7641
+    },
+    {
+      "epoch": 0.561953084785646,
+      "grad_norm": 1.0607962608337402,
+      "learning_rate": 8.487544821690279e-05,
+      "loss": 0.3587,
+      "step": 7642
+    },
+    {
+      "epoch": 0.5620266196043827,
+      "grad_norm": 0.6922910809516907,
+      "learning_rate": 8.485190643934218e-05,
+      "loss": 0.3399,
+      "step": 7643
+    },
+    {
+      "epoch": 0.5621001544231193,
+      "grad_norm": 0.9217420220375061,
+      "learning_rate": 8.482836552099746e-05,
+      "loss": 0.3438,
+      "step": 7644
+    },
+    {
+      "epoch": 0.5621736892418561,
+      "grad_norm": 1.4364522695541382,
+      "learning_rate": 8.480482546320385e-05,
+      "loss": 0.3358,
+      "step": 7645
+    },
+    {
+      "epoch": 0.5622472240605927,
+      "grad_norm": 0.8251013159751892,
+      "learning_rate": 8.478128626729656e-05,
+      "loss": 0.3331,
+      "step": 7646
+    },
+    {
+      "epoch": 0.5623207588793293,
+      "grad_norm": 1.5636528730392456,
+      "learning_rate": 8.475774793461079e-05,
+      "loss": 0.3652,
+      "step": 7647
+    },
+    {
+      "epoch": 0.562394293698066,
+      "grad_norm": 0.9658908247947693,
+      "learning_rate": 8.473421046648165e-05,
+      "loss": 0.3393,
+      "step": 7648
+    },
+    {
+      "epoch": 0.5624678285168027,
+      "grad_norm": 1.6255276203155518,
+      "learning_rate": 8.471067386424421e-05,
+      "loss": 0.3526,
+      "step": 7649
+    },
+    {
+      "epoch": 0.5625413633355394,
+      "grad_norm": 1.1823607683181763,
+      "learning_rate": 8.468713812923344e-05,
+      "loss": 0.3229,
+      "step": 7650
+    },
+    {
+      "epoch": 0.562614898154276,
+      "grad_norm": 1.4129124879837036,
+      "learning_rate": 8.46636032627844e-05,
+      "loss": 0.3681,
+      "step": 7651
+    },
+    {
+      "epoch": 0.5626884329730127,
+      "grad_norm": 1.1703908443450928,
+      "learning_rate": 8.464006926623195e-05,
+      "loss": 0.3539,
+      "step": 7652
+    },
+    {
+      "epoch": 0.5627619677917494,
+      "grad_norm": 1.9206047058105469,
+      "learning_rate": 8.461653614091097e-05,
+      "loss": 0.3602,
+      "step": 7653
+    },
+    {
+      "epoch": 0.5628355026104861,
+      "grad_norm": 1.502793550491333,
+      "learning_rate": 8.45930038881563e-05,
+      "loss": 0.3594,
+      "step": 7654
+    },
+    {
+      "epoch": 0.5629090374292227,
+      "grad_norm": 1.443351149559021,
+      "learning_rate": 8.456947250930273e-05,
+      "loss": 0.3373,
+      "step": 7655
+    },
+    {
+      "epoch": 0.5629825722479594,
+      "grad_norm": 1.3772151470184326,
+      "learning_rate": 8.454594200568493e-05,
+      "loss": 0.338,
+      "step": 7656
+    },
+    {
+      "epoch": 0.5630561070666961,
+      "grad_norm": 1.4561470746994019,
+      "learning_rate": 8.452241237863763e-05,
+      "loss": 0.3446,
+      "step": 7657
+    },
+    {
+      "epoch": 0.5631296418854328,
+      "grad_norm": 1.3147649765014648,
+      "learning_rate": 8.449888362949541e-05,
+      "loss": 0.3757,
+      "step": 7658
+    },
+    {
+      "epoch": 0.5632031767041694,
+      "grad_norm": 1.1935348510742188,
+      "learning_rate": 8.447535575959295e-05,
+      "loss": 0.3672,
+      "step": 7659
+    },
+    {
+      "epoch": 0.5632767115229061,
+      "grad_norm": 1.1019657850265503,
+      "learning_rate": 8.445182877026461e-05,
+      "loss": 0.3374,
+      "step": 7660
+    },
+    {
+      "epoch": 0.5633502463416428,
+      "grad_norm": 1.587770938873291,
+      "learning_rate": 8.442830266284498e-05,
+      "loss": 0.3585,
+      "step": 7661
+    },
+    {
+      "epoch": 0.5634237811603794,
+      "grad_norm": 1.3335374593734741,
+      "learning_rate": 8.440477743866846e-05,
+      "loss": 0.357,
+      "step": 7662
+    },
+    {
+      "epoch": 0.5634973159791161,
+      "grad_norm": 1.4650659561157227,
+      "learning_rate": 8.43812530990694e-05,
+      "loss": 0.3659,
+      "step": 7663
+    },
+    {
+      "epoch": 0.5635708507978527,
+      "grad_norm": 1.431960105895996,
+      "learning_rate": 8.435772964538216e-05,
+      "loss": 0.3774,
+      "step": 7664
+    },
+    {
+      "epoch": 0.5636443856165895,
+      "grad_norm": 1.6357907056808472,
+      "learning_rate": 8.433420707894101e-05,
+      "loss": 0.3405,
+      "step": 7665
+    },
+    {
+      "epoch": 0.5637179204353261,
+      "grad_norm": 1.2361090183258057,
+      "learning_rate": 8.431068540108015e-05,
+      "loss": 0.3388,
+      "step": 7666
+    },
+    {
+      "epoch": 0.5637914552540628,
+      "grad_norm": 1.5144414901733398,
+      "learning_rate": 8.42871646131338e-05,
+      "loss": 0.3352,
+      "step": 7667
+    },
+    {
+      "epoch": 0.5638649900727994,
+      "grad_norm": 1.5667564868927002,
+      "learning_rate": 8.426364471643604e-05,
+      "loss": 0.3757,
+      "step": 7668
+    },
+    {
+      "epoch": 0.5639385248915362,
+      "grad_norm": 1.2037935256958008,
+      "learning_rate": 8.424012571232094e-05,
+      "loss": 0.3444,
+      "step": 7669
+    },
+    {
+      "epoch": 0.5640120597102728,
+      "grad_norm": 0.7977166175842285,
+      "learning_rate": 8.421660760212257e-05,
+      "loss": 0.324,
+      "step": 7670
+    },
+    {
+      "epoch": 0.5640855945290095,
+      "grad_norm": 1.2007055282592773,
+      "learning_rate": 8.419309038717489e-05,
+      "loss": 0.3435,
+      "step": 7671
+    },
+    {
+      "epoch": 0.5641591293477461,
+      "grad_norm": 0.6762022972106934,
+      "learning_rate": 8.416957406881179e-05,
+      "loss": 0.346,
+      "step": 7672
+    },
+    {
+      "epoch": 0.5642326641664829,
+      "grad_norm": 1.6045796871185303,
+      "learning_rate": 8.414605864836712e-05,
+      "loss": 0.3304,
+      "step": 7673
+    },
+    {
+      "epoch": 0.5643061989852195,
+      "grad_norm": 0.7256388664245605,
+      "learning_rate": 8.412254412717476e-05,
+      "loss": 0.3545,
+      "step": 7674
+    },
+    {
+      "epoch": 0.5643797338039561,
+      "grad_norm": 1.317524790763855,
+      "learning_rate": 8.409903050656847e-05,
+      "loss": 0.3526,
+      "step": 7675
+    },
+    {
+      "epoch": 0.5644532686226928,
+      "grad_norm": 0.8778370022773743,
+      "learning_rate": 8.40755177878819e-05,
+      "loss": 0.3417,
+      "step": 7676
+    },
+    {
+      "epoch": 0.5645268034414295,
+      "grad_norm": 2.0145339965820312,
+      "learning_rate": 8.405200597244881e-05,
+      "loss": 0.3661,
+      "step": 7677
+    },
+    {
+      "epoch": 0.5646003382601662,
+      "grad_norm": 1.4402700662612915,
+      "learning_rate": 8.402849506160277e-05,
+      "loss": 0.341,
+      "step": 7678
+    },
+    {
+      "epoch": 0.5646738730789028,
+      "grad_norm": 1.6107815504074097,
+      "learning_rate": 8.400498505667732e-05,
+      "loss": 0.3503,
+      "step": 7679
+    },
+    {
+      "epoch": 0.5647474078976396,
+      "grad_norm": 1.809687614440918,
+      "learning_rate": 8.398147595900602e-05,
+      "loss": 0.3467,
+      "step": 7680
+    },
+    {
+      "epoch": 0.5648209427163762,
+      "grad_norm": 1.1830226182937622,
+      "learning_rate": 8.39579677699223e-05,
+      "loss": 0.3696,
+      "step": 7681
+    },
+    {
+      "epoch": 0.5648944775351129,
+      "grad_norm": 0.8406214118003845,
+      "learning_rate": 8.393446049075956e-05,
+      "loss": 0.3423,
+      "step": 7682
+    },
+    {
+      "epoch": 0.5649680123538495,
+      "grad_norm": 1.4573893547058105,
+      "learning_rate": 8.391095412285124e-05,
+      "loss": 0.342,
+      "step": 7683
+    },
+    {
+      "epoch": 0.5650415471725863,
+      "grad_norm": 0.8829978704452515,
+      "learning_rate": 8.388744866753053e-05,
+      "loss": 0.3688,
+      "step": 7684
+    },
+    {
+      "epoch": 0.5651150819913229,
+      "grad_norm": 1.7846860885620117,
+      "learning_rate": 8.386394412613076e-05,
+      "loss": 0.3661,
+      "step": 7685
+    },
+    {
+      "epoch": 0.5651886168100596,
+      "grad_norm": 1.4276819229125977,
+      "learning_rate": 8.384044049998508e-05,
+      "loss": 0.3312,
+      "step": 7686
+    },
+    {
+      "epoch": 0.5652621516287962,
+      "grad_norm": 1.5395148992538452,
+      "learning_rate": 8.381693779042668e-05,
+      "loss": 0.3539,
+      "step": 7687
+    },
+    {
+      "epoch": 0.565335686447533,
+      "grad_norm": 1.4472095966339111,
+      "learning_rate": 8.379343599878866e-05,
+      "loss": 0.3925,
+      "step": 7688
+    },
+    {
+      "epoch": 0.5654092212662696,
+      "grad_norm": 1.4851000308990479,
+      "learning_rate": 8.376993512640402e-05,
+      "loss": 0.3679,
+      "step": 7689
+    },
+    {
+      "epoch": 0.5654827560850062,
+      "grad_norm": 1.4717342853546143,
+      "learning_rate": 8.37464351746058e-05,
+      "loss": 0.3638,
+      "step": 7690
+    },
+    {
+      "epoch": 0.5655562909037429,
+      "grad_norm": 1.4194706678390503,
+      "learning_rate": 8.372293614472696e-05,
+      "loss": 0.3797,
+      "step": 7691
+    },
+    {
+      "epoch": 0.5656298257224796,
+      "grad_norm": 1.307923674583435,
+      "learning_rate": 8.369943803810033e-05,
+      "loss": 0.3622,
+      "step": 7692
+    },
+    {
+      "epoch": 0.5657033605412163,
+      "grad_norm": 1.4116398096084595,
+      "learning_rate": 8.36759408560588e-05,
+      "loss": 0.3783,
+      "step": 7693
+    },
+    {
+      "epoch": 0.5657768953599529,
+      "grad_norm": 1.1456377506256104,
+      "learning_rate": 8.365244459993513e-05,
+      "loss": 0.382,
+      "step": 7694
+    },
+    {
+      "epoch": 0.5658504301786896,
+      "grad_norm": 1.6152911186218262,
+      "learning_rate": 8.362894927106204e-05,
+      "loss": 0.3839,
+      "step": 7695
+    },
+    {
+      "epoch": 0.5659239649974263,
+      "grad_norm": 1.3798789978027344,
+      "learning_rate": 8.360545487077223e-05,
+      "loss": 0.3617,
+      "step": 7696
+    },
+    {
+      "epoch": 0.565997499816163,
+      "grad_norm": 1.645367980003357,
+      "learning_rate": 8.358196140039831e-05,
+      "loss": 0.356,
+      "step": 7697
+    },
+    {
+      "epoch": 0.5660710346348996,
+      "grad_norm": 1.6282235383987427,
+      "learning_rate": 8.355846886127289e-05,
+      "loss": 0.389,
+      "step": 7698
+    },
+    {
+      "epoch": 0.5661445694536363,
+      "grad_norm": 1.0492875576019287,
+      "learning_rate": 8.353497725472842e-05,
+      "loss": 0.3464,
+      "step": 7699
+    },
+    {
+      "epoch": 0.566218104272373,
+      "grad_norm": 1.1752017736434937,
+      "learning_rate": 8.351148658209747e-05,
+      "loss": 0.3792,
+      "step": 7700
+    },
+    {
+      "epoch": 0.5662916390911097,
+      "grad_norm": 1.054031252861023,
+      "learning_rate": 8.348799684471238e-05,
+      "loss": 0.3595,
+      "step": 7701
+    },
+    {
+      "epoch": 0.5663651739098463,
+      "grad_norm": 1.0846956968307495,
+      "learning_rate": 8.346450804390551e-05,
+      "loss": 0.3414,
+      "step": 7702
+    },
+    {
+      "epoch": 0.5664387087285829,
+      "grad_norm": 1.5485564470291138,
+      "learning_rate": 8.344102018100924e-05,
+      "loss": 0.3442,
+      "step": 7703
+    },
+    {
+      "epoch": 0.5665122435473197,
+      "grad_norm": 1.5371967554092407,
+      "learning_rate": 8.341753325735576e-05,
+      "loss": 0.3502,
+      "step": 7704
+    },
+    {
+      "epoch": 0.5665857783660563,
+      "grad_norm": 1.3962205648422241,
+      "learning_rate": 8.339404727427729e-05,
+      "loss": 0.3373,
+      "step": 7705
+    },
+    {
+      "epoch": 0.566659313184793,
+      "grad_norm": 1.0833207368850708,
+      "learning_rate": 8.337056223310602e-05,
+      "loss": 0.3698,
+      "step": 7706
+    },
+    {
+      "epoch": 0.5667328480035296,
+      "grad_norm": 1.7732266187667847,
+      "learning_rate": 8.334707813517398e-05,
+      "loss": 0.3641,
+      "step": 7707
+    },
+    {
+      "epoch": 0.5668063828222664,
+      "grad_norm": 1.7349650859832764,
+      "learning_rate": 8.332359498181324e-05,
+      "loss": 0.3679,
+      "step": 7708
+    },
+    {
+      "epoch": 0.566879917641003,
+      "grad_norm": 1.268524408340454,
+      "learning_rate": 8.330011277435578e-05,
+      "loss": 0.3702,
+      "step": 7709
+    },
+    {
+      "epoch": 0.5669534524597397,
+      "grad_norm": 1.3128830194473267,
+      "learning_rate": 8.327663151413357e-05,
+      "loss": 0.355,
+      "step": 7710
+    },
+    {
+      "epoch": 0.5670269872784763,
+      "grad_norm": 1.2923696041107178,
+      "learning_rate": 8.325315120247845e-05,
+      "loss": 0.3831,
+      "step": 7711
+    },
+    {
+      "epoch": 0.5671005220972131,
+      "grad_norm": 1.219282627105713,
+      "learning_rate": 8.322967184072227e-05,
+      "loss": 0.3439,
+      "step": 7712
+    },
+    {
+      "epoch": 0.5671740569159497,
+      "grad_norm": 1.7268366813659668,
+      "learning_rate": 8.320619343019682e-05,
+      "loss": 0.3533,
+      "step": 7713
+    },
+    {
+      "epoch": 0.5672475917346864,
+      "grad_norm": 1.5755723714828491,
+      "learning_rate": 8.318271597223379e-05,
+      "loss": 0.3469,
+      "step": 7714
+    },
+    {
+      "epoch": 0.567321126553423,
+      "grad_norm": 1.7249614000320435,
+      "learning_rate": 8.315923946816484e-05,
+      "loss": 0.3441,
+      "step": 7715
+    },
+    {
+      "epoch": 0.5673946613721598,
+      "grad_norm": 1.2155020236968994,
+      "learning_rate": 8.313576391932164e-05,
+      "loss": 0.3565,
+      "step": 7716
+    },
+    {
+      "epoch": 0.5674681961908964,
+      "grad_norm": 1.6053850650787354,
+      "learning_rate": 8.311228932703565e-05,
+      "loss": 0.3704,
+      "step": 7717
+    },
+    {
+      "epoch": 0.567541731009633,
+      "grad_norm": 1.5827713012695312,
+      "learning_rate": 8.308881569263852e-05,
+      "loss": 0.3939,
+      "step": 7718
+    },
+    {
+      "epoch": 0.5676152658283697,
+      "grad_norm": 1.4826618432998657,
+      "learning_rate": 8.306534301746155e-05,
+      "loss": 0.3459,
+      "step": 7719
+    },
+    {
+      "epoch": 0.5676888006471064,
+      "grad_norm": 1.5668129920959473,
+      "learning_rate": 8.30418713028362e-05,
+      "loss": 0.3799,
+      "step": 7720
+    },
+    {
+      "epoch": 0.5677623354658431,
+      "grad_norm": 1.3190041780471802,
+      "learning_rate": 8.301840055009382e-05,
+      "loss": 0.3893,
+      "step": 7721
+    },
+    {
+      "epoch": 0.5678358702845797,
+      "grad_norm": 1.032352328300476,
+      "learning_rate": 8.299493076056566e-05,
+      "loss": 0.3673,
+      "step": 7722
+    },
+    {
+      "epoch": 0.5679094051033164,
+      "grad_norm": 1.3711907863616943,
+      "learning_rate": 8.2971461935583e-05,
+      "loss": 0.3952,
+      "step": 7723
+    },
+    {
+      "epoch": 0.5679829399220531,
+      "grad_norm": 1.286271095275879,
+      "learning_rate": 8.294799407647698e-05,
+      "loss": 0.3655,
+      "step": 7724
+    },
+    {
+      "epoch": 0.5680564747407898,
+      "grad_norm": 1.3104757070541382,
+      "learning_rate": 8.292452718457871e-05,
+      "loss": 0.3731,
+      "step": 7725
+    },
+    {
+      "epoch": 0.5681300095595264,
+      "grad_norm": 1.1338286399841309,
+      "learning_rate": 8.290106126121931e-05,
+      "loss": 0.3654,
+      "step": 7726
+    },
+    {
+      "epoch": 0.5682035443782631,
+      "grad_norm": 1.6224207878112793,
+      "learning_rate": 8.287759630772976e-05,
+      "loss": 0.3638,
+      "step": 7727
+    },
+    {
+      "epoch": 0.5682770791969998,
+      "grad_norm": 1.4472523927688599,
+      "learning_rate": 8.285413232544099e-05,
+      "loss": 0.3727,
+      "step": 7728
+    },
+    {
+      "epoch": 0.5683506140157365,
+      "grad_norm": 1.5813696384429932,
+      "learning_rate": 8.283066931568396e-05,
+      "loss": 0.3804,
+      "step": 7729
+    },
+    {
+      "epoch": 0.5684241488344731,
+      "grad_norm": 1.568503499031067,
+      "learning_rate": 8.280720727978948e-05,
+      "loss": 0.3753,
+      "step": 7730
+    },
+    {
+      "epoch": 0.5684976836532097,
+      "grad_norm": 0.9750174880027771,
+      "learning_rate": 8.278374621908835e-05,
+      "loss": 0.3554,
+      "step": 7731
+    },
+    {
+      "epoch": 0.5685712184719465,
+      "grad_norm": 0.942540168762207,
+      "learning_rate": 8.276028613491127e-05,
+      "loss": 0.3809,
+      "step": 7732
+    },
+    {
+      "epoch": 0.5686447532906831,
+      "grad_norm": 1.1997575759887695,
+      "learning_rate": 8.273682702858898e-05,
+      "loss": 0.373,
+      "step": 7733
+    },
+    {
+      "epoch": 0.5687182881094198,
+      "grad_norm": 0.6452527642250061,
+      "learning_rate": 8.271336890145208e-05,
+      "loss": 0.3642,
+      "step": 7734
+    },
+    {
+      "epoch": 0.5687918229281564,
+      "grad_norm": 1.3071502447128296,
+      "learning_rate": 8.268991175483112e-05,
+      "loss": 0.3741,
+      "step": 7735
+    },
+    {
+      "epoch": 0.5688653577468932,
+      "grad_norm": 0.5256643295288086,
+      "learning_rate": 8.266645559005664e-05,
+      "loss": 0.3941,
+      "step": 7736
+    },
+    {
+      "epoch": 0.5689388925656298,
+      "grad_norm": 0.7618046402931213,
+      "learning_rate": 8.264300040845912e-05,
+      "loss": 0.363,
+      "step": 7737
+    },
+    {
+      "epoch": 0.5690124273843665,
+      "grad_norm": 0.7929079532623291,
+      "learning_rate": 8.261954621136889e-05,
+      "loss": 0.3736,
+      "step": 7738
+    },
+    {
+      "epoch": 0.5690859622031031,
+      "grad_norm": 0.6193045973777771,
+      "learning_rate": 8.259609300011638e-05,
+      "loss": 0.3363,
+      "step": 7739
+    },
+    {
+      "epoch": 0.5691594970218399,
+      "grad_norm": 1.0193917751312256,
+      "learning_rate": 8.257264077603181e-05,
+      "loss": 0.3665,
+      "step": 7740
+    },
+    {
+      "epoch": 0.5692330318405765,
+      "grad_norm": 0.5270761847496033,
+      "learning_rate": 8.25491895404455e-05,
+      "loss": 0.3557,
+      "step": 7741
+    },
+    {
+      "epoch": 0.5693065666593132,
+      "grad_norm": 0.9522221684455872,
+      "learning_rate": 8.252573929468753e-05,
+      "loss": 0.3381,
+      "step": 7742
+    },
+    {
+      "epoch": 0.5693801014780498,
+      "grad_norm": 0.8869754076004028,
+      "learning_rate": 8.250229004008809e-05,
+      "loss": 0.3462,
+      "step": 7743
+    },
+    {
+      "epoch": 0.5694536362967866,
+      "grad_norm": 0.4682360887527466,
+      "learning_rate": 8.247884177797722e-05,
+      "loss": 0.3837,
+      "step": 7744
+    },
+    {
+      "epoch": 0.5695271711155232,
+      "grad_norm": 0.49778205156326294,
+      "learning_rate": 8.245539450968491e-05,
+      "loss": 0.3518,
+      "step": 7745
+    },
+    {
+      "epoch": 0.5696007059342598,
+      "grad_norm": 0.4949684143066406,
+      "learning_rate": 8.243194823654116e-05,
+      "loss": 0.3567,
+      "step": 7746
+    },
+    {
+      "epoch": 0.5696742407529966,
+      "grad_norm": 0.47970688343048096,
+      "learning_rate": 8.240850295987586e-05,
+      "loss": 0.3622,
+      "step": 7747
+    },
+    {
+      "epoch": 0.5697477755717332,
+      "grad_norm": 0.5441577434539795,
+      "learning_rate": 8.23850586810188e-05,
+      "loss": 0.3711,
+      "step": 7748
+    },
+    {
+      "epoch": 0.5698213103904699,
+      "grad_norm": 0.44171667098999023,
+      "learning_rate": 8.236161540129983e-05,
+      "loss": 0.3564,
+      "step": 7749
+    },
+    {
+      "epoch": 0.5698948452092065,
+      "grad_norm": 0.4826618731021881,
+      "learning_rate": 8.233817312204865e-05,
+      "loss": 0.3598,
+      "step": 7750
+    },
+    {
+      "epoch": 0.5699683800279433,
+      "grad_norm": 0.7069659233093262,
+      "learning_rate": 8.231473184459489e-05,
+      "loss": 0.3638,
+      "step": 7751
+    },
+    {
+      "epoch": 0.5700419148466799,
+      "grad_norm": 0.5544124245643616,
+      "learning_rate": 8.229129157026824e-05,
+      "loss": 0.3543,
+      "step": 7752
+    },
+    {
+      "epoch": 0.5701154496654166,
+      "grad_norm": 1.1144914627075195,
+      "learning_rate": 8.226785230039826e-05,
+      "loss": 0.383,
+      "step": 7753
+    },
+    {
+      "epoch": 0.5701889844841532,
+      "grad_norm": 0.9502038359642029,
+      "learning_rate": 8.224441403631437e-05,
+      "loss": 0.3796,
+      "step": 7754
+    },
+    {
+      "epoch": 0.57026251930289,
+      "grad_norm": 0.6581213474273682,
+      "learning_rate": 8.222097677934603e-05,
+      "loss": 0.3502,
+      "step": 7755
+    },
+    {
+      "epoch": 0.5703360541216266,
+      "grad_norm": 0.9100571870803833,
+      "learning_rate": 8.219754053082267e-05,
+      "loss": 0.3776,
+      "step": 7756
+    },
+    {
+      "epoch": 0.5704095889403633,
+      "grad_norm": 0.9517785906791687,
+      "learning_rate": 8.21741052920736e-05,
+      "loss": 0.3544,
+      "step": 7757
+    },
+    {
+      "epoch": 0.5704831237590999,
+      "grad_norm": 0.7520815134048462,
+      "learning_rate": 8.215067106442808e-05,
+      "loss": 0.3674,
+      "step": 7758
+    },
+    {
+      "epoch": 0.5705566585778366,
+      "grad_norm": 0.7943175435066223,
+      "learning_rate": 8.212723784921536e-05,
+      "loss": 0.3635,
+      "step": 7759
+    },
+    {
+      "epoch": 0.5706301933965733,
+      "grad_norm": 0.7980525493621826,
+      "learning_rate": 8.210380564776455e-05,
+      "loss": 0.3532,
+      "step": 7760
+    },
+    {
+      "epoch": 0.5707037282153099,
+      "grad_norm": 0.6586311459541321,
+      "learning_rate": 8.208037446140477e-05,
+      "loss": 0.3595,
+      "step": 7761
+    },
+    {
+      "epoch": 0.5707772630340466,
+      "grad_norm": 0.5389314889907837,
+      "learning_rate": 8.205694429146506e-05,
+      "loss": 0.3402,
+      "step": 7762
+    },
+    {
+      "epoch": 0.5708507978527833,
+      "grad_norm": 0.7593215107917786,
+      "learning_rate": 8.203351513927441e-05,
+      "loss": 0.3825,
+      "step": 7763
+    },
+    {
+      "epoch": 0.57092433267152,
+      "grad_norm": 0.8209600448608398,
+      "learning_rate": 8.201008700616179e-05,
+      "loss": 0.3739,
+      "step": 7764
+    },
+    {
+      "epoch": 0.5709978674902566,
+      "grad_norm": 0.837499737739563,
+      "learning_rate": 8.198665989345597e-05,
+      "loss": 0.3441,
+      "step": 7765
+    },
+    {
+      "epoch": 0.5710714023089933,
+      "grad_norm": 0.5714189410209656,
+      "learning_rate": 8.196323380248582e-05,
+      "loss": 0.3689,
+      "step": 7766
+    },
+    {
+      "epoch": 0.57114493712773,
+      "grad_norm": 0.6219752430915833,
+      "learning_rate": 8.193980873458008e-05,
+      "loss": 0.3703,
+      "step": 7767
+    },
+    {
+      "epoch": 0.5712184719464667,
+      "grad_norm": 0.719826877117157,
+      "learning_rate": 8.191638469106743e-05,
+      "loss": 0.405,
+      "step": 7768
+    },
+    {
+      "epoch": 0.5712920067652033,
+      "grad_norm": 0.49243003129959106,
+      "learning_rate": 8.189296167327656e-05,
+      "loss": 0.3835,
+      "step": 7769
+    },
+    {
+      "epoch": 0.57136554158394,
+      "grad_norm": 0.6292635202407837,
+      "learning_rate": 8.186953968253599e-05,
+      "loss": 0.3772,
+      "step": 7770
+    },
+    {
+      "epoch": 0.5714390764026767,
+      "grad_norm": 0.478537380695343,
+      "learning_rate": 8.184611872017426e-05,
+      "loss": 0.3511,
+      "step": 7771
+    },
+    {
+      "epoch": 0.5715126112214134,
+      "grad_norm": 0.8555082082748413,
+      "learning_rate": 8.182269878751985e-05,
+      "loss": 0.3637,
+      "step": 7772
+    },
+    {
+      "epoch": 0.57158614604015,
+      "grad_norm": 0.6781911849975586,
+      "learning_rate": 8.179927988590115e-05,
+      "loss": 0.369,
+      "step": 7773
+    },
+    {
+      "epoch": 0.5716596808588866,
+      "grad_norm": 0.7124538421630859,
+      "learning_rate": 8.177586201664648e-05,
+      "loss": 0.3438,
+      "step": 7774
+    },
+    {
+      "epoch": 0.5717332156776234,
+      "grad_norm": 0.6906933784484863,
+      "learning_rate": 8.175244518108418e-05,
+      "loss": 0.3922,
+      "step": 7775
+    },
+    {
+      "epoch": 0.57180675049636,
+      "grad_norm": 1.1711000204086304,
+      "learning_rate": 8.172902938054245e-05,
+      "loss": 0.3866,
+      "step": 7776
+    },
+    {
+      "epoch": 0.5718802853150967,
+      "grad_norm": 0.6021994352340698,
+      "learning_rate": 8.170561461634944e-05,
+      "loss": 0.3969,
+      "step": 7777
+    },
+    {
+      "epoch": 0.5719538201338333,
+      "grad_norm": 0.7893120050430298,
+      "learning_rate": 8.168220088983325e-05,
+      "loss": 0.3658,
+      "step": 7778
+    },
+    {
+      "epoch": 0.5720273549525701,
+      "grad_norm": 0.6408349275588989,
+      "learning_rate": 8.165878820232198e-05,
+      "loss": 0.362,
+      "step": 7779
+    },
+    {
+      "epoch": 0.5721008897713067,
+      "grad_norm": 0.7570788860321045,
+      "learning_rate": 8.163537655514362e-05,
+      "loss": 0.38,
+      "step": 7780
+    },
+    {
+      "epoch": 0.5721744245900434,
+      "grad_norm": 0.828637957572937,
+      "learning_rate": 8.161196594962605e-05,
+      "loss": 0.3792,
+      "step": 7781
+    },
+    {
+      "epoch": 0.57224795940878,
+      "grad_norm": 0.6898981332778931,
+      "learning_rate": 8.15885563870972e-05,
+      "loss": 0.3931,
+      "step": 7782
+    },
+    {
+      "epoch": 0.5723214942275168,
+      "grad_norm": 0.7384957075119019,
+      "learning_rate": 8.156514786888486e-05,
+      "loss": 0.3761,
+      "step": 7783
+    },
+    {
+      "epoch": 0.5723950290462534,
+      "grad_norm": 0.5850455164909363,
+      "learning_rate": 8.154174039631675e-05,
+      "loss": 0.3791,
+      "step": 7784
+    },
+    {
+      "epoch": 0.5724685638649901,
+      "grad_norm": 0.5721343159675598,
+      "learning_rate": 8.151833397072064e-05,
+      "loss": 0.366,
+      "step": 7785
+    },
+    {
+      "epoch": 0.5725420986837267,
+      "grad_norm": 0.8576267957687378,
+      "learning_rate": 8.149492859342411e-05,
+      "loss": 0.3798,
+      "step": 7786
+    },
+    {
+      "epoch": 0.5726156335024634,
+      "grad_norm": 1.0207594633102417,
+      "learning_rate": 8.147152426575477e-05,
+      "loss": 0.3604,
+      "step": 7787
+    },
+    {
+      "epoch": 0.5726891683212001,
+      "grad_norm": 0.6456494927406311,
+      "learning_rate": 8.144812098904017e-05,
+      "loss": 0.3564,
+      "step": 7788
+    },
+    {
+      "epoch": 0.5727627031399367,
+      "grad_norm": 1.437841773033142,
+      "learning_rate": 8.142471876460769e-05,
+      "loss": 0.363,
+      "step": 7789
+    },
+    {
+      "epoch": 0.5728362379586734,
+      "grad_norm": 0.8544130325317383,
+      "learning_rate": 8.140131759378477e-05,
+      "loss": 0.3947,
+      "step": 7790
+    },
+    {
+      "epoch": 0.5729097727774101,
+      "grad_norm": 0.9979299902915955,
+      "learning_rate": 8.137791747789871e-05,
+      "loss": 0.3744,
+      "step": 7791
+    },
+    {
+      "epoch": 0.5729833075961468,
+      "grad_norm": 1.0539854764938354,
+      "learning_rate": 8.135451841827686e-05,
+      "loss": 0.3499,
+      "step": 7792
+    },
+    {
+      "epoch": 0.5730568424148834,
+      "grad_norm": 0.8720017671585083,
+      "learning_rate": 8.13311204162464e-05,
+      "loss": 0.4155,
+      "step": 7793
+    },
+    {
+      "epoch": 0.5731303772336201,
+      "grad_norm": 1.0275461673736572,
+      "learning_rate": 8.130772347313447e-05,
+      "loss": 0.3805,
+      "step": 7794
+    },
+    {
+      "epoch": 0.5732039120523568,
+      "grad_norm": 1.4037115573883057,
+      "learning_rate": 8.128432759026822e-05,
+      "loss": 0.3689,
+      "step": 7795
+    },
+    {
+      "epoch": 0.5732774468710935,
+      "grad_norm": 0.7360805869102478,
+      "learning_rate": 8.126093276897465e-05,
+      "loss": 0.3773,
+      "step": 7796
+    },
+    {
+      "epoch": 0.5733509816898301,
+      "grad_norm": 1.7134994268417358,
+      "learning_rate": 8.123753901058074e-05,
+      "loss": 0.3903,
+      "step": 7797
+    },
+    {
+      "epoch": 0.5734245165085669,
+      "grad_norm": 0.894292950630188,
+      "learning_rate": 8.121414631641342e-05,
+      "loss": 0.3762,
+      "step": 7798
+    },
+    {
+      "epoch": 0.5734980513273035,
+      "grad_norm": 1.8636047840118408,
+      "learning_rate": 8.119075468779957e-05,
+      "loss": 0.3874,
+      "step": 7799
+    },
+    {
+      "epoch": 0.5735715861460402,
+      "grad_norm": 1.4409728050231934,
+      "learning_rate": 8.116736412606596e-05,
+      "loss": 0.3815,
+      "step": 7800
+    },
+    {
+      "epoch": 0.5736451209647768,
+      "grad_norm": 1.3361090421676636,
+      "learning_rate": 8.11439746325393e-05,
+      "loss": 0.3531,
+      "step": 7801
+    },
+    {
+      "epoch": 0.5737186557835134,
+      "grad_norm": 1.498543620109558,
+      "learning_rate": 8.112058620854633e-05,
+      "loss": 0.3832,
+      "step": 7802
+    },
+    {
+      "epoch": 0.5737921906022502,
+      "grad_norm": 0.9636831283569336,
+      "learning_rate": 8.109719885541361e-05,
+      "loss": 0.3772,
+      "step": 7803
+    },
+    {
+      "epoch": 0.5738657254209868,
+      "grad_norm": 0.8541214466094971,
+      "learning_rate": 8.107381257446773e-05,
+      "loss": 0.3617,
+      "step": 7804
+    },
+    {
+      "epoch": 0.5739392602397235,
+      "grad_norm": 0.7552184462547302,
+      "learning_rate": 8.105042736703517e-05,
+      "loss": 0.3653,
+      "step": 7805
+    },
+    {
+      "epoch": 0.5740127950584601,
+      "grad_norm": 0.7331070899963379,
+      "learning_rate": 8.102704323444238e-05,
+      "loss": 0.3572,
+      "step": 7806
+    },
+    {
+      "epoch": 0.5740863298771969,
+      "grad_norm": 0.8776545524597168,
+      "learning_rate": 8.10036601780157e-05,
+      "loss": 0.394,
+      "step": 7807
+    },
+    {
+      "epoch": 0.5741598646959335,
+      "grad_norm": 0.703478217124939,
+      "learning_rate": 8.098027819908147e-05,
+      "loss": 0.3528,
+      "step": 7808
+    },
+    {
+      "epoch": 0.5742333995146702,
+      "grad_norm": 0.5338261127471924,
+      "learning_rate": 8.095689729896595e-05,
+      "loss": 0.381,
+      "step": 7809
+    },
+    {
+      "epoch": 0.5743069343334068,
+      "grad_norm": 0.6662909388542175,
+      "learning_rate": 8.093351747899528e-05,
+      "loss": 0.4133,
+      "step": 7810
+    },
+    {
+      "epoch": 0.5743804691521436,
+      "grad_norm": 0.5690781474113464,
+      "learning_rate": 8.091013874049566e-05,
+      "loss": 0.3706,
+      "step": 7811
+    },
+    {
+      "epoch": 0.5744540039708802,
+      "grad_norm": 0.5729175806045532,
+      "learning_rate": 8.088676108479311e-05,
+      "loss": 0.3801,
+      "step": 7812
+    },
+    {
+      "epoch": 0.5745275387896169,
+      "grad_norm": 0.7595308423042297,
+      "learning_rate": 8.086338451321364e-05,
+      "loss": 0.3784,
+      "step": 7813
+    },
+    {
+      "epoch": 0.5746010736083536,
+      "grad_norm": 0.6314162611961365,
+      "learning_rate": 8.084000902708318e-05,
+      "loss": 0.3493,
+      "step": 7814
+    },
+    {
+      "epoch": 0.5746746084270902,
+      "grad_norm": 1.1295338869094849,
+      "learning_rate": 8.081663462772763e-05,
+      "loss": 0.4028,
+      "step": 7815
+    },
+    {
+      "epoch": 0.5747481432458269,
+      "grad_norm": 1.1058101654052734,
+      "learning_rate": 8.079326131647282e-05,
+      "loss": 0.3918,
+      "step": 7816
+    },
+    {
+      "epoch": 0.5748216780645635,
+      "grad_norm": 0.5273606777191162,
+      "learning_rate": 8.076988909464446e-05,
+      "loss": 0.3656,
+      "step": 7817
+    },
+    {
+      "epoch": 0.5748952128833003,
+      "grad_norm": 1.0255643129348755,
+      "learning_rate": 8.074651796356833e-05,
+      "loss": 0.3724,
+      "step": 7818
+    },
+    {
+      "epoch": 0.5749687477020369,
+      "grad_norm": 0.6332260370254517,
+      "learning_rate": 8.072314792457002e-05,
+      "loss": 0.3614,
+      "step": 7819
+    },
+    {
+      "epoch": 0.5750422825207736,
+      "grad_norm": 0.6734099388122559,
+      "learning_rate": 8.069977897897507e-05,
+      "loss": 0.3767,
+      "step": 7820
+    },
+    {
+      "epoch": 0.5751158173395102,
+      "grad_norm": 0.5936859250068665,
+      "learning_rate": 8.067641112810905e-05,
+      "loss": 0.3466,
+      "step": 7821
+    },
+    {
+      "epoch": 0.575189352158247,
+      "grad_norm": 0.5805171132087708,
+      "learning_rate": 8.065304437329736e-05,
+      "loss": 0.3967,
+      "step": 7822
+    },
+    {
+      "epoch": 0.5752628869769836,
+      "grad_norm": 0.5643659234046936,
+      "learning_rate": 8.062967871586548e-05,
+      "loss": 0.3606,
+      "step": 7823
+    },
+    {
+      "epoch": 0.5753364217957203,
+      "grad_norm": 0.6515311002731323,
+      "learning_rate": 8.06063141571386e-05,
+      "loss": 0.3538,
+      "step": 7824
+    },
+    {
+      "epoch": 0.5754099566144569,
+      "grad_norm": 0.6337628364562988,
+      "learning_rate": 8.058295069844206e-05,
+      "loss": 0.3877,
+      "step": 7825
+    },
+    {
+      "epoch": 0.5754834914331937,
+      "grad_norm": 0.49579378962516785,
+      "learning_rate": 8.055958834110107e-05,
+      "loss": 0.3619,
+      "step": 7826
+    },
+    {
+      "epoch": 0.5755570262519303,
+      "grad_norm": 0.65865159034729,
+      "learning_rate": 8.053622708644069e-05,
+      "loss": 0.3719,
+      "step": 7827
+    },
+    {
+      "epoch": 0.575630561070667,
+      "grad_norm": 0.6282687187194824,
+      "learning_rate": 8.05128669357861e-05,
+      "loss": 0.3579,
+      "step": 7828
+    },
+    {
+      "epoch": 0.5757040958894036,
+      "grad_norm": 0.4881400763988495,
+      "learning_rate": 8.048950789046225e-05,
+      "loss": 0.372,
+      "step": 7829
+    },
+    {
+      "epoch": 0.5757776307081403,
+      "grad_norm": 0.5442676544189453,
+      "learning_rate": 8.046614995179409e-05,
+      "loss": 0.378,
+      "step": 7830
+    },
+    {
+      "epoch": 0.575851165526877,
+      "grad_norm": 0.5283936858177185,
+      "learning_rate": 8.044279312110654e-05,
+      "loss": 0.373,
+      "step": 7831
+    },
+    {
+      "epoch": 0.5759247003456136,
+      "grad_norm": 0.49429166316986084,
+      "learning_rate": 8.04194373997244e-05,
+      "loss": 0.3758,
+      "step": 7832
+    },
+    {
+      "epoch": 0.5759982351643503,
+      "grad_norm": 0.49466216564178467,
+      "learning_rate": 8.039608278897242e-05,
+      "loss": 0.3493,
+      "step": 7833
+    },
+    {
+      "epoch": 0.576071769983087,
+      "grad_norm": 0.6050233840942383,
+      "learning_rate": 8.037272929017532e-05,
+      "loss": 0.3914,
+      "step": 7834
+    },
+    {
+      "epoch": 0.5761453048018237,
+      "grad_norm": 0.49107837677001953,
+      "learning_rate": 8.034937690465775e-05,
+      "loss": 0.3642,
+      "step": 7835
+    },
+    {
+      "epoch": 0.5762188396205603,
+      "grad_norm": 0.5432994961738586,
+      "learning_rate": 8.032602563374426e-05,
+      "loss": 0.3431,
+      "step": 7836
+    },
+    {
+      "epoch": 0.576292374439297,
+      "grad_norm": 0.4670559763908386,
+      "learning_rate": 8.030267547875933e-05,
+      "loss": 0.3625,
+      "step": 7837
+    },
+    {
+      "epoch": 0.5763659092580337,
+      "grad_norm": 0.5336586833000183,
+      "learning_rate": 8.027932644102744e-05,
+      "loss": 0.3709,
+      "step": 7838
+    },
+    {
+      "epoch": 0.5764394440767704,
+      "grad_norm": 0.5619916915893555,
+      "learning_rate": 8.025597852187298e-05,
+      "loss": 0.3637,
+      "step": 7839
+    },
+    {
+      "epoch": 0.576512978895507,
+      "grad_norm": 0.6204789876937866,
+      "learning_rate": 8.023263172262023e-05,
+      "loss": 0.3954,
+      "step": 7840
+    },
+    {
+      "epoch": 0.5765865137142437,
+      "grad_norm": 0.6015584468841553,
+      "learning_rate": 8.020928604459349e-05,
+      "loss": 0.341,
+      "step": 7841
+    },
+    {
+      "epoch": 0.5766600485329804,
+      "grad_norm": 0.5978754162788391,
+      "learning_rate": 8.018594148911694e-05,
+      "loss": 0.3643,
+      "step": 7842
+    },
+    {
+      "epoch": 0.576733583351717,
+      "grad_norm": 0.8018456101417542,
+      "learning_rate": 8.016259805751467e-05,
+      "loss": 0.3643,
+      "step": 7843
+    },
+    {
+      "epoch": 0.5768071181704537,
+      "grad_norm": 0.6691281199455261,
+      "learning_rate": 8.013925575111079e-05,
+      "loss": 0.3459,
+      "step": 7844
+    },
+    {
+      "epoch": 0.5768806529891903,
+      "grad_norm": 0.46637487411499023,
+      "learning_rate": 8.011591457122927e-05,
+      "loss": 0.3385,
+      "step": 7845
+    },
+    {
+      "epoch": 0.5769541878079271,
+      "grad_norm": 0.6991901993751526,
+      "learning_rate": 8.009257451919412e-05,
+      "loss": 0.3576,
+      "step": 7846
+    },
+    {
+      "epoch": 0.5770277226266637,
+      "grad_norm": 0.7704020142555237,
+      "learning_rate": 8.006923559632909e-05,
+      "loss": 0.3633,
+      "step": 7847
+    },
+    {
+      "epoch": 0.5771012574454004,
+      "grad_norm": 0.5979422330856323,
+      "learning_rate": 8.004589780395809e-05,
+      "loss": 0.3584,
+      "step": 7848
+    },
+    {
+      "epoch": 0.577174792264137,
+      "grad_norm": 0.5145589709281921,
+      "learning_rate": 8.00225611434048e-05,
+      "loss": 0.3758,
+      "step": 7849
+    },
+    {
+      "epoch": 0.5772483270828738,
+      "grad_norm": 0.5681321024894714,
+      "learning_rate": 7.99992256159929e-05,
+      "loss": 0.3867,
+      "step": 7850
+    },
+    {
+      "epoch": 0.5773218619016104,
+      "grad_norm": 0.7436895966529846,
+      "learning_rate": 7.997589122304608e-05,
+      "loss": 0.3758,
+      "step": 7851
+    },
+    {
+      "epoch": 0.5773953967203471,
+      "grad_norm": 0.49626803398132324,
+      "learning_rate": 7.99525579658878e-05,
+      "loss": 0.3817,
+      "step": 7852
+    },
+    {
+      "epoch": 0.5774689315390837,
+      "grad_norm": 0.6401667594909668,
+      "learning_rate": 7.992922584584161e-05,
+      "loss": 0.3592,
+      "step": 7853
+    },
+    {
+      "epoch": 0.5775424663578205,
+      "grad_norm": 0.6727571487426758,
+      "learning_rate": 7.990589486423091e-05,
+      "loss": 0.3572,
+      "step": 7854
+    },
+    {
+      "epoch": 0.5776160011765571,
+      "grad_norm": 0.6022122502326965,
+      "learning_rate": 7.988256502237907e-05,
+      "loss": 0.3619,
+      "step": 7855
+    },
+    {
+      "epoch": 0.5776895359952938,
+      "grad_norm": 0.5321819186210632,
+      "learning_rate": 7.985923632160934e-05,
+      "loss": 0.3654,
+      "step": 7856
+    },
+    {
+      "epoch": 0.5777630708140304,
+      "grad_norm": 0.6191771030426025,
+      "learning_rate": 7.983590876324502e-05,
+      "loss": 0.3501,
+      "step": 7857
+    },
+    {
+      "epoch": 0.5778366056327671,
+      "grad_norm": 0.6655152440071106,
+      "learning_rate": 7.981258234860925e-05,
+      "loss": 0.3396,
+      "step": 7858
+    },
+    {
+      "epoch": 0.5779101404515038,
+      "grad_norm": 0.5138206481933594,
+      "learning_rate": 7.978925707902506e-05,
+      "loss": 0.372,
+      "step": 7859
+    },
+    {
+      "epoch": 0.5779836752702404,
+      "grad_norm": 0.6670894622802734,
+      "learning_rate": 7.976593295581555e-05,
+      "loss": 0.3463,
+      "step": 7860
+    },
+    {
+      "epoch": 0.5780572100889771,
+      "grad_norm": 0.5477434992790222,
+      "learning_rate": 7.97426099803037e-05,
+      "loss": 0.3706,
+      "step": 7861
+    },
+    {
+      "epoch": 0.5781307449077138,
+      "grad_norm": 0.7446560263633728,
+      "learning_rate": 7.971928815381239e-05,
+      "loss": 0.3739,
+      "step": 7862
+    },
+    {
+      "epoch": 0.5782042797264505,
+      "grad_norm": 0.494760125875473,
+      "learning_rate": 7.969596747766442e-05,
+      "loss": 0.3452,
+      "step": 7863
+    },
+    {
+      "epoch": 0.5782778145451871,
+      "grad_norm": 0.642528235912323,
+      "learning_rate": 7.967264795318265e-05,
+      "loss": 0.3896,
+      "step": 7864
+    },
+    {
+      "epoch": 0.5783513493639239,
+      "grad_norm": 0.5648360252380371,
+      "learning_rate": 7.964932958168971e-05,
+      "loss": 0.3608,
+      "step": 7865
+    },
+    {
+      "epoch": 0.5784248841826605,
+      "grad_norm": 0.6409399509429932,
+      "learning_rate": 7.962601236450827e-05,
+      "loss": 0.3805,
+      "step": 7866
+    },
+    {
+      "epoch": 0.5784984190013972,
+      "grad_norm": 0.8179351091384888,
+      "learning_rate": 7.960269630296093e-05,
+      "loss": 0.3562,
+      "step": 7867
+    },
+    {
+      "epoch": 0.5785719538201338,
+      "grad_norm": 0.673319935798645,
+      "learning_rate": 7.957938139837015e-05,
+      "loss": 0.3649,
+      "step": 7868
+    },
+    {
+      "epoch": 0.5786454886388706,
+      "grad_norm": 0.5224233269691467,
+      "learning_rate": 7.955606765205844e-05,
+      "loss": 0.3734,
+      "step": 7869
+    },
+    {
+      "epoch": 0.5787190234576072,
+      "grad_norm": 0.8495147228240967,
+      "learning_rate": 7.953275506534817e-05,
+      "loss": 0.3588,
+      "step": 7870
+    },
+    {
+      "epoch": 0.5787925582763438,
+      "grad_norm": 0.5413752198219299,
+      "learning_rate": 7.950944363956161e-05,
+      "loss": 0.3506,
+      "step": 7871
+    },
+    {
+      "epoch": 0.5788660930950805,
+      "grad_norm": 0.8820645809173584,
+      "learning_rate": 7.948613337602103e-05,
+      "loss": 0.3798,
+      "step": 7872
+    },
+    {
+      "epoch": 0.5789396279138171,
+      "grad_norm": 0.6929572820663452,
+      "learning_rate": 7.946282427604859e-05,
+      "loss": 0.4055,
+      "step": 7873
+    },
+    {
+      "epoch": 0.5790131627325539,
+      "grad_norm": 0.7468129396438599,
+      "learning_rate": 7.943951634096645e-05,
+      "loss": 0.3773,
+      "step": 7874
+    },
+    {
+      "epoch": 0.5790866975512905,
+      "grad_norm": 0.8426890969276428,
+      "learning_rate": 7.941620957209666e-05,
+      "loss": 0.3846,
+      "step": 7875
+    },
+    {
+      "epoch": 0.5791602323700272,
+      "grad_norm": 0.5318155288696289,
+      "learning_rate": 7.939290397076115e-05,
+      "loss": 0.3668,
+      "step": 7876
+    },
+    {
+      "epoch": 0.5792337671887638,
+      "grad_norm": 0.7023219466209412,
+      "learning_rate": 7.936959953828189e-05,
+      "loss": 0.3632,
+      "step": 7877
+    },
+    {
+      "epoch": 0.5793073020075006,
+      "grad_norm": 0.6870683431625366,
+      "learning_rate": 7.934629627598073e-05,
+      "loss": 0.361,
+      "step": 7878
+    },
+    {
+      "epoch": 0.5793808368262372,
+      "grad_norm": 0.5772577524185181,
+      "learning_rate": 7.932299418517941e-05,
+      "loss": 0.3579,
+      "step": 7879
+    },
+    {
+      "epoch": 0.5794543716449739,
+      "grad_norm": 0.8082666397094727,
+      "learning_rate": 7.92996932671997e-05,
+      "loss": 0.3719,
+      "step": 7880
+    },
+    {
+      "epoch": 0.5795279064637106,
+      "grad_norm": 0.5423339605331421,
+      "learning_rate": 7.927639352336327e-05,
+      "loss": 0.3933,
+      "step": 7881
+    },
+    {
+      "epoch": 0.5796014412824473,
+      "grad_norm": 0.5628049373626709,
+      "learning_rate": 7.925309495499162e-05,
+      "loss": 0.352,
+      "step": 7882
+    },
+    {
+      "epoch": 0.5796749761011839,
+      "grad_norm": 0.6247488856315613,
+      "learning_rate": 7.922979756340633e-05,
+      "loss": 0.3679,
+      "step": 7883
+    },
+    {
+      "epoch": 0.5797485109199206,
+      "grad_norm": 0.5714010000228882,
+      "learning_rate": 7.920650134992884e-05,
+      "loss": 0.3835,
+      "step": 7884
+    },
+    {
+      "epoch": 0.5798220457386573,
+      "grad_norm": 0.5632257461547852,
+      "learning_rate": 7.918320631588055e-05,
+      "loss": 0.3606,
+      "step": 7885
+    },
+    {
+      "epoch": 0.5798955805573939,
+      "grad_norm": 0.46902820467948914,
+      "learning_rate": 7.915991246258275e-05,
+      "loss": 0.3921,
+      "step": 7886
+    },
+    {
+      "epoch": 0.5799691153761306,
+      "grad_norm": 0.8518468141555786,
+      "learning_rate": 7.913661979135673e-05,
+      "loss": 0.3488,
+      "step": 7887
+    },
+    {
+      "epoch": 0.5800426501948672,
+      "grad_norm": 0.9833033680915833,
+      "learning_rate": 7.911332830352364e-05,
+      "loss": 0.3733,
+      "step": 7888
+    },
+    {
+      "epoch": 0.580116185013604,
+      "grad_norm": 0.5358754992485046,
+      "learning_rate": 7.90900380004046e-05,
+      "loss": 0.3743,
+      "step": 7889
+    },
+    {
+      "epoch": 0.5801897198323406,
+      "grad_norm": 0.9246276617050171,
+      "learning_rate": 7.906674888332069e-05,
+      "loss": 0.3844,
+      "step": 7890
+    },
+    {
+      "epoch": 0.5802632546510773,
+      "grad_norm": 0.5035247802734375,
+      "learning_rate": 7.904346095359285e-05,
+      "loss": 0.3671,
+      "step": 7891
+    },
+    {
+      "epoch": 0.5803367894698139,
+      "grad_norm": 0.8894498348236084,
+      "learning_rate": 7.902017421254204e-05,
+      "loss": 0.3791,
+      "step": 7892
+    },
+    {
+      "epoch": 0.5804103242885507,
+      "grad_norm": 0.568418025970459,
+      "learning_rate": 7.899688866148912e-05,
+      "loss": 0.3658,
+      "step": 7893
+    },
+    {
+      "epoch": 0.5804838591072873,
+      "grad_norm": 0.5713819265365601,
+      "learning_rate": 7.897360430175483e-05,
+      "loss": 0.3584,
+      "step": 7894
+    },
+    {
+      "epoch": 0.580557393926024,
+      "grad_norm": 0.5822025537490845,
+      "learning_rate": 7.895032113465986e-05,
+      "loss": 0.3725,
+      "step": 7895
+    },
+    {
+      "epoch": 0.5806309287447606,
+      "grad_norm": 0.5798414349555969,
+      "learning_rate": 7.892703916152489e-05,
+      "loss": 0.3719,
+      "step": 7896
+    },
+    {
+      "epoch": 0.5807044635634974,
+      "grad_norm": 1.0978277921676636,
+      "learning_rate": 7.890375838367052e-05,
+      "loss": 0.3577,
+      "step": 7897
+    },
+    {
+      "epoch": 0.580777998382234,
+      "grad_norm": 0.6961476802825928,
+      "learning_rate": 7.888047880241723e-05,
+      "loss": 0.357,
+      "step": 7898
+    },
+    {
+      "epoch": 0.5808515332009706,
+      "grad_norm": 0.6798830628395081,
+      "learning_rate": 7.885720041908547e-05,
+      "loss": 0.376,
+      "step": 7899
+    },
+    {
+      "epoch": 0.5809250680197073,
+      "grad_norm": 0.780113935470581,
+      "learning_rate": 7.883392323499561e-05,
+      "loss": 0.3648,
+      "step": 7900
+    },
+    {
+      "epoch": 0.580998602838444,
+      "grad_norm": 0.47175320982933044,
+      "learning_rate": 7.881064725146798e-05,
+      "loss": 0.3662,
+      "step": 7901
+    },
+    {
+      "epoch": 0.5810721376571807,
+      "grad_norm": 0.8591800928115845,
+      "learning_rate": 7.878737246982277e-05,
+      "loss": 0.3623,
+      "step": 7902
+    },
+    {
+      "epoch": 0.5811456724759173,
+      "grad_norm": 1.0523569583892822,
+      "learning_rate": 7.87640988913802e-05,
+      "loss": 0.3459,
+      "step": 7903
+    },
+    {
+      "epoch": 0.581219207294654,
+      "grad_norm": 0.4687853157520294,
+      "learning_rate": 7.874082651746033e-05,
+      "loss": 0.3717,
+      "step": 7904
+    },
+    {
+      "epoch": 0.5812927421133907,
+      "grad_norm": 0.9535640478134155,
+      "learning_rate": 7.871755534938329e-05,
+      "loss": 0.3789,
+      "step": 7905
+    },
+    {
+      "epoch": 0.5813662769321274,
+      "grad_norm": 0.6445412039756775,
+      "learning_rate": 7.869428538846891e-05,
+      "loss": 0.3741,
+      "step": 7906
+    },
+    {
+      "epoch": 0.581439811750864,
+      "grad_norm": 0.6554338335990906,
+      "learning_rate": 7.867101663603716e-05,
+      "loss": 0.3726,
+      "step": 7907
+    },
+    {
+      "epoch": 0.5815133465696007,
+      "grad_norm": 0.9287732839584351,
+      "learning_rate": 7.864774909340784e-05,
+      "loss": 0.3699,
+      "step": 7908
+    },
+    {
+      "epoch": 0.5815868813883374,
+      "grad_norm": 0.5982078909873962,
+      "learning_rate": 7.862448276190073e-05,
+      "loss": 0.3816,
+      "step": 7909
+    },
+    {
+      "epoch": 0.5816604162070741,
+      "grad_norm": 1.4350864887237549,
+      "learning_rate": 7.860121764283552e-05,
+      "loss": 0.353,
+      "step": 7910
+    },
+    {
+      "epoch": 0.5817339510258107,
+      "grad_norm": 0.6711572408676147,
+      "learning_rate": 7.857795373753184e-05,
+      "loss": 0.3601,
+      "step": 7911
+    },
+    {
+      "epoch": 0.5818074858445474,
+      "grad_norm": 1.355353832244873,
+      "learning_rate": 7.85546910473092e-05,
+      "loss": 0.3579,
+      "step": 7912
+    },
+    {
+      "epoch": 0.5818810206632841,
+      "grad_norm": 0.7710068225860596,
+      "learning_rate": 7.853142957348714e-05,
+      "loss": 0.3623,
+      "step": 7913
+    },
+    {
+      "epoch": 0.5819545554820207,
+      "grad_norm": 0.7388173341751099,
+      "learning_rate": 7.850816931738504e-05,
+      "loss": 0.375,
+      "step": 7914
+    },
+    {
+      "epoch": 0.5820280903007574,
+      "grad_norm": 1.0937561988830566,
+      "learning_rate": 7.848491028032226e-05,
+      "loss": 0.3589,
+      "step": 7915
+    },
+    {
+      "epoch": 0.582101625119494,
+      "grad_norm": 0.5906800031661987,
+      "learning_rate": 7.846165246361812e-05,
+      "loss": 0.382,
+      "step": 7916
+    },
+    {
+      "epoch": 0.5821751599382308,
+      "grad_norm": 1.581094741821289,
+      "learning_rate": 7.843839586859174e-05,
+      "loss": 0.3537,
+      "step": 7917
+    },
+    {
+      "epoch": 0.5822486947569674,
+      "grad_norm": 0.6620028614997864,
+      "learning_rate": 7.841514049656231e-05,
+      "loss": 0.3694,
+      "step": 7918
+    },
+    {
+      "epoch": 0.5823222295757041,
+      "grad_norm": 1.5853426456451416,
+      "learning_rate": 7.839188634884888e-05,
+      "loss": 0.3809,
+      "step": 7919
+    },
+    {
+      "epoch": 0.5823957643944407,
+      "grad_norm": 1.1306424140930176,
+      "learning_rate": 7.836863342677046e-05,
+      "loss": 0.3953,
+      "step": 7920
+    },
+    {
+      "epoch": 0.5824692992131775,
+      "grad_norm": 1.5273572206497192,
+      "learning_rate": 7.834538173164599e-05,
+      "loss": 0.3819,
+      "step": 7921
+    },
+    {
+      "epoch": 0.5825428340319141,
+      "grad_norm": 1.6183546781539917,
+      "learning_rate": 7.832213126479431e-05,
+      "loss": 0.3739,
+      "step": 7922
+    },
+    {
+      "epoch": 0.5826163688506508,
+      "grad_norm": 0.9244197607040405,
+      "learning_rate": 7.829888202753422e-05,
+      "loss": 0.3652,
+      "step": 7923
+    },
+    {
+      "epoch": 0.5826899036693874,
+      "grad_norm": 0.8411330580711365,
+      "learning_rate": 7.827563402118446e-05,
+      "loss": 0.3677,
+      "step": 7924
+    },
+    {
+      "epoch": 0.5827634384881242,
+      "grad_norm": 0.8185566067695618,
+      "learning_rate": 7.825238724706365e-05,
+      "loss": 0.3975,
+      "step": 7925
+    },
+    {
+      "epoch": 0.5828369733068608,
+      "grad_norm": 0.7996388673782349,
+      "learning_rate": 7.82291417064904e-05,
+      "loss": 0.4051,
+      "step": 7926
+    },
+    {
+      "epoch": 0.5829105081255974,
+      "grad_norm": 1.2547111511230469,
+      "learning_rate": 7.820589740078317e-05,
+      "loss": 0.3598,
+      "step": 7927
+    },
+    {
+      "epoch": 0.5829840429443341,
+      "grad_norm": 0.7523272633552551,
+      "learning_rate": 7.818265433126053e-05,
+      "loss": 0.3756,
+      "step": 7928
+    },
+    {
+      "epoch": 0.5830575777630708,
+      "grad_norm": 1.4698442220687866,
+      "learning_rate": 7.815941249924069e-05,
+      "loss": 0.3454,
+      "step": 7929
+    },
+    {
+      "epoch": 0.5831311125818075,
+      "grad_norm": 1.0817277431488037,
+      "learning_rate": 7.813617190604204e-05,
+      "loss": 0.3817,
+      "step": 7930
+    },
+    {
+      "epoch": 0.5832046474005441,
+      "grad_norm": 1.6882094144821167,
+      "learning_rate": 7.811293255298279e-05,
+      "loss": 0.3722,
+      "step": 7931
+    },
+    {
+      "epoch": 0.5832781822192808,
+      "grad_norm": 1.6201244592666626,
+      "learning_rate": 7.808969444138109e-05,
+      "loss": 0.3467,
+      "step": 7932
+    },
+    {
+      "epoch": 0.5833517170380175,
+      "grad_norm": 0.8159083724021912,
+      "learning_rate": 7.806645757255506e-05,
+      "loss": 0.3886,
+      "step": 7933
+    },
+    {
+      "epoch": 0.5834252518567542,
+      "grad_norm": 1.3346034288406372,
+      "learning_rate": 7.804322194782271e-05,
+      "loss": 0.335,
+      "step": 7934
+    },
+    {
+      "epoch": 0.5834987866754908,
+      "grad_norm": 0.667958676815033,
+      "learning_rate": 7.801998756850195e-05,
+      "loss": 0.3614,
+      "step": 7935
+    },
+    {
+      "epoch": 0.5835723214942276,
+      "grad_norm": 1.1388239860534668,
+      "learning_rate": 7.799675443591072e-05,
+      "loss": 0.3583,
+      "step": 7936
+    },
+    {
+      "epoch": 0.5836458563129642,
+      "grad_norm": 0.982695460319519,
+      "learning_rate": 7.797352255136678e-05,
+      "loss": 0.3667,
+      "step": 7937
+    },
+    {
+      "epoch": 0.5837193911317009,
+      "grad_norm": 0.5001276731491089,
+      "learning_rate": 7.795029191618792e-05,
+      "loss": 0.341,
+      "step": 7938
+    },
+    {
+      "epoch": 0.5837929259504375,
+      "grad_norm": 0.7822884917259216,
+      "learning_rate": 7.792706253169175e-05,
+      "loss": 0.3609,
+      "step": 7939
+    },
+    {
+      "epoch": 0.5838664607691743,
+      "grad_norm": 0.49519696831703186,
+      "learning_rate": 7.790383439919591e-05,
+      "loss": 0.373,
+      "step": 7940
+    },
+    {
+      "epoch": 0.5839399955879109,
+      "grad_norm": 1.0268676280975342,
+      "learning_rate": 7.788060752001789e-05,
+      "loss": 0.3469,
+      "step": 7941
+    },
+    {
+      "epoch": 0.5840135304066475,
+      "grad_norm": 0.7917190194129944,
+      "learning_rate": 7.785738189547513e-05,
+      "loss": 0.3759,
+      "step": 7942
+    },
+    {
+      "epoch": 0.5840870652253842,
+      "grad_norm": 0.902504563331604,
+      "learning_rate": 7.783415752688506e-05,
+      "loss": 0.3667,
+      "step": 7943
+    },
+    {
+      "epoch": 0.5841606000441208,
+      "grad_norm": 1.3853756189346313,
+      "learning_rate": 7.781093441556497e-05,
+      "loss": 0.3511,
+      "step": 7944
+    },
+    {
+      "epoch": 0.5842341348628576,
+      "grad_norm": 0.7292165160179138,
+      "learning_rate": 7.778771256283208e-05,
+      "loss": 0.356,
+      "step": 7945
+    },
+    {
+      "epoch": 0.5843076696815942,
+      "grad_norm": 1.5143500566482544,
+      "learning_rate": 7.776449197000359e-05,
+      "loss": 0.3406,
+      "step": 7946
+    },
+    {
+      "epoch": 0.5843812045003309,
+      "grad_norm": 0.8999934196472168,
+      "learning_rate": 7.774127263839657e-05,
+      "loss": 0.3484,
+      "step": 7947
+    },
+    {
+      "epoch": 0.5844547393190676,
+      "grad_norm": 2.051689863204956,
+      "learning_rate": 7.771805456932806e-05,
+      "loss": 0.3628,
+      "step": 7948
+    },
+    {
+      "epoch": 0.5845282741378043,
+      "grad_norm": 1.8291244506835938,
+      "learning_rate": 7.769483776411502e-05,
+      "loss": 0.3942,
+      "step": 7949
+    },
+    {
+      "epoch": 0.5846018089565409,
+      "grad_norm": 0.857469379901886,
+      "learning_rate": 7.767162222407429e-05,
+      "loss": 0.3483,
+      "step": 7950
+    },
+    {
+      "epoch": 0.5846753437752776,
+      "grad_norm": 1.290911078453064,
+      "learning_rate": 7.764840795052278e-05,
+      "loss": 0.3618,
+      "step": 7951
+    },
+    {
+      "epoch": 0.5847488785940143,
+      "grad_norm": 0.7161484956741333,
+      "learning_rate": 7.762519494477711e-05,
+      "loss": 0.3614,
+      "step": 7952
+    },
+    {
+      "epoch": 0.584822413412751,
+      "grad_norm": 1.3883317708969116,
+      "learning_rate": 7.760198320815399e-05,
+      "loss": 0.3768,
+      "step": 7953
+    },
+    {
+      "epoch": 0.5848959482314876,
+      "grad_norm": 0.7918971180915833,
+      "learning_rate": 7.757877274197006e-05,
+      "loss": 0.3561,
+      "step": 7954
+    },
+    {
+      "epoch": 0.5849694830502242,
+      "grad_norm": 1.2547792196273804,
+      "learning_rate": 7.755556354754175e-05,
+      "loss": 0.3412,
+      "step": 7955
+    },
+    {
+      "epoch": 0.585043017868961,
+      "grad_norm": 0.7687346339225769,
+      "learning_rate": 7.753235562618558e-05,
+      "loss": 0.3677,
+      "step": 7956
+    },
+    {
+      "epoch": 0.5851165526876976,
+      "grad_norm": 1.075657844543457,
+      "learning_rate": 7.750914897921792e-05,
+      "loss": 0.3666,
+      "step": 7957
+    },
+    {
+      "epoch": 0.5851900875064343,
+      "grad_norm": 0.7096709609031677,
+      "learning_rate": 7.748594360795505e-05,
+      "loss": 0.3517,
+      "step": 7958
+    },
+    {
+      "epoch": 0.5852636223251709,
+      "grad_norm": 1.1270078420639038,
+      "learning_rate": 7.746273951371323e-05,
+      "loss": 0.3581,
+      "step": 7959
+    },
+    {
+      "epoch": 0.5853371571439077,
+      "grad_norm": 0.7164536714553833,
+      "learning_rate": 7.74395366978086e-05,
+      "loss": 0.3634,
+      "step": 7960
+    },
+    {
+      "epoch": 0.5854106919626443,
+      "grad_norm": 0.7525386214256287,
+      "learning_rate": 7.741633516155726e-05,
+      "loss": 0.3471,
+      "step": 7961
+    },
+    {
+      "epoch": 0.585484226781381,
+      "grad_norm": 0.8836906552314758,
+      "learning_rate": 7.739313490627525e-05,
+      "loss": 0.3772,
+      "step": 7962
+    },
+    {
+      "epoch": 0.5855577616001176,
+      "grad_norm": 0.5003070831298828,
+      "learning_rate": 7.736993593327848e-05,
+      "loss": 0.3774,
+      "step": 7963
+    },
+    {
+      "epoch": 0.5856312964188544,
+      "grad_norm": 0.8313140869140625,
+      "learning_rate": 7.734673824388282e-05,
+      "loss": 0.3526,
+      "step": 7964
+    },
+    {
+      "epoch": 0.585704831237591,
+      "grad_norm": 0.7142065167427063,
+      "learning_rate": 7.732354183940405e-05,
+      "loss": 0.358,
+      "step": 7965
+    },
+    {
+      "epoch": 0.5857783660563277,
+      "grad_norm": 0.4974057972431183,
+      "learning_rate": 7.730034672115794e-05,
+      "loss": 0.3593,
+      "step": 7966
+    },
+    {
+      "epoch": 0.5858519008750643,
+      "grad_norm": 0.48272940516471863,
+      "learning_rate": 7.727715289046011e-05,
+      "loss": 0.3642,
+      "step": 7967
+    },
+    {
+      "epoch": 0.5859254356938011,
+      "grad_norm": 0.6494429111480713,
+      "learning_rate": 7.725396034862614e-05,
+      "loss": 0.3556,
+      "step": 7968
+    },
+    {
+      "epoch": 0.5859989705125377,
+      "grad_norm": 0.6384482383728027,
+      "learning_rate": 7.723076909697155e-05,
+      "loss": 0.3751,
+      "step": 7969
+    },
+    {
+      "epoch": 0.5860725053312743,
+      "grad_norm": 0.6608214378356934,
+      "learning_rate": 7.72075791368118e-05,
+      "loss": 0.3596,
+      "step": 7970
+    },
+    {
+      "epoch": 0.586146040150011,
+      "grad_norm": 0.5962188243865967,
+      "learning_rate": 7.718439046946215e-05,
+      "loss": 0.334,
+      "step": 7971
+    },
+    {
+      "epoch": 0.5862195749687477,
+      "grad_norm": 0.7549698948860168,
+      "learning_rate": 7.7161203096238e-05,
+      "loss": 0.3448,
+      "step": 7972
+    },
+    {
+      "epoch": 0.5862931097874844,
+      "grad_norm": 0.5648935437202454,
+      "learning_rate": 7.71380170184545e-05,
+      "loss": 0.391,
+      "step": 7973
+    },
+    {
+      "epoch": 0.586366644606221,
+      "grad_norm": 0.8171812295913696,
+      "learning_rate": 7.71148322374268e-05,
+      "loss": 0.3411,
+      "step": 7974
+    },
+    {
+      "epoch": 0.5864401794249577,
+      "grad_norm": 0.5935189723968506,
+      "learning_rate": 7.709164875447e-05,
+      "loss": 0.3905,
+      "step": 7975
+    },
+    {
+      "epoch": 0.5865137142436944,
+      "grad_norm": 0.6696495413780212,
+      "learning_rate": 7.706846657089902e-05,
+      "loss": 0.3649,
+      "step": 7976
+    },
+    {
+      "epoch": 0.5865872490624311,
+      "grad_norm": 0.6371015906333923,
+      "learning_rate": 7.704528568802884e-05,
+      "loss": 0.3774,
+      "step": 7977
+    },
+    {
+      "epoch": 0.5866607838811677,
+      "grad_norm": 0.6274473667144775,
+      "learning_rate": 7.702210610717424e-05,
+      "loss": 0.3637,
+      "step": 7978
+    },
+    {
+      "epoch": 0.5867343186999044,
+      "grad_norm": 0.85896897315979,
+      "learning_rate": 7.699892782965008e-05,
+      "loss": 0.3704,
+      "step": 7979
+    },
+    {
+      "epoch": 0.5868078535186411,
+      "grad_norm": 0.6312841176986694,
+      "learning_rate": 7.697575085677098e-05,
+      "loss": 0.3711,
+      "step": 7980
+    },
+    {
+      "epoch": 0.5868813883373778,
+      "grad_norm": 0.889290988445282,
+      "learning_rate": 7.695257518985158e-05,
+      "loss": 0.3451,
+      "step": 7981
+    },
+    {
+      "epoch": 0.5869549231561144,
+      "grad_norm": 1.2061396837234497,
+      "learning_rate": 7.692940083020645e-05,
+      "loss": 0.3677,
+      "step": 7982
+    },
+    {
+      "epoch": 0.587028457974851,
+      "grad_norm": 0.45513370633125305,
+      "learning_rate": 7.690622777915006e-05,
+      "loss": 0.3661,
+      "step": 7983
+    },
+    {
+      "epoch": 0.5871019927935878,
+      "grad_norm": 1.25619637966156,
+      "learning_rate": 7.68830560379968e-05,
+      "loss": 0.3575,
+      "step": 7984
+    },
+    {
+      "epoch": 0.5871755276123244,
+      "grad_norm": 0.48157745599746704,
+      "learning_rate": 7.685988560806098e-05,
+      "loss": 0.365,
+      "step": 7985
+    },
+    {
+      "epoch": 0.5872490624310611,
+      "grad_norm": 1.2179251909255981,
+      "learning_rate": 7.683671649065692e-05,
+      "loss": 0.3739,
+      "step": 7986
+    },
+    {
+      "epoch": 0.5873225972497977,
+      "grad_norm": 0.6798186898231506,
+      "learning_rate": 7.681354868709871e-05,
+      "loss": 0.3695,
+      "step": 7987
+    },
+    {
+      "epoch": 0.5873961320685345,
+      "grad_norm": 0.6564776301383972,
+      "learning_rate": 7.679038219870048e-05,
+      "loss": 0.3567,
+      "step": 7988
+    },
+    {
+      "epoch": 0.5874696668872711,
+      "grad_norm": 0.6051353216171265,
+      "learning_rate": 7.676721702677628e-05,
+      "loss": 0.3631,
+      "step": 7989
+    },
+    {
+      "epoch": 0.5875432017060078,
+      "grad_norm": 0.7147058844566345,
+      "learning_rate": 7.674405317264005e-05,
+      "loss": 0.3516,
+      "step": 7990
+    },
+    {
+      "epoch": 0.5876167365247444,
+      "grad_norm": 0.7338356971740723,
+      "learning_rate": 7.672089063760564e-05,
+      "loss": 0.3744,
+      "step": 7991
+    },
+    {
+      "epoch": 0.5876902713434812,
+      "grad_norm": 0.5722053050994873,
+      "learning_rate": 7.669772942298691e-05,
+      "loss": 0.3725,
+      "step": 7992
+    },
+    {
+      "epoch": 0.5877638061622178,
+      "grad_norm": 1.0266309976577759,
+      "learning_rate": 7.667456953009755e-05,
+      "loss": 0.3585,
+      "step": 7993
+    },
+    {
+      "epoch": 0.5878373409809545,
+      "grad_norm": 0.6099303960800171,
+      "learning_rate": 7.66514109602512e-05,
+      "loss": 0.325,
+      "step": 7994
+    },
+    {
+      "epoch": 0.5879108757996911,
+      "grad_norm": 0.7801576852798462,
+      "learning_rate": 7.66282537147615e-05,
+      "loss": 0.3489,
+      "step": 7995
+    },
+    {
+      "epoch": 0.5879844106184279,
+      "grad_norm": 1.3784855604171753,
+      "learning_rate": 7.660509779494187e-05,
+      "loss": 0.3514,
+      "step": 7996
+    },
+    {
+      "epoch": 0.5880579454371645,
+      "grad_norm": 0.6588388681411743,
+      "learning_rate": 7.658194320210582e-05,
+      "loss": 0.3703,
+      "step": 7997
+    },
+    {
+      "epoch": 0.5881314802559011,
+      "grad_norm": 1.4542032480239868,
+      "learning_rate": 7.655878993756669e-05,
+      "loss": 0.3447,
+      "step": 7998
+    },
+    {
+      "epoch": 0.5882050150746378,
+      "grad_norm": 0.7762621641159058,
+      "learning_rate": 7.65356380026377e-05,
+      "loss": 0.3514,
+      "step": 7999
+    },
+    {
+      "epoch": 0.5882785498933745,
+      "grad_norm": 1.5217646360397339,
+      "learning_rate": 7.651248739863208e-05,
+      "loss": 0.3678,
+      "step": 8000
+    },
+    {
+      "epoch": 0.5883520847121112,
+      "grad_norm": 1.1571040153503418,
+      "learning_rate": 7.648933812686297e-05,
+      "loss": 0.3428,
+      "step": 8001
+    },
+    {
+      "epoch": 0.5884256195308478,
+      "grad_norm": 1.5211336612701416,
+      "learning_rate": 7.646619018864341e-05,
+      "loss": 0.3449,
+      "step": 8002
+    },
+    {
+      "epoch": 0.5884991543495846,
+      "grad_norm": 1.5511574745178223,
+      "learning_rate": 7.644304358528638e-05,
+      "loss": 0.3641,
+      "step": 8003
+    },
+    {
+      "epoch": 0.5885726891683212,
+      "grad_norm": 1.3498018980026245,
+      "learning_rate": 7.641989831810476e-05,
+      "loss": 0.3425,
+      "step": 8004
+    },
+    {
+      "epoch": 0.5886462239870579,
+      "grad_norm": 1.2019609212875366,
+      "learning_rate": 7.63967543884114e-05,
+      "loss": 0.3674,
+      "step": 8005
+    },
+    {
+      "epoch": 0.5887197588057945,
+      "grad_norm": 1.5852868556976318,
+      "learning_rate": 7.637361179751907e-05,
+      "loss": 0.3554,
+      "step": 8006
+    },
+    {
+      "epoch": 0.5887932936245313,
+      "grad_norm": 1.6830224990844727,
+      "learning_rate": 7.635047054674037e-05,
+      "loss": 0.3489,
+      "step": 8007
+    },
+    {
+      "epoch": 0.5888668284432679,
+      "grad_norm": 1.035674810409546,
+      "learning_rate": 7.632733063738797e-05,
+      "loss": 0.3738,
+      "step": 8008
+    },
+    {
+      "epoch": 0.5889403632620046,
+      "grad_norm": 1.2226834297180176,
+      "learning_rate": 7.630419207077434e-05,
+      "loss": 0.3629,
+      "step": 8009
+    },
+    {
+      "epoch": 0.5890138980807412,
+      "grad_norm": 0.8698837161064148,
+      "learning_rate": 7.6281054848212e-05,
+      "loss": 0.3356,
+      "step": 8010
+    },
+    {
+      "epoch": 0.5890874328994778,
+      "grad_norm": 0.9657955765724182,
+      "learning_rate": 7.62579189710132e-05,
+      "loss": 0.3677,
+      "step": 8011
+    },
+    {
+      "epoch": 0.5891609677182146,
+      "grad_norm": 1.2080270051956177,
+      "learning_rate": 7.623478444049031e-05,
+      "loss": 0.3541,
+      "step": 8012
+    },
+    {
+      "epoch": 0.5892345025369512,
+      "grad_norm": 0.6750339865684509,
+      "learning_rate": 7.621165125795553e-05,
+      "loss": 0.3509,
+      "step": 8013
+    },
+    {
+      "epoch": 0.5893080373556879,
+      "grad_norm": 1.4617353677749634,
+      "learning_rate": 7.618851942472096e-05,
+      "loss": 0.374,
+      "step": 8014
+    },
+    {
+      "epoch": 0.5893815721744246,
+      "grad_norm": 0.7898066639900208,
+      "learning_rate": 7.616538894209874e-05,
+      "loss": 0.3416,
+      "step": 8015
+    },
+    {
+      "epoch": 0.5894551069931613,
+      "grad_norm": 1.6650078296661377,
+      "learning_rate": 7.61422598114008e-05,
+      "loss": 0.337,
+      "step": 8016
+    },
+    {
+      "epoch": 0.5895286418118979,
+      "grad_norm": 1.123032808303833,
+      "learning_rate": 7.611913203393905e-05,
+      "loss": 0.3609,
+      "step": 8017
+    },
+    {
+      "epoch": 0.5896021766306346,
+      "grad_norm": 2.149937152862549,
+      "learning_rate": 7.609600561102534e-05,
+      "loss": 0.3394,
+      "step": 8018
+    },
+    {
+      "epoch": 0.5896757114493713,
+      "grad_norm": 1.9463454484939575,
+      "learning_rate": 7.607288054397141e-05,
+      "loss": 0.3406,
+      "step": 8019
+    },
+    {
+      "epoch": 0.589749246268108,
+      "grad_norm": 0.8876090049743652,
+      "learning_rate": 7.604975683408897e-05,
+      "loss": 0.3269,
+      "step": 8020
+    },
+    {
+      "epoch": 0.5898227810868446,
+      "grad_norm": 1.0842511653900146,
+      "learning_rate": 7.602663448268961e-05,
+      "loss": 0.3407,
+      "step": 8021
+    },
+    {
+      "epoch": 0.5898963159055813,
+      "grad_norm": 0.9440042972564697,
+      "learning_rate": 7.600351349108483e-05,
+      "loss": 0.388,
+      "step": 8022
+    },
+    {
+      "epoch": 0.589969850724318,
+      "grad_norm": 0.8076239228248596,
+      "learning_rate": 7.598039386058609e-05,
+      "loss": 0.3549,
+      "step": 8023
+    },
+    {
+      "epoch": 0.5900433855430547,
+      "grad_norm": 1.038030982017517,
+      "learning_rate": 7.595727559250473e-05,
+      "loss": 0.3392,
+      "step": 8024
+    },
+    {
+      "epoch": 0.5901169203617913,
+      "grad_norm": 0.9829807281494141,
+      "learning_rate": 7.593415868815209e-05,
+      "loss": 0.3395,
+      "step": 8025
+    },
+    {
+      "epoch": 0.5901904551805279,
+      "grad_norm": 1.397996425628662,
+      "learning_rate": 7.591104314883937e-05,
+      "loss": 0.3387,
+      "step": 8026
+    },
+    {
+      "epoch": 0.5902639899992647,
+      "grad_norm": 0.9924814105033875,
+      "learning_rate": 7.588792897587768e-05,
+      "loss": 0.3359,
+      "step": 8027
+    },
+    {
+      "epoch": 0.5903375248180013,
+      "grad_norm": 1.7686562538146973,
+      "learning_rate": 7.586481617057815e-05,
+      "loss": 0.3803,
+      "step": 8028
+    },
+    {
+      "epoch": 0.590411059636738,
+      "grad_norm": 1.9678699970245361,
+      "learning_rate": 7.584170473425169e-05,
+      "loss": 0.3684,
+      "step": 8029
+    },
+    {
+      "epoch": 0.5904845944554746,
+      "grad_norm": 0.82186359167099,
+      "learning_rate": 7.581859466820922e-05,
+      "loss": 0.3281,
+      "step": 8030
+    },
+    {
+      "epoch": 0.5905581292742114,
+      "grad_norm": 1.1043391227722168,
+      "learning_rate": 7.57954859737616e-05,
+      "loss": 0.3488,
+      "step": 8031
+    },
+    {
+      "epoch": 0.590631664092948,
+      "grad_norm": 0.7103190422058105,
+      "learning_rate": 7.577237865221954e-05,
+      "loss": 0.3461,
+      "step": 8032
+    },
+    {
+      "epoch": 0.5907051989116847,
+      "grad_norm": 0.8563942909240723,
+      "learning_rate": 7.57492727048938e-05,
+      "loss": 0.3406,
+      "step": 8033
+    },
+    {
+      "epoch": 0.5907787337304213,
+      "grad_norm": 0.7966180443763733,
+      "learning_rate": 7.572616813309484e-05,
+      "loss": 0.3398,
+      "step": 8034
+    },
+    {
+      "epoch": 0.5908522685491581,
+      "grad_norm": 0.5967215895652771,
+      "learning_rate": 7.570306493813326e-05,
+      "loss": 0.3363,
+      "step": 8035
+    },
+    {
+      "epoch": 0.5909258033678947,
+      "grad_norm": 0.8825503587722778,
+      "learning_rate": 7.567996312131948e-05,
+      "loss": 0.3159,
+      "step": 8036
+    },
+    {
+      "epoch": 0.5909993381866314,
+      "grad_norm": 0.8002889752388,
+      "learning_rate": 7.565686268396382e-05,
+      "loss": 0.3247,
+      "step": 8037
+    },
+    {
+      "epoch": 0.591072873005368,
+      "grad_norm": 1.3011163473129272,
+      "learning_rate": 7.563376362737664e-05,
+      "loss": 0.3346,
+      "step": 8038
+    },
+    {
+      "epoch": 0.5911464078241047,
+      "grad_norm": 0.8248789310455322,
+      "learning_rate": 7.56106659528681e-05,
+      "loss": 0.3115,
+      "step": 8039
+    },
+    {
+      "epoch": 0.5912199426428414,
+      "grad_norm": 1.3986241817474365,
+      "learning_rate": 7.55875696617483e-05,
+      "loss": 0.3903,
+      "step": 8040
+    },
+    {
+      "epoch": 0.591293477461578,
+      "grad_norm": 1.049381136894226,
+      "learning_rate": 7.556447475532733e-05,
+      "loss": 0.3544,
+      "step": 8041
+    },
+    {
+      "epoch": 0.5913670122803147,
+      "grad_norm": 1.2094253301620483,
+      "learning_rate": 7.554138123491512e-05,
+      "loss": 0.3651,
+      "step": 8042
+    },
+    {
+      "epoch": 0.5914405470990514,
+      "grad_norm": 1.1242952346801758,
+      "learning_rate": 7.55182891018216e-05,
+      "loss": 0.3414,
+      "step": 8043
+    },
+    {
+      "epoch": 0.5915140819177881,
+      "grad_norm": 1.3271981477737427,
+      "learning_rate": 7.549519835735656e-05,
+      "loss": 0.3437,
+      "step": 8044
+    },
+    {
+      "epoch": 0.5915876167365247,
+      "grad_norm": 1.071462869644165,
+      "learning_rate": 7.547210900282976e-05,
+      "loss": 0.351,
+      "step": 8045
+    },
+    {
+      "epoch": 0.5916611515552614,
+      "grad_norm": 1.2564846277236938,
+      "learning_rate": 7.544902103955079e-05,
+      "loss": 0.3383,
+      "step": 8046
+    },
+    {
+      "epoch": 0.5917346863739981,
+      "grad_norm": 1.0747278928756714,
+      "learning_rate": 7.542593446882924e-05,
+      "loss": 0.3537,
+      "step": 8047
+    },
+    {
+      "epoch": 0.5918082211927348,
+      "grad_norm": 1.0940550565719604,
+      "learning_rate": 7.540284929197465e-05,
+      "loss": 0.3673,
+      "step": 8048
+    },
+    {
+      "epoch": 0.5918817560114714,
+      "grad_norm": 1.1997473239898682,
+      "learning_rate": 7.537976551029639e-05,
+      "loss": 0.3683,
+      "step": 8049
+    },
+    {
+      "epoch": 0.5919552908302081,
+      "grad_norm": 1.4027988910675049,
+      "learning_rate": 7.53566831251038e-05,
+      "loss": 0.3498,
+      "step": 8050
+    },
+    {
+      "epoch": 0.5920288256489448,
+      "grad_norm": 1.3766862154006958,
+      "learning_rate": 7.533360213770617e-05,
+      "loss": 0.3286,
+      "step": 8051
+    },
+    {
+      "epoch": 0.5921023604676815,
+      "grad_norm": 0.9734534025192261,
+      "learning_rate": 7.531052254941266e-05,
+      "loss": 0.3804,
+      "step": 8052
+    },
+    {
+      "epoch": 0.5921758952864181,
+      "grad_norm": 1.226743221282959,
+      "learning_rate": 7.528744436153234e-05,
+      "loss": 0.3162,
+      "step": 8053
+    },
+    {
+      "epoch": 0.5922494301051547,
+      "grad_norm": 0.8048805594444275,
+      "learning_rate": 7.526436757537429e-05,
+      "loss": 0.3538,
+      "step": 8054
+    },
+    {
+      "epoch": 0.5923229649238915,
+      "grad_norm": 0.9351117014884949,
+      "learning_rate": 7.524129219224739e-05,
+      "loss": 0.3615,
+      "step": 8055
+    },
+    {
+      "epoch": 0.5923964997426281,
+      "grad_norm": 0.9732881784439087,
+      "learning_rate": 7.521821821346054e-05,
+      "loss": 0.3791,
+      "step": 8056
+    },
+    {
+      "epoch": 0.5924700345613648,
+      "grad_norm": 0.4898485541343689,
+      "learning_rate": 7.519514564032253e-05,
+      "loss": 0.3415,
+      "step": 8057
+    },
+    {
+      "epoch": 0.5925435693801014,
+      "grad_norm": 0.8546133637428284,
+      "learning_rate": 7.5172074474142e-05,
+      "loss": 0.337,
+      "step": 8058
+    },
+    {
+      "epoch": 0.5926171041988382,
+      "grad_norm": 0.4625721573829651,
+      "learning_rate": 7.514900471622763e-05,
+      "loss": 0.3356,
+      "step": 8059
+    },
+    {
+      "epoch": 0.5926906390175748,
+      "grad_norm": 0.8417932987213135,
+      "learning_rate": 7.512593636788791e-05,
+      "loss": 0.3333,
+      "step": 8060
+    },
+    {
+      "epoch": 0.5927641738363115,
+      "grad_norm": 0.4352606236934662,
+      "learning_rate": 7.510286943043136e-05,
+      "loss": 0.3511,
+      "step": 8061
+    },
+    {
+      "epoch": 0.5928377086550481,
+      "grad_norm": 1.026644229888916,
+      "learning_rate": 7.507980390516633e-05,
+      "loss": 0.3179,
+      "step": 8062
+    },
+    {
+      "epoch": 0.5929112434737849,
+      "grad_norm": 0.5216363072395325,
+      "learning_rate": 7.50567397934011e-05,
+      "loss": 0.3571,
+      "step": 8063
+    },
+    {
+      "epoch": 0.5929847782925215,
+      "grad_norm": 0.850949764251709,
+      "learning_rate": 7.50336770964439e-05,
+      "loss": 0.3518,
+      "step": 8064
+    },
+    {
+      "epoch": 0.5930583131112582,
+      "grad_norm": 0.8114153146743774,
+      "learning_rate": 7.50106158156029e-05,
+      "loss": 0.3494,
+      "step": 8065
+    },
+    {
+      "epoch": 0.5931318479299948,
+      "grad_norm": 0.687869668006897,
+      "learning_rate": 7.498755595218616e-05,
+      "loss": 0.3389,
+      "step": 8066
+    },
+    {
+      "epoch": 0.5932053827487315,
+      "grad_norm": 1.1054757833480835,
+      "learning_rate": 7.496449750750164e-05,
+      "loss": 0.3453,
+      "step": 8067
+    },
+    {
+      "epoch": 0.5932789175674682,
+      "grad_norm": 0.5508403778076172,
+      "learning_rate": 7.494144048285726e-05,
+      "loss": 0.332,
+      "step": 8068
+    },
+    {
+      "epoch": 0.5933524523862048,
+      "grad_norm": 1.205690860748291,
+      "learning_rate": 7.49183848795608e-05,
+      "loss": 0.3572,
+      "step": 8069
+    },
+    {
+      "epoch": 0.5934259872049416,
+      "grad_norm": 0.6248819231987,
+      "learning_rate": 7.489533069892001e-05,
+      "loss": 0.3603,
+      "step": 8070
+    },
+    {
+      "epoch": 0.5934995220236782,
+      "grad_norm": 1.1153548955917358,
+      "learning_rate": 7.487227794224257e-05,
+      "loss": 0.348,
+      "step": 8071
+    },
+    {
+      "epoch": 0.5935730568424149,
+      "grad_norm": 0.534546434879303,
+      "learning_rate": 7.484922661083603e-05,
+      "loss": 0.3509,
+      "step": 8072
+    },
+    {
+      "epoch": 0.5936465916611515,
+      "grad_norm": 1.3348228931427002,
+      "learning_rate": 7.482617670600789e-05,
+      "loss": 0.3603,
+      "step": 8073
+    },
+    {
+      "epoch": 0.5937201264798883,
+      "grad_norm": 0.47766679525375366,
+      "learning_rate": 7.480312822906559e-05,
+      "loss": 0.3593,
+      "step": 8074
+    },
+    {
+      "epoch": 0.5937936612986249,
+      "grad_norm": 0.8593456149101257,
+      "learning_rate": 7.478008118131646e-05,
+      "loss": 0.3548,
+      "step": 8075
+    },
+    {
+      "epoch": 0.5938671961173616,
+      "grad_norm": 0.6130463480949402,
+      "learning_rate": 7.47570355640677e-05,
+      "loss": 0.3718,
+      "step": 8076
+    },
+    {
+      "epoch": 0.5939407309360982,
+      "grad_norm": 0.9347648024559021,
+      "learning_rate": 7.473399137862656e-05,
+      "loss": 0.3752,
+      "step": 8077
+    },
+    {
+      "epoch": 0.594014265754835,
+      "grad_norm": 0.7319321036338806,
+      "learning_rate": 7.471094862630007e-05,
+      "loss": 0.3558,
+      "step": 8078
+    },
+    {
+      "epoch": 0.5940878005735716,
+      "grad_norm": 0.5560859441757202,
+      "learning_rate": 7.468790730839528e-05,
+      "loss": 0.3384,
+      "step": 8079
+    },
+    {
+      "epoch": 0.5941613353923083,
+      "grad_norm": 0.9075961112976074,
+      "learning_rate": 7.466486742621913e-05,
+      "loss": 0.3642,
+      "step": 8080
+    },
+    {
+      "epoch": 0.5942348702110449,
+      "grad_norm": 0.5779425501823425,
+      "learning_rate": 7.464182898107841e-05,
+      "loss": 0.3587,
+      "step": 8081
+    },
+    {
+      "epoch": 0.5943084050297816,
+      "grad_norm": 0.6059698462486267,
+      "learning_rate": 7.46187919742799e-05,
+      "loss": 0.3387,
+      "step": 8082
+    },
+    {
+      "epoch": 0.5943819398485183,
+      "grad_norm": 0.7878549098968506,
+      "learning_rate": 7.459575640713029e-05,
+      "loss": 0.3671,
+      "step": 8083
+    },
+    {
+      "epoch": 0.5944554746672549,
+      "grad_norm": 0.4489938020706177,
+      "learning_rate": 7.457272228093621e-05,
+      "loss": 0.3294,
+      "step": 8084
+    },
+    {
+      "epoch": 0.5945290094859916,
+      "grad_norm": 0.535186231136322,
+      "learning_rate": 7.454968959700414e-05,
+      "loss": 0.3372,
+      "step": 8085
+    },
+    {
+      "epoch": 0.5946025443047283,
+      "grad_norm": 0.5254487991333008,
+      "learning_rate": 7.452665835664052e-05,
+      "loss": 0.3401,
+      "step": 8086
+    },
+    {
+      "epoch": 0.594676079123465,
+      "grad_norm": 0.6448150277137756,
+      "learning_rate": 7.450362856115175e-05,
+      "loss": 0.359,
+      "step": 8087
+    },
+    {
+      "epoch": 0.5947496139422016,
+      "grad_norm": 0.5135288834571838,
+      "learning_rate": 7.448060021184407e-05,
+      "loss": 0.3545,
+      "step": 8088
+    },
+    {
+      "epoch": 0.5948231487609383,
+      "grad_norm": 0.5218680500984192,
+      "learning_rate": 7.445757331002365e-05,
+      "loss": 0.3388,
+      "step": 8089
+    },
+    {
+      "epoch": 0.594896683579675,
+      "grad_norm": 0.5282544493675232,
+      "learning_rate": 7.443454785699667e-05,
+      "loss": 0.3644,
+      "step": 8090
+    },
+    {
+      "epoch": 0.5949702183984117,
+      "grad_norm": 0.4556291401386261,
+      "learning_rate": 7.441152385406908e-05,
+      "loss": 0.353,
+      "step": 8091
+    },
+    {
+      "epoch": 0.5950437532171483,
+      "grad_norm": 0.4951363503932953,
+      "learning_rate": 7.438850130254693e-05,
+      "loss": 0.339,
+      "step": 8092
+    },
+    {
+      "epoch": 0.595117288035885,
+      "grad_norm": 0.46346867084503174,
+      "learning_rate": 7.436548020373594e-05,
+      "loss": 0.3427,
+      "step": 8093
+    },
+    {
+      "epoch": 0.5951908228546217,
+      "grad_norm": 0.5590572357177734,
+      "learning_rate": 7.434246055894199e-05,
+      "loss": 0.3867,
+      "step": 8094
+    },
+    {
+      "epoch": 0.5952643576733583,
+      "grad_norm": 0.5127354264259338,
+      "learning_rate": 7.431944236947077e-05,
+      "loss": 0.3275,
+      "step": 8095
+    },
+    {
+      "epoch": 0.595337892492095,
+      "grad_norm": 0.7629472613334656,
+      "learning_rate": 7.429642563662784e-05,
+      "loss": 0.3804,
+      "step": 8096
+    },
+    {
+      "epoch": 0.5954114273108316,
+      "grad_norm": 0.47217097878456116,
+      "learning_rate": 7.42734103617188e-05,
+      "loss": 0.3756,
+      "step": 8097
+    },
+    {
+      "epoch": 0.5954849621295684,
+      "grad_norm": 0.8962590098381042,
+      "learning_rate": 7.425039654604908e-05,
+      "loss": 0.3655,
+      "step": 8098
+    },
+    {
+      "epoch": 0.595558496948305,
+      "grad_norm": 0.7494411468505859,
+      "learning_rate": 7.422738419092403e-05,
+      "loss": 0.3528,
+      "step": 8099
+    },
+    {
+      "epoch": 0.5956320317670417,
+      "grad_norm": 0.6800077557563782,
+      "learning_rate": 7.420437329764896e-05,
+      "loss": 0.3684,
+      "step": 8100
+    },
+    {
+      "epoch": 0.5957055665857783,
+      "grad_norm": 0.9809805154800415,
+      "learning_rate": 7.418136386752903e-05,
+      "loss": 0.349,
+      "step": 8101
+    },
+    {
+      "epoch": 0.5957791014045151,
+      "grad_norm": 0.7965381741523743,
+      "learning_rate": 7.415835590186941e-05,
+      "loss": 0.3597,
+      "step": 8102
+    },
+    {
+      "epoch": 0.5958526362232517,
+      "grad_norm": 0.7120128273963928,
+      "learning_rate": 7.413534940197515e-05,
+      "loss": 0.3722,
+      "step": 8103
+    },
+    {
+      "epoch": 0.5959261710419884,
+      "grad_norm": 0.738084614276886,
+      "learning_rate": 7.411234436915114e-05,
+      "loss": 0.3587,
+      "step": 8104
+    },
+    {
+      "epoch": 0.595999705860725,
+      "grad_norm": 0.6092620491981506,
+      "learning_rate": 7.408934080470229e-05,
+      "loss": 0.3613,
+      "step": 8105
+    },
+    {
+      "epoch": 0.5960732406794618,
+      "grad_norm": 1.0174109935760498,
+      "learning_rate": 7.406633870993333e-05,
+      "loss": 0.3453,
+      "step": 8106
+    },
+    {
+      "epoch": 0.5961467754981984,
+      "grad_norm": 0.5364693999290466,
+      "learning_rate": 7.404333808614905e-05,
+      "loss": 0.3641,
+      "step": 8107
+    },
+    {
+      "epoch": 0.5962203103169351,
+      "grad_norm": 0.8425420522689819,
+      "learning_rate": 7.402033893465401e-05,
+      "loss": 0.361,
+      "step": 8108
+    },
+    {
+      "epoch": 0.5962938451356717,
+      "grad_norm": 0.900999128818512,
+      "learning_rate": 7.399734125675274e-05,
+      "loss": 0.3693,
+      "step": 8109
+    },
+    {
+      "epoch": 0.5963673799544084,
+      "grad_norm": 0.7106398344039917,
+      "learning_rate": 7.397434505374977e-05,
+      "loss": 0.3557,
+      "step": 8110
+    },
+    {
+      "epoch": 0.5964409147731451,
+      "grad_norm": 0.8519819974899292,
+      "learning_rate": 7.395135032694937e-05,
+      "loss": 0.3548,
+      "step": 8111
+    },
+    {
+      "epoch": 0.5965144495918817,
+      "grad_norm": 0.7711750864982605,
+      "learning_rate": 7.392835707765587e-05,
+      "loss": 0.3462,
+      "step": 8112
+    },
+    {
+      "epoch": 0.5965879844106184,
+      "grad_norm": 0.7870955467224121,
+      "learning_rate": 7.390536530717349e-05,
+      "loss": 0.3384,
+      "step": 8113
+    },
+    {
+      "epoch": 0.5966615192293551,
+      "grad_norm": 0.5435266494750977,
+      "learning_rate": 7.38823750168063e-05,
+      "loss": 0.3393,
+      "step": 8114
+    },
+    {
+      "epoch": 0.5967350540480918,
+      "grad_norm": 0.6179154515266418,
+      "learning_rate": 7.385938620785844e-05,
+      "loss": 0.35,
+      "step": 8115
+    },
+    {
+      "epoch": 0.5968085888668284,
+      "grad_norm": 0.5087405443191528,
+      "learning_rate": 7.383639888163371e-05,
+      "loss": 0.3478,
+      "step": 8116
+    },
+    {
+      "epoch": 0.5968821236855651,
+      "grad_norm": 0.517057478427887,
+      "learning_rate": 7.381341303943607e-05,
+      "loss": 0.3629,
+      "step": 8117
+    },
+    {
+      "epoch": 0.5969556585043018,
+      "grad_norm": 0.9029625058174133,
+      "learning_rate": 7.379042868256928e-05,
+      "loss": 0.3447,
+      "step": 8118
+    },
+    {
+      "epoch": 0.5970291933230385,
+      "grad_norm": 0.46024876832962036,
+      "learning_rate": 7.376744581233701e-05,
+      "loss": 0.3428,
+      "step": 8119
+    },
+    {
+      "epoch": 0.5971027281417751,
+      "grad_norm": 0.6138572692871094,
+      "learning_rate": 7.374446443004293e-05,
+      "loss": 0.333,
+      "step": 8120
+    },
+    {
+      "epoch": 0.5971762629605118,
+      "grad_norm": 0.47742000222206116,
+      "learning_rate": 7.372148453699052e-05,
+      "loss": 0.3184,
+      "step": 8121
+    },
+    {
+      "epoch": 0.5972497977792485,
+      "grad_norm": 0.768800675868988,
+      "learning_rate": 7.369850613448323e-05,
+      "loss": 0.3337,
+      "step": 8122
+    },
+    {
+      "epoch": 0.5973233325979851,
+      "grad_norm": 0.925486147403717,
+      "learning_rate": 7.367552922382443e-05,
+      "loss": 0.3804,
+      "step": 8123
+    },
+    {
+      "epoch": 0.5973968674167218,
+      "grad_norm": 0.49657413363456726,
+      "learning_rate": 7.365255380631739e-05,
+      "loss": 0.3761,
+      "step": 8124
+    },
+    {
+      "epoch": 0.5974704022354584,
+      "grad_norm": 0.7977949380874634,
+      "learning_rate": 7.362957988326533e-05,
+      "loss": 0.3656,
+      "step": 8125
+    },
+    {
+      "epoch": 0.5975439370541952,
+      "grad_norm": 0.6967710852622986,
+      "learning_rate": 7.360660745597132e-05,
+      "loss": 0.3728,
+      "step": 8126
+    },
+    {
+      "epoch": 0.5976174718729318,
+      "grad_norm": 0.4779576361179352,
+      "learning_rate": 7.358363652573843e-05,
+      "loss": 0.3539,
+      "step": 8127
+    },
+    {
+      "epoch": 0.5976910066916685,
+      "grad_norm": 0.46770334243774414,
+      "learning_rate": 7.35606670938695e-05,
+      "loss": 0.3287,
+      "step": 8128
+    },
+    {
+      "epoch": 0.5977645415104051,
+      "grad_norm": 0.42773905396461487,
+      "learning_rate": 7.353769916166744e-05,
+      "loss": 0.372,
+      "step": 8129
+    },
+    {
+      "epoch": 0.5978380763291419,
+      "grad_norm": 0.4261116087436676,
+      "learning_rate": 7.3514732730435e-05,
+      "loss": 0.3525,
+      "step": 8130
+    },
+    {
+      "epoch": 0.5979116111478785,
+      "grad_norm": 0.545883059501648,
+      "learning_rate": 7.349176780147491e-05,
+      "loss": 0.367,
+      "step": 8131
+    },
+    {
+      "epoch": 0.5979851459666152,
+      "grad_norm": 0.49030861258506775,
+      "learning_rate": 7.346880437608968e-05,
+      "loss": 0.3666,
+      "step": 8132
+    },
+    {
+      "epoch": 0.5980586807853518,
+      "grad_norm": 0.4906505048274994,
+      "learning_rate": 7.34458424555819e-05,
+      "loss": 0.3356,
+      "step": 8133
+    },
+    {
+      "epoch": 0.5981322156040886,
+      "grad_norm": 0.4739971160888672,
+      "learning_rate": 7.342288204125395e-05,
+      "loss": 0.3475,
+      "step": 8134
+    },
+    {
+      "epoch": 0.5982057504228252,
+      "grad_norm": 0.5181059837341309,
+      "learning_rate": 7.339992313440816e-05,
+      "loss": 0.3484,
+      "step": 8135
+    },
+    {
+      "epoch": 0.5982792852415619,
+      "grad_norm": 0.5256387591362,
+      "learning_rate": 7.337696573634683e-05,
+      "loss": 0.3575,
+      "step": 8136
+    },
+    {
+      "epoch": 0.5983528200602986,
+      "grad_norm": 0.7012495398521423,
+      "learning_rate": 7.335400984837207e-05,
+      "loss": 0.3372,
+      "step": 8137
+    },
+    {
+      "epoch": 0.5984263548790352,
+      "grad_norm": 0.5153701901435852,
+      "learning_rate": 7.333105547178605e-05,
+      "loss": 0.3714,
+      "step": 8138
+    },
+    {
+      "epoch": 0.5984998896977719,
+      "grad_norm": 0.5329612493515015,
+      "learning_rate": 7.330810260789065e-05,
+      "loss": 0.3897,
+      "step": 8139
+    },
+    {
+      "epoch": 0.5985734245165085,
+      "grad_norm": 0.6161844730377197,
+      "learning_rate": 7.328515125798787e-05,
+      "loss": 0.3618,
+      "step": 8140
+    },
+    {
+      "epoch": 0.5986469593352453,
+      "grad_norm": 0.6370679140090942,
+      "learning_rate": 7.326220142337949e-05,
+      "loss": 0.3505,
+      "step": 8141
+    },
+    {
+      "epoch": 0.5987204941539819,
+      "grad_norm": 0.6432853937149048,
+      "learning_rate": 7.323925310536724e-05,
+      "loss": 0.36,
+      "step": 8142
+    },
+    {
+      "epoch": 0.5987940289727186,
+      "grad_norm": 0.5881980061531067,
+      "learning_rate": 7.321630630525282e-05,
+      "loss": 0.3305,
+      "step": 8143
+    },
+    {
+      "epoch": 0.5988675637914552,
+      "grad_norm": 0.6485658288002014,
+      "learning_rate": 7.319336102433779e-05,
+      "loss": 0.3908,
+      "step": 8144
+    },
+    {
+      "epoch": 0.598941098610192,
+      "grad_norm": 0.7363612055778503,
+      "learning_rate": 7.317041726392357e-05,
+      "loss": 0.3593,
+      "step": 8145
+    },
+    {
+      "epoch": 0.5990146334289286,
+      "grad_norm": 0.513533890247345,
+      "learning_rate": 7.314747502531163e-05,
+      "loss": 0.3512,
+      "step": 8146
+    },
+    {
+      "epoch": 0.5990881682476653,
+      "grad_norm": 0.8218700885772705,
+      "learning_rate": 7.312453430980322e-05,
+      "loss": 0.3825,
+      "step": 8147
+    },
+    {
+      "epoch": 0.5991617030664019,
+      "grad_norm": 1.1454977989196777,
+      "learning_rate": 7.31015951186996e-05,
+      "loss": 0.3536,
+      "step": 8148
+    },
+    {
+      "epoch": 0.5992352378851387,
+      "grad_norm": 0.6251192092895508,
+      "learning_rate": 7.307865745330192e-05,
+      "loss": 0.3634,
+      "step": 8149
+    },
+    {
+      "epoch": 0.5993087727038753,
+      "grad_norm": 0.5545515418052673,
+      "learning_rate": 7.30557213149112e-05,
+      "loss": 0.3553,
+      "step": 8150
+    },
+    {
+      "epoch": 0.5993823075226119,
+      "grad_norm": 0.539008617401123,
+      "learning_rate": 7.303278670482839e-05,
+      "loss": 0.3578,
+      "step": 8151
+    },
+    {
+      "epoch": 0.5994558423413486,
+      "grad_norm": 0.6885058879852295,
+      "learning_rate": 7.300985362435433e-05,
+      "loss": 0.3294,
+      "step": 8152
+    },
+    {
+      "epoch": 0.5995293771600853,
+      "grad_norm": 0.9257411956787109,
+      "learning_rate": 7.29869220747899e-05,
+      "loss": 0.4107,
+      "step": 8153
+    },
+    {
+      "epoch": 0.599602911978822,
+      "grad_norm": 0.510121762752533,
+      "learning_rate": 7.296399205743577e-05,
+      "loss": 0.346,
+      "step": 8154
+    },
+    {
+      "epoch": 0.5996764467975586,
+      "grad_norm": 0.8854896426200867,
+      "learning_rate": 7.29410635735925e-05,
+      "loss": 0.3744,
+      "step": 8155
+    },
+    {
+      "epoch": 0.5997499816162953,
+      "grad_norm": 0.5998499989509583,
+      "learning_rate": 7.291813662456069e-05,
+      "loss": 0.3329,
+      "step": 8156
+    },
+    {
+      "epoch": 0.599823516435032,
+      "grad_norm": 0.8373947739601135,
+      "learning_rate": 7.289521121164075e-05,
+      "loss": 0.3635,
+      "step": 8157
+    },
+    {
+      "epoch": 0.5998970512537687,
+      "grad_norm": 1.0486185550689697,
+      "learning_rate": 7.2872287336133e-05,
+      "loss": 0.3567,
+      "step": 8158
+    },
+    {
+      "epoch": 0.5999705860725053,
+      "grad_norm": 0.6137644648551941,
+      "learning_rate": 7.284936499933778e-05,
+      "loss": 0.3799,
+      "step": 8159
+    },
+    {
+      "epoch": 0.600044120891242,
+      "grad_norm": 0.6576284170150757,
+      "learning_rate": 7.282644420255522e-05,
+      "loss": 0.3555,
+      "step": 8160
+    },
+    {
+      "epoch": 0.6001176557099787,
+      "grad_norm": 0.6704086065292358,
+      "learning_rate": 7.280352494708542e-05,
+      "loss": 0.3533,
+      "step": 8161
+    },
+    {
+      "epoch": 0.6001911905287154,
+      "grad_norm": 0.7451910972595215,
+      "learning_rate": 7.278060723422841e-05,
+      "loss": 0.355,
+      "step": 8162
+    },
+    {
+      "epoch": 0.600264725347452,
+      "grad_norm": 0.4076879918575287,
+      "learning_rate": 7.275769106528407e-05,
+      "loss": 0.3349,
+      "step": 8163
+    },
+    {
+      "epoch": 0.6003382601661887,
+      "grad_norm": 0.8162896037101746,
+      "learning_rate": 7.273477644155225e-05,
+      "loss": 0.4143,
+      "step": 8164
+    },
+    {
+      "epoch": 0.6004117949849254,
+      "grad_norm": 0.5747522711753845,
+      "learning_rate": 7.271186336433265e-05,
+      "loss": 0.3533,
+      "step": 8165
+    },
+    {
+      "epoch": 0.600485329803662,
+      "grad_norm": 0.6850835084915161,
+      "learning_rate": 7.268895183492498e-05,
+      "loss": 0.3561,
+      "step": 8166
+    },
+    {
+      "epoch": 0.6005588646223987,
+      "grad_norm": 0.6450134515762329,
+      "learning_rate": 7.26660418546288e-05,
+      "loss": 0.3488,
+      "step": 8167
+    },
+    {
+      "epoch": 0.6006323994411353,
+      "grad_norm": 0.7432376146316528,
+      "learning_rate": 7.264313342474353e-05,
+      "loss": 0.3516,
+      "step": 8168
+    },
+    {
+      "epoch": 0.6007059342598721,
+      "grad_norm": 0.4381636381149292,
+      "learning_rate": 7.262022654656863e-05,
+      "loss": 0.3311,
+      "step": 8169
+    },
+    {
+      "epoch": 0.6007794690786087,
+      "grad_norm": 0.6479642987251282,
+      "learning_rate": 7.259732122140337e-05,
+      "loss": 0.3443,
+      "step": 8170
+    },
+    {
+      "epoch": 0.6008530038973454,
+      "grad_norm": 0.4364616870880127,
+      "learning_rate": 7.257441745054699e-05,
+      "loss": 0.3651,
+      "step": 8171
+    },
+    {
+      "epoch": 0.600926538716082,
+      "grad_norm": 0.7454835772514343,
+      "learning_rate": 7.255151523529858e-05,
+      "loss": 0.3555,
+      "step": 8172
+    },
+    {
+      "epoch": 0.6010000735348188,
+      "grad_norm": 0.503334641456604,
+      "learning_rate": 7.25286145769572e-05,
+      "loss": 0.3343,
+      "step": 8173
+    },
+    {
+      "epoch": 0.6010736083535554,
+      "grad_norm": 0.6129637360572815,
+      "learning_rate": 7.250571547682182e-05,
+      "loss": 0.3476,
+      "step": 8174
+    },
+    {
+      "epoch": 0.6011471431722921,
+      "grad_norm": 0.7604959011077881,
+      "learning_rate": 7.24828179361912e-05,
+      "loss": 0.3532,
+      "step": 8175
+    },
+    {
+      "epoch": 0.6012206779910287,
+      "grad_norm": 0.5289852023124695,
+      "learning_rate": 7.245992195636424e-05,
+      "loss": 0.3833,
+      "step": 8176
+    },
+    {
+      "epoch": 0.6012942128097655,
+      "grad_norm": 0.9110402464866638,
+      "learning_rate": 7.243702753863957e-05,
+      "loss": 0.3534,
+      "step": 8177
+    },
+    {
+      "epoch": 0.6013677476285021,
+      "grad_norm": 1.5147359371185303,
+      "learning_rate": 7.241413468431574e-05,
+      "loss": 0.3699,
+      "step": 8178
+    },
+    {
+      "epoch": 0.6014412824472388,
+      "grad_norm": 0.7181137204170227,
+      "learning_rate": 7.239124339469136e-05,
+      "loss": 0.3308,
+      "step": 8179
+    },
+    {
+      "epoch": 0.6015148172659754,
+      "grad_norm": 2.1370275020599365,
+      "learning_rate": 7.236835367106477e-05,
+      "loss": 0.3683,
+      "step": 8180
+    },
+    {
+      "epoch": 0.6015883520847121,
+      "grad_norm": 1.2899870872497559,
+      "learning_rate": 7.23454655147343e-05,
+      "loss": 0.3534,
+      "step": 8181
+    },
+    {
+      "epoch": 0.6016618869034488,
+      "grad_norm": 1.774027705192566,
+      "learning_rate": 7.232257892699824e-05,
+      "loss": 0.3502,
+      "step": 8182
+    },
+    {
+      "epoch": 0.6017354217221854,
+      "grad_norm": 1.913740634918213,
+      "learning_rate": 7.229969390915467e-05,
+      "loss": 0.3662,
+      "step": 8183
+    },
+    {
+      "epoch": 0.6018089565409221,
+      "grad_norm": 1.0699596405029297,
+      "learning_rate": 7.227681046250174e-05,
+      "loss": 0.3434,
+      "step": 8184
+    },
+    {
+      "epoch": 0.6018824913596588,
+      "grad_norm": 1.165242075920105,
+      "learning_rate": 7.22539285883374e-05,
+      "loss": 0.3541,
+      "step": 8185
+    },
+    {
+      "epoch": 0.6019560261783955,
+      "grad_norm": 1.2242375612258911,
+      "learning_rate": 7.223104828795948e-05,
+      "loss": 0.3326,
+      "step": 8186
+    },
+    {
+      "epoch": 0.6020295609971321,
+      "grad_norm": 0.9133359789848328,
+      "learning_rate": 7.220816956266581e-05,
+      "loss": 0.3447,
+      "step": 8187
+    },
+    {
+      "epoch": 0.6021030958158688,
+      "grad_norm": 0.9726787805557251,
+      "learning_rate": 7.218529241375407e-05,
+      "loss": 0.3284,
+      "step": 8188
+    },
+    {
+      "epoch": 0.6021766306346055,
+      "grad_norm": 0.7423802614212036,
+      "learning_rate": 7.216241684252192e-05,
+      "loss": 0.3383,
+      "step": 8189
+    },
+    {
+      "epoch": 0.6022501654533422,
+      "grad_norm": 0.9540519118309021,
+      "learning_rate": 7.213954285026686e-05,
+      "loss": 0.3426,
+      "step": 8190
+    },
+    {
+      "epoch": 0.6023237002720788,
+      "grad_norm": 0.5517004728317261,
+      "learning_rate": 7.211667043828632e-05,
+      "loss": 0.3356,
+      "step": 8191
+    },
+    {
+      "epoch": 0.6023972350908156,
+      "grad_norm": 1.0914781093597412,
+      "learning_rate": 7.209379960787768e-05,
+      "loss": 0.3304,
+      "step": 8192
+    },
+    {
+      "epoch": 0.6024707699095522,
+      "grad_norm": 0.5079007148742676,
+      "learning_rate": 7.207093036033815e-05,
+      "loss": 0.3416,
+      "step": 8193
+    },
+    {
+      "epoch": 0.6025443047282888,
+      "grad_norm": 0.993626594543457,
+      "learning_rate": 7.204806269696494e-05,
+      "loss": 0.3613,
+      "step": 8194
+    },
+    {
+      "epoch": 0.6026178395470255,
+      "grad_norm": 1.085702657699585,
+      "learning_rate": 7.202519661905512e-05,
+      "loss": 0.3508,
+      "step": 8195
+    },
+    {
+      "epoch": 0.6026913743657621,
+      "grad_norm": 0.6778366565704346,
+      "learning_rate": 7.200233212790566e-05,
+      "loss": 0.3375,
+      "step": 8196
+    },
+    {
+      "epoch": 0.6027649091844989,
+      "grad_norm": 1.4388376474380493,
+      "learning_rate": 7.197946922481351e-05,
+      "loss": 0.3672,
+      "step": 8197
+    },
+    {
+      "epoch": 0.6028384440032355,
+      "grad_norm": 0.7044104933738708,
+      "learning_rate": 7.195660791107539e-05,
+      "loss": 0.3552,
+      "step": 8198
+    },
+    {
+      "epoch": 0.6029119788219722,
+      "grad_norm": 1.5372604131698608,
+      "learning_rate": 7.193374818798808e-05,
+      "loss": 0.3319,
+      "step": 8199
+    },
+    {
+      "epoch": 0.6029855136407088,
+      "grad_norm": 0.873512327671051,
+      "learning_rate": 7.191089005684819e-05,
+      "loss": 0.3418,
+      "step": 8200
+    },
+    {
+      "epoch": 0.6030590484594456,
+      "grad_norm": 1.5681811571121216,
+      "learning_rate": 7.188803351895225e-05,
+      "loss": 0.3521,
+      "step": 8201
+    },
+    {
+      "epoch": 0.6031325832781822,
+      "grad_norm": 1.437869668006897,
+      "learning_rate": 7.186517857559672e-05,
+      "loss": 0.3531,
+      "step": 8202
+    },
+    {
+      "epoch": 0.6032061180969189,
+      "grad_norm": 1.2787344455718994,
+      "learning_rate": 7.184232522807797e-05,
+      "loss": 0.3406,
+      "step": 8203
+    },
+    {
+      "epoch": 0.6032796529156556,
+      "grad_norm": 1.391981601715088,
+      "learning_rate": 7.181947347769222e-05,
+      "loss": 0.3497,
+      "step": 8204
+    },
+    {
+      "epoch": 0.6033531877343923,
+      "grad_norm": 1.3080456256866455,
+      "learning_rate": 7.179662332573569e-05,
+      "loss": 0.3582,
+      "step": 8205
+    },
+    {
+      "epoch": 0.6034267225531289,
+      "grad_norm": 0.8808373808860779,
+      "learning_rate": 7.177377477350445e-05,
+      "loss": 0.3503,
+      "step": 8206
+    },
+    {
+      "epoch": 0.6035002573718656,
+      "grad_norm": 1.2983462810516357,
+      "learning_rate": 7.17509278222945e-05,
+      "loss": 0.3569,
+      "step": 8207
+    },
+    {
+      "epoch": 0.6035737921906023,
+      "grad_norm": 0.7284552454948425,
+      "learning_rate": 7.172808247340175e-05,
+      "loss": 0.3505,
+      "step": 8208
+    },
+    {
+      "epoch": 0.6036473270093389,
+      "grad_norm": 1.8926198482513428,
+      "learning_rate": 7.170523872812198e-05,
+      "loss": 0.3536,
+      "step": 8209
+    },
+    {
+      "epoch": 0.6037208618280756,
+      "grad_norm": 1.0462331771850586,
+      "learning_rate": 7.168239658775093e-05,
+      "loss": 0.3388,
+      "step": 8210
+    },
+    {
+      "epoch": 0.6037943966468122,
+      "grad_norm": 2.1389596462249756,
+      "learning_rate": 7.165955605358421e-05,
+      "loss": 0.379,
+      "step": 8211
+    },
+    {
+      "epoch": 0.603867931465549,
+      "grad_norm": 2.190845012664795,
+      "learning_rate": 7.163671712691741e-05,
+      "loss": 0.3505,
+      "step": 8212
+    },
+    {
+      "epoch": 0.6039414662842856,
+      "grad_norm": 0.5252201557159424,
+      "learning_rate": 7.161387980904594e-05,
+      "loss": 0.3312,
+      "step": 8213
+    },
+    {
+      "epoch": 0.6040150011030223,
+      "grad_norm": 1.9001613855361938,
+      "learning_rate": 7.159104410126513e-05,
+      "loss": 0.3479,
+      "step": 8214
+    },
+    {
+      "epoch": 0.6040885359217589,
+      "grad_norm": 1.1542948484420776,
+      "learning_rate": 7.156821000487032e-05,
+      "loss": 0.3454,
+      "step": 8215
+    },
+    {
+      "epoch": 0.6041620707404957,
+      "grad_norm": 1.8883124589920044,
+      "learning_rate": 7.154537752115662e-05,
+      "loss": 0.3469,
+      "step": 8216
+    },
+    {
+      "epoch": 0.6042356055592323,
+      "grad_norm": 2.54520845413208,
+      "learning_rate": 7.152254665141916e-05,
+      "loss": 0.3439,
+      "step": 8217
+    },
+    {
+      "epoch": 0.604309140377969,
+      "grad_norm": 0.8817645907402039,
+      "learning_rate": 7.149971739695289e-05,
+      "loss": 0.3464,
+      "step": 8218
+    },
+    {
+      "epoch": 0.6043826751967056,
+      "grad_norm": 1.4419281482696533,
+      "learning_rate": 7.147688975905274e-05,
+      "loss": 0.3522,
+      "step": 8219
+    },
+    {
+      "epoch": 0.6044562100154424,
+      "grad_norm": 0.829383909702301,
+      "learning_rate": 7.145406373901354e-05,
+      "loss": 0.347,
+      "step": 8220
+    },
+    {
+      "epoch": 0.604529744834179,
+      "grad_norm": 1.2477589845657349,
+      "learning_rate": 7.143123933812991e-05,
+      "loss": 0.3195,
+      "step": 8221
+    },
+    {
+      "epoch": 0.6046032796529156,
+      "grad_norm": 1.1932352781295776,
+      "learning_rate": 7.140841655769656e-05,
+      "loss": 0.3469,
+      "step": 8222
+    },
+    {
+      "epoch": 0.6046768144716523,
+      "grad_norm": 0.8046215176582336,
+      "learning_rate": 7.138559539900799e-05,
+      "loss": 0.3484,
+      "step": 8223
+    },
+    {
+      "epoch": 0.604750349290389,
+      "grad_norm": 1.2741742134094238,
+      "learning_rate": 7.136277586335863e-05,
+      "loss": 0.3289,
+      "step": 8224
+    },
+    {
+      "epoch": 0.6048238841091257,
+      "grad_norm": 0.708217442035675,
+      "learning_rate": 7.133995795204287e-05,
+      "loss": 0.3519,
+      "step": 8225
+    },
+    {
+      "epoch": 0.6048974189278623,
+      "grad_norm": 1.2425906658172607,
+      "learning_rate": 7.131714166635494e-05,
+      "loss": 0.3438,
+      "step": 8226
+    },
+    {
+      "epoch": 0.604970953746599,
+      "grad_norm": 0.6490116715431213,
+      "learning_rate": 7.129432700758897e-05,
+      "loss": 0.2939,
+      "step": 8227
+    },
+    {
+      "epoch": 0.6050444885653357,
+      "grad_norm": 0.9496816396713257,
+      "learning_rate": 7.12715139770391e-05,
+      "loss": 0.3413,
+      "step": 8228
+    },
+    {
+      "epoch": 0.6051180233840724,
+      "grad_norm": 0.7752984166145325,
+      "learning_rate": 7.124870257599926e-05,
+      "loss": 0.3252,
+      "step": 8229
+    },
+    {
+      "epoch": 0.605191558202809,
+      "grad_norm": 0.6910108923912048,
+      "learning_rate": 7.122589280576337e-05,
+      "loss": 0.344,
+      "step": 8230
+    },
+    {
+      "epoch": 0.6052650930215457,
+      "grad_norm": 0.8270978331565857,
+      "learning_rate": 7.12030846676252e-05,
+      "loss": 0.3437,
+      "step": 8231
+    },
+    {
+      "epoch": 0.6053386278402824,
+      "grad_norm": 0.6913051605224609,
+      "learning_rate": 7.118027816287849e-05,
+      "loss": 0.3454,
+      "step": 8232
+    },
+    {
+      "epoch": 0.6054121626590191,
+      "grad_norm": 0.5846638083457947,
+      "learning_rate": 7.115747329281679e-05,
+      "loss": 0.3089,
+      "step": 8233
+    },
+    {
+      "epoch": 0.6054856974777557,
+      "grad_norm": 0.5339980721473694,
+      "learning_rate": 7.113467005873362e-05,
+      "loss": 0.3317,
+      "step": 8234
+    },
+    {
+      "epoch": 0.6055592322964924,
+      "grad_norm": 0.688757061958313,
+      "learning_rate": 7.111186846192244e-05,
+      "loss": 0.3208,
+      "step": 8235
+    },
+    {
+      "epoch": 0.6056327671152291,
+      "grad_norm": 0.520949125289917,
+      "learning_rate": 7.108906850367658e-05,
+      "loss": 0.3414,
+      "step": 8236
+    },
+    {
+      "epoch": 0.6057063019339657,
+      "grad_norm": 0.8777709007263184,
+      "learning_rate": 7.106627018528925e-05,
+      "loss": 0.3356,
+      "step": 8237
+    },
+    {
+      "epoch": 0.6057798367527024,
+      "grad_norm": 0.5613953471183777,
+      "learning_rate": 7.104347350805363e-05,
+      "loss": 0.3292,
+      "step": 8238
+    },
+    {
+      "epoch": 0.605853371571439,
+      "grad_norm": 0.8390777707099915,
+      "learning_rate": 7.102067847326274e-05,
+      "loss": 0.3305,
+      "step": 8239
+    },
+    {
+      "epoch": 0.6059269063901758,
+      "grad_norm": 0.5666837096214294,
+      "learning_rate": 7.099788508220952e-05,
+      "loss": 0.3458,
+      "step": 8240
+    },
+    {
+      "epoch": 0.6060004412089124,
+      "grad_norm": 0.6122527718544006,
+      "learning_rate": 7.097509333618691e-05,
+      "loss": 0.3335,
+      "step": 8241
+    },
+    {
+      "epoch": 0.6060739760276491,
+      "grad_norm": 0.6474093198776245,
+      "learning_rate": 7.09523032364876e-05,
+      "loss": 0.3475,
+      "step": 8242
+    },
+    {
+      "epoch": 0.6061475108463857,
+      "grad_norm": 0.6022659540176392,
+      "learning_rate": 7.092951478440439e-05,
+      "loss": 0.3449,
+      "step": 8243
+    },
+    {
+      "epoch": 0.6062210456651225,
+      "grad_norm": 0.4482566714286804,
+      "learning_rate": 7.09067279812297e-05,
+      "loss": 0.3249,
+      "step": 8244
+    },
+    {
+      "epoch": 0.6062945804838591,
+      "grad_norm": 0.6123313307762146,
+      "learning_rate": 7.088394282825612e-05,
+      "loss": 0.3506,
+      "step": 8245
+    },
+    {
+      "epoch": 0.6063681153025958,
+      "grad_norm": 0.5080642700195312,
+      "learning_rate": 7.086115932677604e-05,
+      "loss": 0.3341,
+      "step": 8246
+    },
+    {
+      "epoch": 0.6064416501213324,
+      "grad_norm": 0.6829054951667786,
+      "learning_rate": 7.083837747808173e-05,
+      "loss": 0.3484,
+      "step": 8247
+    },
+    {
+      "epoch": 0.6065151849400692,
+      "grad_norm": 0.4268295466899872,
+      "learning_rate": 7.081559728346545e-05,
+      "loss": 0.3208,
+      "step": 8248
+    },
+    {
+      "epoch": 0.6065887197588058,
+      "grad_norm": 0.5009310245513916,
+      "learning_rate": 7.079281874421928e-05,
+      "loss": 0.3639,
+      "step": 8249
+    },
+    {
+      "epoch": 0.6066622545775424,
+      "grad_norm": 0.5002111792564392,
+      "learning_rate": 7.077004186163525e-05,
+      "loss": 0.3741,
+      "step": 8250
+    },
+    {
+      "epoch": 0.6067357893962791,
+      "grad_norm": 0.5719836950302124,
+      "learning_rate": 7.07472666370053e-05,
+      "loss": 0.3522,
+      "step": 8251
+    },
+    {
+      "epoch": 0.6068093242150158,
+      "grad_norm": 0.6134761571884155,
+      "learning_rate": 7.072449307162125e-05,
+      "loss": 0.3313,
+      "step": 8252
+    },
+    {
+      "epoch": 0.6068828590337525,
+      "grad_norm": 0.49271267652511597,
+      "learning_rate": 7.070172116677484e-05,
+      "loss": 0.3279,
+      "step": 8253
+    },
+    {
+      "epoch": 0.6069563938524891,
+      "grad_norm": 0.6498926877975464,
+      "learning_rate": 7.067895092375776e-05,
+      "loss": 0.3399,
+      "step": 8254
+    },
+    {
+      "epoch": 0.6070299286712258,
+      "grad_norm": 0.46176454424858093,
+      "learning_rate": 7.065618234386154e-05,
+      "loss": 0.3673,
+      "step": 8255
+    },
+    {
+      "epoch": 0.6071034634899625,
+      "grad_norm": 0.711781919002533,
+      "learning_rate": 7.06334154283776e-05,
+      "loss": 0.3251,
+      "step": 8256
+    },
+    {
+      "epoch": 0.6071769983086992,
+      "grad_norm": 0.7409199476242065,
+      "learning_rate": 7.06106501785973e-05,
+      "loss": 0.3483,
+      "step": 8257
+    },
+    {
+      "epoch": 0.6072505331274358,
+      "grad_norm": 0.5020418763160706,
+      "learning_rate": 7.058788659581196e-05,
+      "loss": 0.3252,
+      "step": 8258
+    },
+    {
+      "epoch": 0.6073240679461726,
+      "grad_norm": 0.5455499887466431,
+      "learning_rate": 7.056512468131276e-05,
+      "loss": 0.3274,
+      "step": 8259
+    },
+    {
+      "epoch": 0.6073976027649092,
+      "grad_norm": 0.768730640411377,
+      "learning_rate": 7.05423644363907e-05,
+      "loss": 0.361,
+      "step": 8260
+    },
+    {
+      "epoch": 0.6074711375836459,
+      "grad_norm": 0.45491060614585876,
+      "learning_rate": 7.051960586233683e-05,
+      "loss": 0.329,
+      "step": 8261
+    },
+    {
+      "epoch": 0.6075446724023825,
+      "grad_norm": 0.6517027616500854,
+      "learning_rate": 7.049684896044206e-05,
+      "loss": 0.3219,
+      "step": 8262
+    },
+    {
+      "epoch": 0.6076182072211193,
+      "grad_norm": 0.5464053153991699,
+      "learning_rate": 7.047409373199712e-05,
+      "loss": 0.354,
+      "step": 8263
+    },
+    {
+      "epoch": 0.6076917420398559,
+      "grad_norm": 0.5162655115127563,
+      "learning_rate": 7.045134017829275e-05,
+      "loss": 0.3241,
+      "step": 8264
+    },
+    {
+      "epoch": 0.6077652768585925,
+      "grad_norm": 0.6441061496734619,
+      "learning_rate": 7.042858830061953e-05,
+      "loss": 0.3277,
+      "step": 8265
+    },
+    {
+      "epoch": 0.6078388116773292,
+      "grad_norm": 0.4493354558944702,
+      "learning_rate": 7.040583810026803e-05,
+      "loss": 0.3447,
+      "step": 8266
+    },
+    {
+      "epoch": 0.6079123464960658,
+      "grad_norm": 1.122719168663025,
+      "learning_rate": 7.038308957852864e-05,
+      "loss": 0.3091,
+      "step": 8267
+    },
+    {
+      "epoch": 0.6079858813148026,
+      "grad_norm": 0.8152994513511658,
+      "learning_rate": 7.036034273669163e-05,
+      "loss": 0.3146,
+      "step": 8268
+    },
+    {
+      "epoch": 0.6080594161335392,
+      "grad_norm": 0.5089425444602966,
+      "learning_rate": 7.033759757604728e-05,
+      "loss": 0.3126,
+      "step": 8269
+    },
+    {
+      "epoch": 0.6081329509522759,
+      "grad_norm": 0.7945634126663208,
+      "learning_rate": 7.031485409788567e-05,
+      "loss": 0.3319,
+      "step": 8270
+    },
+    {
+      "epoch": 0.6082064857710126,
+      "grad_norm": 0.4583149552345276,
+      "learning_rate": 7.029211230349688e-05,
+      "loss": 0.3173,
+      "step": 8271
+    },
+    {
+      "epoch": 0.6082800205897493,
+      "grad_norm": 0.7937613129615784,
+      "learning_rate": 7.026937219417083e-05,
+      "loss": 0.3016,
+      "step": 8272
+    },
+    {
+      "epoch": 0.6083535554084859,
+      "grad_norm": 0.5331178903579712,
+      "learning_rate": 7.024663377119735e-05,
+      "loss": 0.3303,
+      "step": 8273
+    },
+    {
+      "epoch": 0.6084270902272226,
+      "grad_norm": 0.9907264709472656,
+      "learning_rate": 7.022389703586622e-05,
+      "loss": 0.3272,
+      "step": 8274
+    },
+    {
+      "epoch": 0.6085006250459593,
+      "grad_norm": 0.780127763748169,
+      "learning_rate": 7.020116198946706e-05,
+      "loss": 0.3322,
+      "step": 8275
+    },
+    {
+      "epoch": 0.608574159864696,
+      "grad_norm": 0.6767337918281555,
+      "learning_rate": 7.017842863328944e-05,
+      "loss": 0.2995,
+      "step": 8276
+    },
+    {
+      "epoch": 0.6086476946834326,
+      "grad_norm": 0.6752906441688538,
+      "learning_rate": 7.015569696862285e-05,
+      "loss": 0.3379,
+      "step": 8277
+    },
+    {
+      "epoch": 0.6087212295021692,
+      "grad_norm": 0.6567636728286743,
+      "learning_rate": 7.013296699675658e-05,
+      "loss": 0.3295,
+      "step": 8278
+    },
+    {
+      "epoch": 0.608794764320906,
+      "grad_norm": 0.901816189289093,
+      "learning_rate": 7.011023871898001e-05,
+      "loss": 0.3277,
+      "step": 8279
+    },
+    {
+      "epoch": 0.6088682991396426,
+      "grad_norm": 0.4943982660770416,
+      "learning_rate": 7.00875121365822e-05,
+      "loss": 0.315,
+      "step": 8280
+    },
+    {
+      "epoch": 0.6089418339583793,
+      "grad_norm": 0.9063783288002014,
+      "learning_rate": 7.006478725085226e-05,
+      "loss": 0.3335,
+      "step": 8281
+    },
+    {
+      "epoch": 0.6090153687771159,
+      "grad_norm": 0.5003710389137268,
+      "learning_rate": 7.00420640630792e-05,
+      "loss": 0.3207,
+      "step": 8282
+    },
+    {
+      "epoch": 0.6090889035958527,
+      "grad_norm": 1.0779533386230469,
+      "learning_rate": 7.001934257455185e-05,
+      "loss": 0.3511,
+      "step": 8283
+    },
+    {
+      "epoch": 0.6091624384145893,
+      "grad_norm": 0.5726550221443176,
+      "learning_rate": 6.999662278655906e-05,
+      "loss": 0.3526,
+      "step": 8284
+    },
+    {
+      "epoch": 0.609235973233326,
+      "grad_norm": 0.8638819456100464,
+      "learning_rate": 6.997390470038949e-05,
+      "loss": 0.3426,
+      "step": 8285
+    },
+    {
+      "epoch": 0.6093095080520626,
+      "grad_norm": 0.9095537066459656,
+      "learning_rate": 6.99511883173317e-05,
+      "loss": 0.3553,
+      "step": 8286
+    },
+    {
+      "epoch": 0.6093830428707994,
+      "grad_norm": 0.5195640921592712,
+      "learning_rate": 6.992847363867425e-05,
+      "loss": 0.3428,
+      "step": 8287
+    },
+    {
+      "epoch": 0.609456577689536,
+      "grad_norm": 0.7165118455886841,
+      "learning_rate": 6.990576066570549e-05,
+      "loss": 0.4049,
+      "step": 8288
+    },
+    {
+      "epoch": 0.6095301125082727,
+      "grad_norm": 0.43285951018333435,
+      "learning_rate": 6.988304939971377e-05,
+      "loss": 0.3081,
+      "step": 8289
+    },
+    {
+      "epoch": 0.6096036473270093,
+      "grad_norm": 0.5148053169250488,
+      "learning_rate": 6.986033984198727e-05,
+      "loss": 0.3423,
+      "step": 8290
+    },
+    {
+      "epoch": 0.6096771821457461,
+      "grad_norm": 0.44212883710861206,
+      "learning_rate": 6.98376319938141e-05,
+      "loss": 0.3057,
+      "step": 8291
+    },
+    {
+      "epoch": 0.6097507169644827,
+      "grad_norm": 0.5050159692764282,
+      "learning_rate": 6.981492585648228e-05,
+      "loss": 0.3405,
+      "step": 8292
+    },
+    {
+      "epoch": 0.6098242517832193,
+      "grad_norm": 0.5297397375106812,
+      "learning_rate": 6.97922214312797e-05,
+      "loss": 0.3394,
+      "step": 8293
+    },
+    {
+      "epoch": 0.609897786601956,
+      "grad_norm": 0.5101082921028137,
+      "learning_rate": 6.97695187194942e-05,
+      "loss": 0.348,
+      "step": 8294
+    },
+    {
+      "epoch": 0.6099713214206927,
+      "grad_norm": 0.5602025985717773,
+      "learning_rate": 6.974681772241351e-05,
+      "loss": 0.3527,
+      "step": 8295
+    },
+    {
+      "epoch": 0.6100448562394294,
+      "grad_norm": 0.5388240218162537,
+      "learning_rate": 6.972411844132524e-05,
+      "loss": 0.3324,
+      "step": 8296
+    },
+    {
+      "epoch": 0.610118391058166,
+      "grad_norm": 0.4448326528072357,
+      "learning_rate": 6.970142087751694e-05,
+      "loss": 0.3435,
+      "step": 8297
+    },
+    {
+      "epoch": 0.6101919258769027,
+      "grad_norm": 0.4838097095489502,
+      "learning_rate": 6.967872503227599e-05,
+      "loss": 0.3422,
+      "step": 8298
+    },
+    {
+      "epoch": 0.6102654606956394,
+      "grad_norm": 0.5387750864028931,
+      "learning_rate": 6.965603090688978e-05,
+      "loss": 0.3441,
+      "step": 8299
+    },
+    {
+      "epoch": 0.6103389955143761,
+      "grad_norm": 0.5192068219184875,
+      "learning_rate": 6.963333850264553e-05,
+      "loss": 0.3253,
+      "step": 8300
+    },
+    {
+      "epoch": 0.6104125303331127,
+      "grad_norm": 0.47793087363243103,
+      "learning_rate": 6.961064782083034e-05,
+      "loss": 0.3305,
+      "step": 8301
+    },
+    {
+      "epoch": 0.6104860651518494,
+      "grad_norm": 0.5752599239349365,
+      "learning_rate": 6.958795886273133e-05,
+      "loss": 0.3181,
+      "step": 8302
+    },
+    {
+      "epoch": 0.6105595999705861,
+      "grad_norm": 0.8545789122581482,
+      "learning_rate": 6.956527162963534e-05,
+      "loss": 0.3827,
+      "step": 8303
+    },
+    {
+      "epoch": 0.6106331347893228,
+      "grad_norm": 0.46315550804138184,
+      "learning_rate": 6.954258612282927e-05,
+      "loss": 0.3305,
+      "step": 8304
+    },
+    {
+      "epoch": 0.6107066696080594,
+      "grad_norm": 1.0119274854660034,
+      "learning_rate": 6.951990234359987e-05,
+      "loss": 0.3392,
+      "step": 8305
+    },
+    {
+      "epoch": 0.610780204426796,
+      "grad_norm": 0.9480305314064026,
+      "learning_rate": 6.949722029323376e-05,
+      "loss": 0.3299,
+      "step": 8306
+    },
+    {
+      "epoch": 0.6108537392455328,
+      "grad_norm": 0.5041278004646301,
+      "learning_rate": 6.947453997301752e-05,
+      "loss": 0.3553,
+      "step": 8307
+    },
+    {
+      "epoch": 0.6109272740642694,
+      "grad_norm": 0.6517362594604492,
+      "learning_rate": 6.945186138423758e-05,
+      "loss": 0.3139,
+      "step": 8308
+    },
+    {
+      "epoch": 0.6110008088830061,
+      "grad_norm": 0.5437297224998474,
+      "learning_rate": 6.942918452818028e-05,
+      "loss": 0.3289,
+      "step": 8309
+    },
+    {
+      "epoch": 0.6110743437017427,
+      "grad_norm": 0.4757782220840454,
+      "learning_rate": 6.940650940613192e-05,
+      "loss": 0.3582,
+      "step": 8310
+    },
+    {
+      "epoch": 0.6111478785204795,
+      "grad_norm": 0.6809840798377991,
+      "learning_rate": 6.938383601937863e-05,
+      "loss": 0.3371,
+      "step": 8311
+    },
+    {
+      "epoch": 0.6112214133392161,
+      "grad_norm": 0.4930642247200012,
+      "learning_rate": 6.936116436920648e-05,
+      "loss": 0.3347,
+      "step": 8312
+    },
+    {
+      "epoch": 0.6112949481579528,
+      "grad_norm": 0.6868883967399597,
+      "learning_rate": 6.933849445690141e-05,
+      "loss": 0.3282,
+      "step": 8313
+    },
+    {
+      "epoch": 0.6113684829766894,
+      "grad_norm": 0.5110467672348022,
+      "learning_rate": 6.931582628374933e-05,
+      "loss": 0.3543,
+      "step": 8314
+    },
+    {
+      "epoch": 0.6114420177954262,
+      "grad_norm": 0.451872855424881,
+      "learning_rate": 6.929315985103592e-05,
+      "loss": 0.3312,
+      "step": 8315
+    },
+    {
+      "epoch": 0.6115155526141628,
+      "grad_norm": 0.5393099784851074,
+      "learning_rate": 6.927049516004688e-05,
+      "loss": 0.3331,
+      "step": 8316
+    },
+    {
+      "epoch": 0.6115890874328995,
+      "grad_norm": 0.541728675365448,
+      "learning_rate": 6.924783221206784e-05,
+      "loss": 0.3333,
+      "step": 8317
+    },
+    {
+      "epoch": 0.6116626222516361,
+      "grad_norm": 0.621240496635437,
+      "learning_rate": 6.922517100838417e-05,
+      "loss": 0.346,
+      "step": 8318
+    },
+    {
+      "epoch": 0.6117361570703729,
+      "grad_norm": 0.5545910000801086,
+      "learning_rate": 6.920251155028127e-05,
+      "loss": 0.3498,
+      "step": 8319
+    },
+    {
+      "epoch": 0.6118096918891095,
+      "grad_norm": 0.740202784538269,
+      "learning_rate": 6.917985383904444e-05,
+      "loss": 0.3884,
+      "step": 8320
+    },
+    {
+      "epoch": 0.6118832267078461,
+      "grad_norm": 0.46962404251098633,
+      "learning_rate": 6.91571978759588e-05,
+      "loss": 0.3481,
+      "step": 8321
+    },
+    {
+      "epoch": 0.6119567615265828,
+      "grad_norm": 0.603115439414978,
+      "learning_rate": 6.913454366230947e-05,
+      "loss": 0.3539,
+      "step": 8322
+    },
+    {
+      "epoch": 0.6120302963453195,
+      "grad_norm": 0.5320051908493042,
+      "learning_rate": 6.911189119938139e-05,
+      "loss": 0.3337,
+      "step": 8323
+    },
+    {
+      "epoch": 0.6121038311640562,
+      "grad_norm": 0.5248119235038757,
+      "learning_rate": 6.908924048845943e-05,
+      "loss": 0.3584,
+      "step": 8324
+    },
+    {
+      "epoch": 0.6121773659827928,
+      "grad_norm": 0.5177549123764038,
+      "learning_rate": 6.906659153082842e-05,
+      "loss": 0.3454,
+      "step": 8325
+    },
+    {
+      "epoch": 0.6122509008015296,
+      "grad_norm": 0.6298940181732178,
+      "learning_rate": 6.904394432777293e-05,
+      "loss": 0.3203,
+      "step": 8326
+    },
+    {
+      "epoch": 0.6123244356202662,
+      "grad_norm": 0.5040342211723328,
+      "learning_rate": 6.902129888057761e-05,
+      "loss": 0.3179,
+      "step": 8327
+    },
+    {
+      "epoch": 0.6123979704390029,
+      "grad_norm": 0.685945451259613,
+      "learning_rate": 6.89986551905269e-05,
+      "loss": 0.3238,
+      "step": 8328
+    },
+    {
+      "epoch": 0.6124715052577395,
+      "grad_norm": 0.6029171943664551,
+      "learning_rate": 6.897601325890518e-05,
+      "loss": 0.3404,
+      "step": 8329
+    },
+    {
+      "epoch": 0.6125450400764763,
+      "grad_norm": 0.530807375907898,
+      "learning_rate": 6.895337308699674e-05,
+      "loss": 0.337,
+      "step": 8330
+    },
+    {
+      "epoch": 0.6126185748952129,
+      "grad_norm": 0.70269376039505,
+      "learning_rate": 6.893073467608575e-05,
+      "loss": 0.3421,
+      "step": 8331
+    },
+    {
+      "epoch": 0.6126921097139496,
+      "grad_norm": 0.6860878467559814,
+      "learning_rate": 6.890809802745624e-05,
+      "loss": 0.3413,
+      "step": 8332
+    },
+    {
+      "epoch": 0.6127656445326862,
+      "grad_norm": 0.6394241452217102,
+      "learning_rate": 6.888546314239226e-05,
+      "loss": 0.3187,
+      "step": 8333
+    },
+    {
+      "epoch": 0.6128391793514228,
+      "grad_norm": 0.5333149433135986,
+      "learning_rate": 6.886283002217761e-05,
+      "loss": 0.3287,
+      "step": 8334
+    },
+    {
+      "epoch": 0.6129127141701596,
+      "grad_norm": 0.7938823103904724,
+      "learning_rate": 6.884019866809615e-05,
+      "loss": 0.3558,
+      "step": 8335
+    },
+    {
+      "epoch": 0.6129862489888962,
+      "grad_norm": 0.7227214574813843,
+      "learning_rate": 6.881756908143149e-05,
+      "loss": 0.3518,
+      "step": 8336
+    },
+    {
+      "epoch": 0.6130597838076329,
+      "grad_norm": 0.531536340713501,
+      "learning_rate": 6.879494126346724e-05,
+      "loss": 0.3352,
+      "step": 8337
+    },
+    {
+      "epoch": 0.6131333186263696,
+      "grad_norm": 0.6247196793556213,
+      "learning_rate": 6.877231521548682e-05,
+      "loss": 0.3659,
+      "step": 8338
+    },
+    {
+      "epoch": 0.6132068534451063,
+      "grad_norm": 0.5598639845848083,
+      "learning_rate": 6.874969093877363e-05,
+      "loss": 0.3275,
+      "step": 8339
+    },
+    {
+      "epoch": 0.6132803882638429,
+      "grad_norm": 0.5121983885765076,
+      "learning_rate": 6.872706843461097e-05,
+      "loss": 0.3614,
+      "step": 8340
+    },
+    {
+      "epoch": 0.6133539230825796,
+      "grad_norm": 0.44049131870269775,
+      "learning_rate": 6.8704447704282e-05,
+      "loss": 0.3283,
+      "step": 8341
+    },
+    {
+      "epoch": 0.6134274579013163,
+      "grad_norm": 0.6327276229858398,
+      "learning_rate": 6.868182874906974e-05,
+      "loss": 0.3521,
+      "step": 8342
+    },
+    {
+      "epoch": 0.613500992720053,
+      "grad_norm": 0.6319581866264343,
+      "learning_rate": 6.865921157025725e-05,
+      "loss": 0.3457,
+      "step": 8343
+    },
+    {
+      "epoch": 0.6135745275387896,
+      "grad_norm": 0.6008892059326172,
+      "learning_rate": 6.863659616912731e-05,
+      "loss": 0.3257,
+      "step": 8344
+    },
+    {
+      "epoch": 0.6136480623575263,
+      "grad_norm": 0.7481988668441772,
+      "learning_rate": 6.861398254696278e-05,
+      "loss": 0.3406,
+      "step": 8345
+    },
+    {
+      "epoch": 0.613721597176263,
+      "grad_norm": 1.314017653465271,
+      "learning_rate": 6.859137070504627e-05,
+      "loss": 0.3471,
+      "step": 8346
+    },
+    {
+      "epoch": 0.6137951319949997,
+      "grad_norm": 0.5789337158203125,
+      "learning_rate": 6.856876064466035e-05,
+      "loss": 0.3384,
+      "step": 8347
+    },
+    {
+      "epoch": 0.6138686668137363,
+      "grad_norm": 0.8660134673118591,
+      "learning_rate": 6.85461523670875e-05,
+      "loss": 0.367,
+      "step": 8348
+    },
+    {
+      "epoch": 0.6139422016324729,
+      "grad_norm": 0.5407134890556335,
+      "learning_rate": 6.852354587361012e-05,
+      "loss": 0.3354,
+      "step": 8349
+    },
+    {
+      "epoch": 0.6140157364512097,
+      "grad_norm": 0.5319637656211853,
+      "learning_rate": 6.85009411655104e-05,
+      "loss": 0.3801,
+      "step": 8350
+    },
+    {
+      "epoch": 0.6140892712699463,
+      "grad_norm": 0.469117134809494,
+      "learning_rate": 6.847833824407055e-05,
+      "loss": 0.3414,
+      "step": 8351
+    },
+    {
+      "epoch": 0.614162806088683,
+      "grad_norm": 0.40109920501708984,
+      "learning_rate": 6.84557371105726e-05,
+      "loss": 0.3497,
+      "step": 8352
+    },
+    {
+      "epoch": 0.6142363409074196,
+      "grad_norm": 0.43230393528938293,
+      "learning_rate": 6.843313776629853e-05,
+      "loss": 0.3444,
+      "step": 8353
+    },
+    {
+      "epoch": 0.6143098757261564,
+      "grad_norm": 0.5139049291610718,
+      "learning_rate": 6.841054021253024e-05,
+      "loss": 0.327,
+      "step": 8354
+    },
+    {
+      "epoch": 0.614383410544893,
+      "grad_norm": 0.43783992528915405,
+      "learning_rate": 6.83879444505494e-05,
+      "loss": 0.3341,
+      "step": 8355
+    },
+    {
+      "epoch": 0.6144569453636297,
+      "grad_norm": 0.6554414629936218,
+      "learning_rate": 6.836535048163774e-05,
+      "loss": 0.3287,
+      "step": 8356
+    },
+    {
+      "epoch": 0.6145304801823663,
+      "grad_norm": 0.6836115121841431,
+      "learning_rate": 6.834275830707676e-05,
+      "loss": 0.3414,
+      "step": 8357
+    },
+    {
+      "epoch": 0.6146040150011031,
+      "grad_norm": 0.6662889719009399,
+      "learning_rate": 6.832016792814797e-05,
+      "loss": 0.3656,
+      "step": 8358
+    },
+    {
+      "epoch": 0.6146775498198397,
+      "grad_norm": 0.6233830451965332,
+      "learning_rate": 6.829757934613267e-05,
+      "loss": 0.3303,
+      "step": 8359
+    },
+    {
+      "epoch": 0.6147510846385764,
+      "grad_norm": 0.7907547354698181,
+      "learning_rate": 6.827499256231216e-05,
+      "loss": 0.3283,
+      "step": 8360
+    },
+    {
+      "epoch": 0.614824619457313,
+      "grad_norm": 0.4650387167930603,
+      "learning_rate": 6.825240757796753e-05,
+      "loss": 0.3519,
+      "step": 8361
+    },
+    {
+      "epoch": 0.6148981542760497,
+      "grad_norm": 0.7839797139167786,
+      "learning_rate": 6.822982439437983e-05,
+      "loss": 0.343,
+      "step": 8362
+    },
+    {
+      "epoch": 0.6149716890947864,
+      "grad_norm": 0.6193326115608215,
+      "learning_rate": 6.820724301283004e-05,
+      "loss": 0.3415,
+      "step": 8363
+    },
+    {
+      "epoch": 0.615045223913523,
+      "grad_norm": 0.8167579770088196,
+      "learning_rate": 6.818466343459897e-05,
+      "loss": 0.3458,
+      "step": 8364
+    },
+    {
+      "epoch": 0.6151187587322597,
+      "grad_norm": 0.7439516186714172,
+      "learning_rate": 6.816208566096735e-05,
+      "loss": 0.383,
+      "step": 8365
+    },
+    {
+      "epoch": 0.6151922935509964,
+      "grad_norm": 0.5862722396850586,
+      "learning_rate": 6.813950969321583e-05,
+      "loss": 0.3533,
+      "step": 8366
+    },
+    {
+      "epoch": 0.6152658283697331,
+      "grad_norm": 0.9367417693138123,
+      "learning_rate": 6.811693553262497e-05,
+      "loss": 0.3857,
+      "step": 8367
+    },
+    {
+      "epoch": 0.6153393631884697,
+      "grad_norm": 0.9503195285797119,
+      "learning_rate": 6.809436318047516e-05,
+      "loss": 0.3574,
+      "step": 8368
+    },
+    {
+      "epoch": 0.6154128980072064,
+      "grad_norm": 0.5763384699821472,
+      "learning_rate": 6.807179263804674e-05,
+      "loss": 0.3128,
+      "step": 8369
+    },
+    {
+      "epoch": 0.6154864328259431,
+      "grad_norm": 0.9705243110656738,
+      "learning_rate": 6.804922390661993e-05,
+      "loss": 0.3667,
+      "step": 8370
+    },
+    {
+      "epoch": 0.6155599676446798,
+      "grad_norm": 0.7000584006309509,
+      "learning_rate": 6.802665698747488e-05,
+      "loss": 0.3364,
+      "step": 8371
+    },
+    {
+      "epoch": 0.6156335024634164,
+      "grad_norm": 0.4271771013736725,
+      "learning_rate": 6.80040918818916e-05,
+      "loss": 0.3452,
+      "step": 8372
+    },
+    {
+      "epoch": 0.6157070372821531,
+      "grad_norm": 0.661976158618927,
+      "learning_rate": 6.798152859114999e-05,
+      "loss": 0.3424,
+      "step": 8373
+    },
+    {
+      "epoch": 0.6157805721008898,
+      "grad_norm": 0.5139139294624329,
+      "learning_rate": 6.795896711652985e-05,
+      "loss": 0.3518,
+      "step": 8374
+    },
+    {
+      "epoch": 0.6158541069196265,
+      "grad_norm": 0.49311748147010803,
+      "learning_rate": 6.79364074593109e-05,
+      "loss": 0.36,
+      "step": 8375
+    },
+    {
+      "epoch": 0.6159276417383631,
+      "grad_norm": 0.48085445165634155,
+      "learning_rate": 6.791384962077277e-05,
+      "loss": 0.3706,
+      "step": 8376
+    },
+    {
+      "epoch": 0.6160011765570997,
+      "grad_norm": 0.570271372795105,
+      "learning_rate": 6.789129360219496e-05,
+      "loss": 0.332,
+      "step": 8377
+    },
+    {
+      "epoch": 0.6160747113758365,
+      "grad_norm": 0.6474676132202148,
+      "learning_rate": 6.786873940485684e-05,
+      "loss": 0.3456,
+      "step": 8378
+    },
+    {
+      "epoch": 0.6161482461945731,
+      "grad_norm": 0.453967809677124,
+      "learning_rate": 6.784618703003775e-05,
+      "loss": 0.3489,
+      "step": 8379
+    },
+    {
+      "epoch": 0.6162217810133098,
+      "grad_norm": 0.5798060297966003,
+      "learning_rate": 6.782363647901683e-05,
+      "loss": 0.3457,
+      "step": 8380
+    },
+    {
+      "epoch": 0.6162953158320464,
+      "grad_norm": 0.5110371112823486,
+      "learning_rate": 6.780108775307325e-05,
+      "loss": 0.325,
+      "step": 8381
+    },
+    {
+      "epoch": 0.6163688506507832,
+      "grad_norm": 0.6858824491500854,
+      "learning_rate": 6.777854085348594e-05,
+      "loss": 0.3299,
+      "step": 8382
+    },
+    {
+      "epoch": 0.6164423854695198,
+      "grad_norm": 0.5039066076278687,
+      "learning_rate": 6.775599578153377e-05,
+      "loss": 0.3623,
+      "step": 8383
+    },
+    {
+      "epoch": 0.6165159202882565,
+      "grad_norm": 0.5570119023323059,
+      "learning_rate": 6.773345253849563e-05,
+      "loss": 0.37,
+      "step": 8384
+    },
+    {
+      "epoch": 0.6165894551069931,
+      "grad_norm": 0.5907719731330872,
+      "learning_rate": 6.771091112565004e-05,
+      "loss": 0.3172,
+      "step": 8385
+    },
+    {
+      "epoch": 0.6166629899257299,
+      "grad_norm": 0.43492263555526733,
+      "learning_rate": 6.768837154427567e-05,
+      "loss": 0.322,
+      "step": 8386
+    },
+    {
+      "epoch": 0.6167365247444665,
+      "grad_norm": 0.5434353351593018,
+      "learning_rate": 6.766583379565097e-05,
+      "loss": 0.3173,
+      "step": 8387
+    },
+    {
+      "epoch": 0.6168100595632032,
+      "grad_norm": 0.8707041144371033,
+      "learning_rate": 6.764329788105428e-05,
+      "loss": 0.3755,
+      "step": 8388
+    },
+    {
+      "epoch": 0.6168835943819398,
+      "grad_norm": 0.7405357360839844,
+      "learning_rate": 6.76207638017639e-05,
+      "loss": 0.3318,
+      "step": 8389
+    },
+    {
+      "epoch": 0.6169571292006765,
+      "grad_norm": 0.631189227104187,
+      "learning_rate": 6.759823155905795e-05,
+      "loss": 0.3269,
+      "step": 8390
+    },
+    {
+      "epoch": 0.6170306640194132,
+      "grad_norm": 0.9455764889717102,
+      "learning_rate": 6.757570115421451e-05,
+      "loss": 0.3336,
+      "step": 8391
+    },
+    {
+      "epoch": 0.6171041988381498,
+      "grad_norm": 0.7538967728614807,
+      "learning_rate": 6.755317258851152e-05,
+      "loss": 0.3002,
+      "step": 8392
+    },
+    {
+      "epoch": 0.6171777336568866,
+      "grad_norm": 0.810497522354126,
+      "learning_rate": 6.753064586322682e-05,
+      "loss": 0.3603,
+      "step": 8393
+    },
+    {
+      "epoch": 0.6172512684756232,
+      "grad_norm": 1.628888487815857,
+      "learning_rate": 6.750812097963815e-05,
+      "loss": 0.3664,
+      "step": 8394
+    },
+    {
+      "epoch": 0.6173248032943599,
+      "grad_norm": 1.2348833084106445,
+      "learning_rate": 6.748559793902319e-05,
+      "loss": 0.3435,
+      "step": 8395
+    },
+    {
+      "epoch": 0.6173983381130965,
+      "grad_norm": 1.8332871198654175,
+      "learning_rate": 6.746307674265939e-05,
+      "loss": 0.3331,
+      "step": 8396
+    },
+    {
+      "epoch": 0.6174718729318333,
+      "grad_norm": 2.061022996902466,
+      "learning_rate": 6.744055739182423e-05,
+      "loss": 0.3516,
+      "step": 8397
+    },
+    {
+      "epoch": 0.6175454077505699,
+      "grad_norm": 0.5989073514938354,
+      "learning_rate": 6.7418039887795e-05,
+      "loss": 0.3492,
+      "step": 8398
+    },
+    {
+      "epoch": 0.6176189425693066,
+      "grad_norm": 1.5299817323684692,
+      "learning_rate": 6.739552423184894e-05,
+      "loss": 0.3515,
+      "step": 8399
+    },
+    {
+      "epoch": 0.6176924773880432,
+      "grad_norm": 0.5938190221786499,
+      "learning_rate": 6.737301042526316e-05,
+      "loss": 0.3453,
+      "step": 8400
+    },
+    {
+      "epoch": 0.61776601220678,
+      "grad_norm": 1.4079252481460571,
+      "learning_rate": 6.735049846931462e-05,
+      "loss": 0.3518,
+      "step": 8401
+    },
+    {
+      "epoch": 0.6178395470255166,
+      "grad_norm": 0.5832533240318298,
+      "learning_rate": 6.732798836528031e-05,
+      "loss": 0.3699,
+      "step": 8402
+    },
+    {
+      "epoch": 0.6179130818442533,
+      "grad_norm": 1.1297763586044312,
+      "learning_rate": 6.730548011443696e-05,
+      "loss": 0.3282,
+      "step": 8403
+    },
+    {
+      "epoch": 0.6179866166629899,
+      "grad_norm": 0.7542088627815247,
+      "learning_rate": 6.72829737180613e-05,
+      "loss": 0.3497,
+      "step": 8404
+    },
+    {
+      "epoch": 0.6180601514817265,
+      "grad_norm": 0.6578119993209839,
+      "learning_rate": 6.726046917742987e-05,
+      "loss": 0.3263,
+      "step": 8405
+    },
+    {
+      "epoch": 0.6181336863004633,
+      "grad_norm": 0.45384201407432556,
+      "learning_rate": 6.723796649381919e-05,
+      "loss": 0.3322,
+      "step": 8406
+    },
+    {
+      "epoch": 0.6182072211191999,
+      "grad_norm": 0.8029436469078064,
+      "learning_rate": 6.721546566850568e-05,
+      "loss": 0.3227,
+      "step": 8407
+    },
+    {
+      "epoch": 0.6182807559379366,
+      "grad_norm": 0.6478369235992432,
+      "learning_rate": 6.719296670276549e-05,
+      "loss": 0.3502,
+      "step": 8408
+    },
+    {
+      "epoch": 0.6183542907566733,
+      "grad_norm": 0.7143117785453796,
+      "learning_rate": 6.717046959787488e-05,
+      "loss": 0.3367,
+      "step": 8409
+    },
+    {
+      "epoch": 0.61842782557541,
+      "grad_norm": 0.8662980198860168,
+      "learning_rate": 6.714797435510987e-05,
+      "loss": 0.3588,
+      "step": 8410
+    },
+    {
+      "epoch": 0.6185013603941466,
+      "grad_norm": 0.5971333384513855,
+      "learning_rate": 6.712548097574642e-05,
+      "loss": 0.3229,
+      "step": 8411
+    },
+    {
+      "epoch": 0.6185748952128833,
+      "grad_norm": 1.1185784339904785,
+      "learning_rate": 6.710298946106039e-05,
+      "loss": 0.3375,
+      "step": 8412
+    },
+    {
+      "epoch": 0.61864843003162,
+      "grad_norm": 0.536197304725647,
+      "learning_rate": 6.708049981232751e-05,
+      "loss": 0.3381,
+      "step": 8413
+    },
+    {
+      "epoch": 0.6187219648503567,
+      "grad_norm": 0.9140260815620422,
+      "learning_rate": 6.705801203082342e-05,
+      "loss": 0.3561,
+      "step": 8414
+    },
+    {
+      "epoch": 0.6187954996690933,
+      "grad_norm": 0.6861332654953003,
+      "learning_rate": 6.703552611782366e-05,
+      "loss": 0.3411,
+      "step": 8415
+    },
+    {
+      "epoch": 0.61886903448783,
+      "grad_norm": 0.7169490456581116,
+      "learning_rate": 6.701304207460364e-05,
+      "loss": 0.3118,
+      "step": 8416
+    },
+    {
+      "epoch": 0.6189425693065667,
+      "grad_norm": 0.9463632702827454,
+      "learning_rate": 6.699055990243868e-05,
+      "loss": 0.3779,
+      "step": 8417
+    },
+    {
+      "epoch": 0.6190161041253033,
+      "grad_norm": 0.45719847083091736,
+      "learning_rate": 6.696807960260402e-05,
+      "loss": 0.3791,
+      "step": 8418
+    },
+    {
+      "epoch": 0.61908963894404,
+      "grad_norm": 0.6164475083351135,
+      "learning_rate": 6.694560117637476e-05,
+      "loss": 0.3149,
+      "step": 8419
+    },
+    {
+      "epoch": 0.6191631737627766,
+      "grad_norm": 0.4954219162464142,
+      "learning_rate": 6.692312462502587e-05,
+      "loss": 0.3374,
+      "step": 8420
+    },
+    {
+      "epoch": 0.6192367085815134,
+      "grad_norm": 0.7151349782943726,
+      "learning_rate": 6.690064994983222e-05,
+      "loss": 0.3569,
+      "step": 8421
+    },
+    {
+      "epoch": 0.61931024340025,
+      "grad_norm": 0.9292550683021545,
+      "learning_rate": 6.687817715206868e-05,
+      "loss": 0.354,
+      "step": 8422
+    },
+    {
+      "epoch": 0.6193837782189867,
+      "grad_norm": 0.511722981929779,
+      "learning_rate": 6.685570623300987e-05,
+      "loss": 0.3364,
+      "step": 8423
+    },
+    {
+      "epoch": 0.6194573130377233,
+      "grad_norm": 0.9432550668716431,
+      "learning_rate": 6.683323719393038e-05,
+      "loss": 0.3126,
+      "step": 8424
+    },
+    {
+      "epoch": 0.6195308478564601,
+      "grad_norm": 0.6192879676818848,
+      "learning_rate": 6.681077003610467e-05,
+      "loss": 0.3461,
+      "step": 8425
+    },
+    {
+      "epoch": 0.6196043826751967,
+      "grad_norm": 0.6254086494445801,
+      "learning_rate": 6.678830476080712e-05,
+      "loss": 0.3581,
+      "step": 8426
+    },
+    {
+      "epoch": 0.6196779174939334,
+      "grad_norm": 0.7703858613967896,
+      "learning_rate": 6.676584136931198e-05,
+      "loss": 0.3253,
+      "step": 8427
+    },
+    {
+      "epoch": 0.61975145231267,
+      "grad_norm": 0.460421621799469,
+      "learning_rate": 6.674337986289339e-05,
+      "loss": 0.3525,
+      "step": 8428
+    },
+    {
+      "epoch": 0.6198249871314068,
+      "grad_norm": 0.7238818407058716,
+      "learning_rate": 6.672092024282537e-05,
+      "loss": 0.3586,
+      "step": 8429
+    },
+    {
+      "epoch": 0.6198985219501434,
+      "grad_norm": 0.572833240032196,
+      "learning_rate": 6.669846251038194e-05,
+      "loss": 0.3259,
+      "step": 8430
+    },
+    {
+      "epoch": 0.6199720567688801,
+      "grad_norm": 0.5188858509063721,
+      "learning_rate": 6.66760066668368e-05,
+      "loss": 0.3439,
+      "step": 8431
+    },
+    {
+      "epoch": 0.6200455915876167,
+      "grad_norm": 0.5451098084449768,
+      "learning_rate": 6.665355271346374e-05,
+      "loss": 0.3317,
+      "step": 8432
+    },
+    {
+      "epoch": 0.6201191264063534,
+      "grad_norm": 0.820017397403717,
+      "learning_rate": 6.663110065153637e-05,
+      "loss": 0.3492,
+      "step": 8433
+    },
+    {
+      "epoch": 0.6201926612250901,
+      "grad_norm": 0.7426794171333313,
+      "learning_rate": 6.660865048232815e-05,
+      "loss": 0.3395,
+      "step": 8434
+    },
+    {
+      "epoch": 0.6202661960438267,
+      "grad_norm": 0.5055623650550842,
+      "learning_rate": 6.658620220711256e-05,
+      "loss": 0.3719,
+      "step": 8435
+    },
+    {
+      "epoch": 0.6203397308625634,
+      "grad_norm": 0.850547730922699,
+      "learning_rate": 6.656375582716283e-05,
+      "loss": 0.3438,
+      "step": 8436
+    },
+    {
+      "epoch": 0.6204132656813001,
+      "grad_norm": 0.4652886390686035,
+      "learning_rate": 6.654131134375214e-05,
+      "loss": 0.3303,
+      "step": 8437
+    },
+    {
+      "epoch": 0.6204868005000368,
+      "grad_norm": 0.6725755333900452,
+      "learning_rate": 6.651886875815361e-05,
+      "loss": 0.358,
+      "step": 8438
+    },
+    {
+      "epoch": 0.6205603353187734,
+      "grad_norm": 0.44186586141586304,
+      "learning_rate": 6.649642807164016e-05,
+      "loss": 0.3394,
+      "step": 8439
+    },
+    {
+      "epoch": 0.6206338701375101,
+      "grad_norm": 0.8604176044464111,
+      "learning_rate": 6.647398928548467e-05,
+      "loss": 0.3392,
+      "step": 8440
+    },
+    {
+      "epoch": 0.6207074049562468,
+      "grad_norm": 0.7330872416496277,
+      "learning_rate": 6.645155240095989e-05,
+      "loss": 0.3422,
+      "step": 8441
+    },
+    {
+      "epoch": 0.6207809397749835,
+      "grad_norm": 0.4496297240257263,
+      "learning_rate": 6.642911741933848e-05,
+      "loss": 0.3322,
+      "step": 8442
+    },
+    {
+      "epoch": 0.6208544745937201,
+      "grad_norm": 0.4743311107158661,
+      "learning_rate": 6.640668434189296e-05,
+      "loss": 0.3312,
+      "step": 8443
+    },
+    {
+      "epoch": 0.6209280094124568,
+      "grad_norm": 0.40196463465690613,
+      "learning_rate": 6.638425316989571e-05,
+      "loss": 0.3391,
+      "step": 8444
+    },
+    {
+      "epoch": 0.6210015442311935,
+      "grad_norm": 0.4925621449947357,
+      "learning_rate": 6.636182390461914e-05,
+      "loss": 0.3232,
+      "step": 8445
+    },
+    {
+      "epoch": 0.6210750790499301,
+      "grad_norm": 0.516791045665741,
+      "learning_rate": 6.63393965473354e-05,
+      "loss": 0.3847,
+      "step": 8446
+    },
+    {
+      "epoch": 0.6211486138686668,
+      "grad_norm": 0.5537052154541016,
+      "learning_rate": 6.631697109931661e-05,
+      "loss": 0.3572,
+      "step": 8447
+    },
+    {
+      "epoch": 0.6212221486874034,
+      "grad_norm": 0.5032663941383362,
+      "learning_rate": 6.629454756183478e-05,
+      "loss": 0.3514,
+      "step": 8448
+    },
+    {
+      "epoch": 0.6212956835061402,
+      "grad_norm": 0.5217093229293823,
+      "learning_rate": 6.627212593616175e-05,
+      "loss": 0.3363,
+      "step": 8449
+    },
+    {
+      "epoch": 0.6213692183248768,
+      "grad_norm": 0.5092450976371765,
+      "learning_rate": 6.624970622356937e-05,
+      "loss": 0.3329,
+      "step": 8450
+    },
+    {
+      "epoch": 0.6214427531436135,
+      "grad_norm": 0.4424246847629547,
+      "learning_rate": 6.622728842532926e-05,
+      "loss": 0.3409,
+      "step": 8451
+    },
+    {
+      "epoch": 0.6215162879623501,
+      "grad_norm": 0.4799439311027527,
+      "learning_rate": 6.620487254271297e-05,
+      "loss": 0.3206,
+      "step": 8452
+    },
+    {
+      "epoch": 0.6215898227810869,
+      "grad_norm": 0.50055330991745,
+      "learning_rate": 6.6182458576992e-05,
+      "loss": 0.3364,
+      "step": 8453
+    },
+    {
+      "epoch": 0.6216633575998235,
+      "grad_norm": 0.517788827419281,
+      "learning_rate": 6.616004652943768e-05,
+      "loss": 0.3424,
+      "step": 8454
+    },
+    {
+      "epoch": 0.6217368924185602,
+      "grad_norm": 0.5375692844390869,
+      "learning_rate": 6.613763640132123e-05,
+      "loss": 0.3573,
+      "step": 8455
+    },
+    {
+      "epoch": 0.6218104272372968,
+      "grad_norm": 0.4875009059906006,
+      "learning_rate": 6.611522819391376e-05,
+      "loss": 0.3206,
+      "step": 8456
+    },
+    {
+      "epoch": 0.6218839620560336,
+      "grad_norm": 0.536723792552948,
+      "learning_rate": 6.609282190848629e-05,
+      "loss": 0.3255,
+      "step": 8457
+    },
+    {
+      "epoch": 0.6219574968747702,
+      "grad_norm": 0.4521346986293793,
+      "learning_rate": 6.607041754630978e-05,
+      "loss": 0.3492,
+      "step": 8458
+    },
+    {
+      "epoch": 0.6220310316935069,
+      "grad_norm": 0.7057673335075378,
+      "learning_rate": 6.6048015108655e-05,
+      "loss": 0.3369,
+      "step": 8459
+    },
+    {
+      "epoch": 0.6221045665122436,
+      "grad_norm": 0.512315571308136,
+      "learning_rate": 6.602561459679259e-05,
+      "loss": 0.3224,
+      "step": 8460
+    },
+    {
+      "epoch": 0.6221781013309802,
+      "grad_norm": 0.7590563893318176,
+      "learning_rate": 6.60032160119932e-05,
+      "loss": 0.3489,
+      "step": 8461
+    },
+    {
+      "epoch": 0.6222516361497169,
+      "grad_norm": 0.6475071310997009,
+      "learning_rate": 6.598081935552726e-05,
+      "loss": 0.3419,
+      "step": 8462
+    },
+    {
+      "epoch": 0.6223251709684535,
+      "grad_norm": 0.5368239879608154,
+      "learning_rate": 6.595842462866518e-05,
+      "loss": 0.3653,
+      "step": 8463
+    },
+    {
+      "epoch": 0.6223987057871903,
+      "grad_norm": 0.8513384461402893,
+      "learning_rate": 6.593603183267716e-05,
+      "loss": 0.3193,
+      "step": 8464
+    },
+    {
+      "epoch": 0.6224722406059269,
+      "grad_norm": 0.7286442518234253,
+      "learning_rate": 6.591364096883339e-05,
+      "loss": 0.3418,
+      "step": 8465
+    },
+    {
+      "epoch": 0.6225457754246636,
+      "grad_norm": 0.6571409702301025,
+      "learning_rate": 6.589125203840386e-05,
+      "loss": 0.3288,
+      "step": 8466
+    },
+    {
+      "epoch": 0.6226193102434002,
+      "grad_norm": 1.1354389190673828,
+      "learning_rate": 6.586886504265849e-05,
+      "loss": 0.3384,
+      "step": 8467
+    },
+    {
+      "epoch": 0.622692845062137,
+      "grad_norm": 0.7229756712913513,
+      "learning_rate": 6.584647998286714e-05,
+      "loss": 0.3339,
+      "step": 8468
+    },
+    {
+      "epoch": 0.6227663798808736,
+      "grad_norm": 0.7477405071258545,
+      "learning_rate": 6.582409686029948e-05,
+      "loss": 0.3445,
+      "step": 8469
+    },
+    {
+      "epoch": 0.6228399146996103,
+      "grad_norm": 0.9569599628448486,
+      "learning_rate": 6.58017156762251e-05,
+      "loss": 0.3338,
+      "step": 8470
+    },
+    {
+      "epoch": 0.6229134495183469,
+      "grad_norm": 0.5043314695358276,
+      "learning_rate": 6.577933643191352e-05,
+      "loss": 0.3472,
+      "step": 8471
+    },
+    {
+      "epoch": 0.6229869843370837,
+      "grad_norm": 1.2951127290725708,
+      "learning_rate": 6.575695912863407e-05,
+      "loss": 0.3499,
+      "step": 8472
+    },
+    {
+      "epoch": 0.6230605191558203,
+      "grad_norm": 0.6567380428314209,
+      "learning_rate": 6.573458376765605e-05,
+      "loss": 0.3401,
+      "step": 8473
+    },
+    {
+      "epoch": 0.6231340539745569,
+      "grad_norm": 0.6729965806007385,
+      "learning_rate": 6.57122103502486e-05,
+      "loss": 0.3265,
+      "step": 8474
+    },
+    {
+      "epoch": 0.6232075887932936,
+      "grad_norm": 0.6155994534492493,
+      "learning_rate": 6.568983887768074e-05,
+      "loss": 0.3376,
+      "step": 8475
+    },
+    {
+      "epoch": 0.6232811236120303,
+      "grad_norm": 0.7629123330116272,
+      "learning_rate": 6.566746935122146e-05,
+      "loss": 0.3418,
+      "step": 8476
+    },
+    {
+      "epoch": 0.623354658430767,
+      "grad_norm": 0.4626353681087494,
+      "learning_rate": 6.564510177213957e-05,
+      "loss": 0.3311,
+      "step": 8477
+    },
+    {
+      "epoch": 0.6234281932495036,
+      "grad_norm": 0.6056297421455383,
+      "learning_rate": 6.562273614170375e-05,
+      "loss": 0.3097,
+      "step": 8478
+    },
+    {
+      "epoch": 0.6235017280682403,
+      "grad_norm": 0.7352742552757263,
+      "learning_rate": 6.56003724611826e-05,
+      "loss": 0.3249,
+      "step": 8479
+    },
+    {
+      "epoch": 0.623575262886977,
+      "grad_norm": 0.7524993419647217,
+      "learning_rate": 6.557801073184461e-05,
+      "loss": 0.3465,
+      "step": 8480
+    },
+    {
+      "epoch": 0.6236487977057137,
+      "grad_norm": 0.6060985326766968,
+      "learning_rate": 6.555565095495821e-05,
+      "loss": 0.3442,
+      "step": 8481
+    },
+    {
+      "epoch": 0.6237223325244503,
+      "grad_norm": 0.693768322467804,
+      "learning_rate": 6.553329313179162e-05,
+      "loss": 0.3136,
+      "step": 8482
+    },
+    {
+      "epoch": 0.623795867343187,
+      "grad_norm": 0.6807841062545776,
+      "learning_rate": 6.551093726361301e-05,
+      "loss": 0.3309,
+      "step": 8483
+    },
+    {
+      "epoch": 0.6238694021619237,
+      "grad_norm": 0.6122098565101624,
+      "learning_rate": 6.548858335169046e-05,
+      "loss": 0.3132,
+      "step": 8484
+    },
+    {
+      "epoch": 0.6239429369806604,
+      "grad_norm": 0.5971434116363525,
+      "learning_rate": 6.546623139729186e-05,
+      "loss": 0.3335,
+      "step": 8485
+    },
+    {
+      "epoch": 0.624016471799397,
+      "grad_norm": 0.5724837779998779,
+      "learning_rate": 6.544388140168507e-05,
+      "loss": 0.3269,
+      "step": 8486
+    },
+    {
+      "epoch": 0.6240900066181337,
+      "grad_norm": 0.8157625794410706,
+      "learning_rate": 6.54215333661378e-05,
+      "loss": 0.328,
+      "step": 8487
+    },
+    {
+      "epoch": 0.6241635414368704,
+      "grad_norm": 0.4963706135749817,
+      "learning_rate": 6.539918729191764e-05,
+      "loss": 0.3367,
+      "step": 8488
+    },
+    {
+      "epoch": 0.624237076255607,
+      "grad_norm": 0.7208766937255859,
+      "learning_rate": 6.537684318029215e-05,
+      "loss": 0.3222,
+      "step": 8489
+    },
+    {
+      "epoch": 0.6243106110743437,
+      "grad_norm": 0.4942074418067932,
+      "learning_rate": 6.535450103252859e-05,
+      "loss": 0.3407,
+      "step": 8490
+    },
+    {
+      "epoch": 0.6243841458930803,
+      "grad_norm": 0.6435662508010864,
+      "learning_rate": 6.533216084989431e-05,
+      "loss": 0.3425,
+      "step": 8491
+    },
+    {
+      "epoch": 0.6244576807118171,
+      "grad_norm": 0.5725953578948975,
+      "learning_rate": 6.530982263365646e-05,
+      "loss": 0.334,
+      "step": 8492
+    },
+    {
+      "epoch": 0.6245312155305537,
+      "grad_norm": 0.6907104253768921,
+      "learning_rate": 6.528748638508206e-05,
+      "loss": 0.3305,
+      "step": 8493
+    },
+    {
+      "epoch": 0.6246047503492904,
+      "grad_norm": 0.5641850233078003,
+      "learning_rate": 6.526515210543809e-05,
+      "loss": 0.3225,
+      "step": 8494
+    },
+    {
+      "epoch": 0.624678285168027,
+      "grad_norm": 0.49998557567596436,
+      "learning_rate": 6.524281979599131e-05,
+      "loss": 0.3243,
+      "step": 8495
+    },
+    {
+      "epoch": 0.6247518199867638,
+      "grad_norm": 0.5979667901992798,
+      "learning_rate": 6.522048945800851e-05,
+      "loss": 0.344,
+      "step": 8496
+    },
+    {
+      "epoch": 0.6248253548055004,
+      "grad_norm": 0.4413849711418152,
+      "learning_rate": 6.519816109275625e-05,
+      "loss": 0.3671,
+      "step": 8497
+    },
+    {
+      "epoch": 0.6248988896242371,
+      "grad_norm": 0.6194885969161987,
+      "learning_rate": 6.517583470150099e-05,
+      "loss": 0.344,
+      "step": 8498
+    },
+    {
+      "epoch": 0.6249724244429737,
+      "grad_norm": 0.5811423659324646,
+      "learning_rate": 6.515351028550916e-05,
+      "loss": 0.3313,
+      "step": 8499
+    },
+    {
+      "epoch": 0.6250459592617105,
+      "grad_norm": 0.651570200920105,
+      "learning_rate": 6.513118784604701e-05,
+      "loss": 0.3483,
+      "step": 8500
+    },
+    {
+      "epoch": 0.6251194940804471,
+      "grad_norm": 0.5101898312568665,
+      "learning_rate": 6.510886738438071e-05,
+      "loss": 0.3449,
+      "step": 8501
+    },
+    {
+      "epoch": 0.6251930288991837,
+      "grad_norm": 0.6413899064064026,
+      "learning_rate": 6.508654890177624e-05,
+      "loss": 0.3504,
+      "step": 8502
+    },
+    {
+      "epoch": 0.6252665637179204,
+      "grad_norm": 0.5245741009712219,
+      "learning_rate": 6.506423239949955e-05,
+      "loss": 0.3342,
+      "step": 8503
+    },
+    {
+      "epoch": 0.6253400985366571,
+      "grad_norm": 0.7054469585418701,
+      "learning_rate": 6.504191787881648e-05,
+      "loss": 0.3487,
+      "step": 8504
+    },
+    {
+      "epoch": 0.6254136333553938,
+      "grad_norm": 0.6113607287406921,
+      "learning_rate": 6.501960534099273e-05,
+      "loss": 0.347,
+      "step": 8505
+    },
+    {
+      "epoch": 0.6254871681741304,
+      "grad_norm": 0.602354109287262,
+      "learning_rate": 6.499729478729385e-05,
+      "loss": 0.3436,
+      "step": 8506
+    },
+    {
+      "epoch": 0.6255607029928671,
+      "grad_norm": 0.8094286918640137,
+      "learning_rate": 6.497498621898537e-05,
+      "loss": 0.3389,
+      "step": 8507
+    },
+    {
+      "epoch": 0.6256342378116038,
+      "grad_norm": 0.5210868120193481,
+      "learning_rate": 6.495267963733263e-05,
+      "loss": 0.3358,
+      "step": 8508
+    },
+    {
+      "epoch": 0.6257077726303405,
+      "grad_norm": 0.5327731966972351,
+      "learning_rate": 6.493037504360089e-05,
+      "loss": 0.3383,
+      "step": 8509
+    },
+    {
+      "epoch": 0.6257813074490771,
+      "grad_norm": 0.7469713091850281,
+      "learning_rate": 6.49080724390553e-05,
+      "loss": 0.3486,
+      "step": 8510
+    },
+    {
+      "epoch": 0.6258548422678138,
+      "grad_norm": 0.5468418002128601,
+      "learning_rate": 6.488577182496085e-05,
+      "loss": 0.362,
+      "step": 8511
+    },
+    {
+      "epoch": 0.6259283770865505,
+      "grad_norm": 0.7598193883895874,
+      "learning_rate": 6.486347320258255e-05,
+      "loss": 0.3371,
+      "step": 8512
+    },
+    {
+      "epoch": 0.6260019119052872,
+      "grad_norm": 0.49397045373916626,
+      "learning_rate": 6.484117657318505e-05,
+      "loss": 0.3238,
+      "step": 8513
+    },
+    {
+      "epoch": 0.6260754467240238,
+      "grad_norm": 0.5724345445632935,
+      "learning_rate": 6.481888193803316e-05,
+      "loss": 0.327,
+      "step": 8514
+    },
+    {
+      "epoch": 0.6261489815427606,
+      "grad_norm": 0.45921051502227783,
+      "learning_rate": 6.479658929839141e-05,
+      "loss": 0.3485,
+      "step": 8515
+    },
+    {
+      "epoch": 0.6262225163614972,
+      "grad_norm": 0.4467390179634094,
+      "learning_rate": 6.477429865552423e-05,
+      "loss": 0.3275,
+      "step": 8516
+    },
+    {
+      "epoch": 0.6262960511802338,
+      "grad_norm": 0.5108541250228882,
+      "learning_rate": 6.475201001069604e-05,
+      "loss": 0.3345,
+      "step": 8517
+    },
+    {
+      "epoch": 0.6263695859989705,
+      "grad_norm": 0.5665149688720703,
+      "learning_rate": 6.472972336517105e-05,
+      "loss": 0.3385,
+      "step": 8518
+    },
+    {
+      "epoch": 0.6264431208177071,
+      "grad_norm": 0.46791914105415344,
+      "learning_rate": 6.470743872021335e-05,
+      "loss": 0.3207,
+      "step": 8519
+    },
+    {
+      "epoch": 0.6265166556364439,
+      "grad_norm": 0.7056413292884827,
+      "learning_rate": 6.468515607708696e-05,
+      "loss": 0.3481,
+      "step": 8520
+    },
+    {
+      "epoch": 0.6265901904551805,
+      "grad_norm": 0.4694414436817169,
+      "learning_rate": 6.46628754370558e-05,
+      "loss": 0.3346,
+      "step": 8521
+    },
+    {
+      "epoch": 0.6266637252739172,
+      "grad_norm": 0.6654601693153381,
+      "learning_rate": 6.464059680138364e-05,
+      "loss": 0.3574,
+      "step": 8522
+    },
+    {
+      "epoch": 0.6267372600926538,
+      "grad_norm": 0.4563503563404083,
+      "learning_rate": 6.461832017133416e-05,
+      "loss": 0.344,
+      "step": 8523
+    },
+    {
+      "epoch": 0.6268107949113906,
+      "grad_norm": 0.6564285755157471,
+      "learning_rate": 6.459604554817091e-05,
+      "loss": 0.3394,
+      "step": 8524
+    },
+    {
+      "epoch": 0.6268843297301272,
+      "grad_norm": 0.4815804958343506,
+      "learning_rate": 6.45737729331573e-05,
+      "loss": 0.3316,
+      "step": 8525
+    },
+    {
+      "epoch": 0.6269578645488639,
+      "grad_norm": 0.5909795165061951,
+      "learning_rate": 6.455150232755667e-05,
+      "loss": 0.333,
+      "step": 8526
+    },
+    {
+      "epoch": 0.6270313993676006,
+      "grad_norm": 0.4938811957836151,
+      "learning_rate": 6.452923373263224e-05,
+      "loss": 0.3377,
+      "step": 8527
+    },
+    {
+      "epoch": 0.6271049341863373,
+      "grad_norm": 0.5509076118469238,
+      "learning_rate": 6.450696714964712e-05,
+      "loss": 0.3584,
+      "step": 8528
+    },
+    {
+      "epoch": 0.6271784690050739,
+      "grad_norm": 0.655487596988678,
+      "learning_rate": 6.448470257986426e-05,
+      "loss": 0.3612,
+      "step": 8529
+    },
+    {
+      "epoch": 0.6272520038238105,
+      "grad_norm": 0.6907817125320435,
+      "learning_rate": 6.446244002454655e-05,
+      "loss": 0.3635,
+      "step": 8530
+    },
+    {
+      "epoch": 0.6273255386425473,
+      "grad_norm": 0.6023157835006714,
+      "learning_rate": 6.444017948495675e-05,
+      "loss": 0.3372,
+      "step": 8531
+    },
+    {
+      "epoch": 0.6273990734612839,
+      "grad_norm": 0.7175008654594421,
+      "learning_rate": 6.44179209623575e-05,
+      "loss": 0.3788,
+      "step": 8532
+    },
+    {
+      "epoch": 0.6274726082800206,
+      "grad_norm": 0.5059716701507568,
+      "learning_rate": 6.439566445801133e-05,
+      "loss": 0.3468,
+      "step": 8533
+    },
+    {
+      "epoch": 0.6275461430987572,
+      "grad_norm": 0.947133481502533,
+      "learning_rate": 6.437340997318064e-05,
+      "loss": 0.3476,
+      "step": 8534
+    },
+    {
+      "epoch": 0.627619677917494,
+      "grad_norm": 0.8290666937828064,
+      "learning_rate": 6.435115750912775e-05,
+      "loss": 0.3373,
+      "step": 8535
+    },
+    {
+      "epoch": 0.6276932127362306,
+      "grad_norm": 0.5548955202102661,
+      "learning_rate": 6.432890706711483e-05,
+      "loss": 0.3297,
+      "step": 8536
+    },
+    {
+      "epoch": 0.6277667475549673,
+      "grad_norm": 1.27918541431427,
+      "learning_rate": 6.430665864840394e-05,
+      "loss": 0.3433,
+      "step": 8537
+    },
+    {
+      "epoch": 0.6278402823737039,
+      "grad_norm": 0.50139319896698,
+      "learning_rate": 6.428441225425703e-05,
+      "loss": 0.3627,
+      "step": 8538
+    },
+    {
+      "epoch": 0.6279138171924407,
+      "grad_norm": 0.8493871092796326,
+      "learning_rate": 6.426216788593594e-05,
+      "loss": 0.336,
+      "step": 8539
+    },
+    {
+      "epoch": 0.6279873520111773,
+      "grad_norm": 0.5063170790672302,
+      "learning_rate": 6.423992554470242e-05,
+      "loss": 0.3384,
+      "step": 8540
+    },
+    {
+      "epoch": 0.628060886829914,
+      "grad_norm": 0.92208331823349,
+      "learning_rate": 6.421768523181806e-05,
+      "loss": 0.3461,
+      "step": 8541
+    },
+    {
+      "epoch": 0.6281344216486506,
+      "grad_norm": 0.9413928985595703,
+      "learning_rate": 6.419544694854434e-05,
+      "loss": 0.3327,
+      "step": 8542
+    },
+    {
+      "epoch": 0.6282079564673874,
+      "grad_norm": 0.6341896653175354,
+      "learning_rate": 6.417321069614269e-05,
+      "loss": 0.3442,
+      "step": 8543
+    },
+    {
+      "epoch": 0.628281491286124,
+      "grad_norm": 1.3163142204284668,
+      "learning_rate": 6.41509764758743e-05,
+      "loss": 0.3408,
+      "step": 8544
+    },
+    {
+      "epoch": 0.6283550261048606,
+      "grad_norm": 0.4436657428741455,
+      "learning_rate": 6.412874428900038e-05,
+      "loss": 0.3432,
+      "step": 8545
+    },
+    {
+      "epoch": 0.6284285609235973,
+      "grad_norm": 1.0758670568466187,
+      "learning_rate": 6.410651413678192e-05,
+      "loss": 0.3434,
+      "step": 8546
+    },
+    {
+      "epoch": 0.628502095742334,
+      "grad_norm": 0.4372706413269043,
+      "learning_rate": 6.408428602047992e-05,
+      "loss": 0.3523,
+      "step": 8547
+    },
+    {
+      "epoch": 0.6285756305610707,
+      "grad_norm": 1.2102776765823364,
+      "learning_rate": 6.406205994135507e-05,
+      "loss": 0.3483,
+      "step": 8548
+    },
+    {
+      "epoch": 0.6286491653798073,
+      "grad_norm": 0.5321853756904602,
+      "learning_rate": 6.403983590066809e-05,
+      "loss": 0.3748,
+      "step": 8549
+    },
+    {
+      "epoch": 0.628722700198544,
+      "grad_norm": 0.9013388752937317,
+      "learning_rate": 6.401761389967959e-05,
+      "loss": 0.3529,
+      "step": 8550
+    },
+    {
+      "epoch": 0.6287962350172807,
+      "grad_norm": 0.671417236328125,
+      "learning_rate": 6.399539393965001e-05,
+      "loss": 0.329,
+      "step": 8551
+    },
+    {
+      "epoch": 0.6288697698360174,
+      "grad_norm": 0.8203094601631165,
+      "learning_rate": 6.397317602183966e-05,
+      "loss": 0.3399,
+      "step": 8552
+    },
+    {
+      "epoch": 0.628943304654754,
+      "grad_norm": 1.1956051588058472,
+      "learning_rate": 6.39509601475088e-05,
+      "loss": 0.3555,
+      "step": 8553
+    },
+    {
+      "epoch": 0.6290168394734907,
+      "grad_norm": 0.5818515419960022,
+      "learning_rate": 6.392874631791752e-05,
+      "loss": 0.3568,
+      "step": 8554
+    },
+    {
+      "epoch": 0.6290903742922274,
+      "grad_norm": 1.0343828201293945,
+      "learning_rate": 6.390653453432583e-05,
+      "loss": 0.3447,
+      "step": 8555
+    },
+    {
+      "epoch": 0.6291639091109641,
+      "grad_norm": 0.936972439289093,
+      "learning_rate": 6.38843247979936e-05,
+      "loss": 0.3421,
+      "step": 8556
+    },
+    {
+      "epoch": 0.6292374439297007,
+      "grad_norm": 0.5565615296363831,
+      "learning_rate": 6.386211711018054e-05,
+      "loss": 0.3461,
+      "step": 8557
+    },
+    {
+      "epoch": 0.6293109787484373,
+      "grad_norm": 0.9924426674842834,
+      "learning_rate": 6.383991147214636e-05,
+      "loss": 0.3169,
+      "step": 8558
+    },
+    {
+      "epoch": 0.6293845135671741,
+      "grad_norm": 0.44848379492759705,
+      "learning_rate": 6.38177078851506e-05,
+      "loss": 0.3217,
+      "step": 8559
+    },
+    {
+      "epoch": 0.6294580483859107,
+      "grad_norm": 0.706409215927124,
+      "learning_rate": 6.37955063504526e-05,
+      "loss": 0.351,
+      "step": 8560
+    },
+    {
+      "epoch": 0.6295315832046474,
+      "grad_norm": 0.5392260551452637,
+      "learning_rate": 6.377330686931169e-05,
+      "loss": 0.3259,
+      "step": 8561
+    },
+    {
+      "epoch": 0.629605118023384,
+      "grad_norm": 0.5775625705718994,
+      "learning_rate": 6.375110944298703e-05,
+      "loss": 0.3452,
+      "step": 8562
+    },
+    {
+      "epoch": 0.6296786528421208,
+      "grad_norm": 0.5617018938064575,
+      "learning_rate": 6.37289140727377e-05,
+      "loss": 0.3407,
+      "step": 8563
+    },
+    {
+      "epoch": 0.6297521876608574,
+      "grad_norm": 0.4351850748062134,
+      "learning_rate": 6.370672075982266e-05,
+      "loss": 0.3267,
+      "step": 8564
+    },
+    {
+      "epoch": 0.6298257224795941,
+      "grad_norm": 0.4370947778224945,
+      "learning_rate": 6.368452950550071e-05,
+      "loss": 0.3262,
+      "step": 8565
+    },
+    {
+      "epoch": 0.6298992572983307,
+      "grad_norm": 0.4438112676143646,
+      "learning_rate": 6.366234031103058e-05,
+      "loss": 0.3764,
+      "step": 8566
+    },
+    {
+      "epoch": 0.6299727921170675,
+      "grad_norm": 0.4162306487560272,
+      "learning_rate": 6.364015317767083e-05,
+      "loss": 0.3231,
+      "step": 8567
+    },
+    {
+      "epoch": 0.6300463269358041,
+      "grad_norm": 0.4828980267047882,
+      "learning_rate": 6.361796810667998e-05,
+      "loss": 0.3324,
+      "step": 8568
+    },
+    {
+      "epoch": 0.6301198617545408,
+      "grad_norm": 0.6714590787887573,
+      "learning_rate": 6.359578509931639e-05,
+      "loss": 0.3141,
+      "step": 8569
+    },
+    {
+      "epoch": 0.6301933965732774,
+      "grad_norm": 0.7519006729125977,
+      "learning_rate": 6.357360415683827e-05,
+      "loss": 0.3134,
+      "step": 8570
+    },
+    {
+      "epoch": 0.6302669313920142,
+      "grad_norm": 0.4111039638519287,
+      "learning_rate": 6.35514252805038e-05,
+      "loss": 0.3228,
+      "step": 8571
+    },
+    {
+      "epoch": 0.6303404662107508,
+      "grad_norm": 0.9733797907829285,
+      "learning_rate": 6.352924847157091e-05,
+      "loss": 0.3344,
+      "step": 8572
+    },
+    {
+      "epoch": 0.6304140010294874,
+      "grad_norm": 0.769050657749176,
+      "learning_rate": 6.350707373129756e-05,
+      "loss": 0.3362,
+      "step": 8573
+    },
+    {
+      "epoch": 0.6304875358482241,
+      "grad_norm": 0.7271032333374023,
+      "learning_rate": 6.34849010609415e-05,
+      "loss": 0.3814,
+      "step": 8574
+    },
+    {
+      "epoch": 0.6305610706669608,
+      "grad_norm": 0.8449476957321167,
+      "learning_rate": 6.346273046176036e-05,
+      "loss": 0.3359,
+      "step": 8575
+    },
+    {
+      "epoch": 0.6306346054856975,
+      "grad_norm": 0.6161856651306152,
+      "learning_rate": 6.344056193501172e-05,
+      "loss": 0.3369,
+      "step": 8576
+    },
+    {
+      "epoch": 0.6307081403044341,
+      "grad_norm": 0.8400689959526062,
+      "learning_rate": 6.341839548195298e-05,
+      "loss": 0.3328,
+      "step": 8577
+    },
+    {
+      "epoch": 0.6307816751231708,
+      "grad_norm": 0.5615174174308777,
+      "learning_rate": 6.339623110384147e-05,
+      "loss": 0.3298,
+      "step": 8578
+    },
+    {
+      "epoch": 0.6308552099419075,
+      "grad_norm": 0.811187207698822,
+      "learning_rate": 6.337406880193437e-05,
+      "loss": 0.3604,
+      "step": 8579
+    },
+    {
+      "epoch": 0.6309287447606442,
+      "grad_norm": 0.9675296545028687,
+      "learning_rate": 6.335190857748871e-05,
+      "loss": 0.3195,
+      "step": 8580
+    },
+    {
+      "epoch": 0.6310022795793808,
+      "grad_norm": 0.43474259972572327,
+      "learning_rate": 6.332975043176151e-05,
+      "loss": 0.3431,
+      "step": 8581
+    },
+    {
+      "epoch": 0.6310758143981176,
+      "grad_norm": 0.7161589860916138,
+      "learning_rate": 6.330759436600957e-05,
+      "loss": 0.3192,
+      "step": 8582
+    },
+    {
+      "epoch": 0.6311493492168542,
+      "grad_norm": 0.6420085430145264,
+      "learning_rate": 6.328544038148958e-05,
+      "loss": 0.3683,
+      "step": 8583
+    },
+    {
+      "epoch": 0.6312228840355909,
+      "grad_norm": 0.39924490451812744,
+      "learning_rate": 6.326328847945817e-05,
+      "loss": 0.3327,
+      "step": 8584
+    },
+    {
+      "epoch": 0.6312964188543275,
+      "grad_norm": 0.5036346912384033,
+      "learning_rate": 6.324113866117178e-05,
+      "loss": 0.3442,
+      "step": 8585
+    },
+    {
+      "epoch": 0.6313699536730641,
+      "grad_norm": 0.45144474506378174,
+      "learning_rate": 6.321899092788681e-05,
+      "loss": 0.3491,
+      "step": 8586
+    },
+    {
+      "epoch": 0.6314434884918009,
+      "grad_norm": 0.4132676422595978,
+      "learning_rate": 6.31968452808595e-05,
+      "loss": 0.327,
+      "step": 8587
+    },
+    {
+      "epoch": 0.6315170233105375,
+      "grad_norm": 0.47793036699295044,
+      "learning_rate": 6.317470172134594e-05,
+      "loss": 0.3398,
+      "step": 8588
+    },
+    {
+      "epoch": 0.6315905581292742,
+      "grad_norm": 0.6159135699272156,
+      "learning_rate": 6.315256025060218e-05,
+      "loss": 0.3367,
+      "step": 8589
+    },
+    {
+      "epoch": 0.6316640929480108,
+      "grad_norm": 0.46380725502967834,
+      "learning_rate": 6.313042086988408e-05,
+      "loss": 0.3318,
+      "step": 8590
+    },
+    {
+      "epoch": 0.6317376277667476,
+      "grad_norm": 0.42916935682296753,
+      "learning_rate": 6.310828358044745e-05,
+      "loss": 0.3472,
+      "step": 8591
+    },
+    {
+      "epoch": 0.6318111625854842,
+      "grad_norm": 0.39487260580062866,
+      "learning_rate": 6.308614838354788e-05,
+      "loss": 0.3147,
+      "step": 8592
+    },
+    {
+      "epoch": 0.6318846974042209,
+      "grad_norm": 0.4897730052471161,
+      "learning_rate": 6.306401528044093e-05,
+      "loss": 0.3166,
+      "step": 8593
+    },
+    {
+      "epoch": 0.6319582322229575,
+      "grad_norm": 0.5327439904212952,
+      "learning_rate": 6.304188427238207e-05,
+      "loss": 0.3607,
+      "step": 8594
+    },
+    {
+      "epoch": 0.6320317670416943,
+      "grad_norm": 0.5348936319351196,
+      "learning_rate": 6.301975536062647e-05,
+      "loss": 0.3487,
+      "step": 8595
+    },
+    {
+      "epoch": 0.6321053018604309,
+      "grad_norm": 0.9512827396392822,
+      "learning_rate": 6.299762854642939e-05,
+      "loss": 0.3625,
+      "step": 8596
+    },
+    {
+      "epoch": 0.6321788366791676,
+      "grad_norm": 0.6593134999275208,
+      "learning_rate": 6.297550383104588e-05,
+      "loss": 0.3237,
+      "step": 8597
+    },
+    {
+      "epoch": 0.6322523714979043,
+      "grad_norm": 0.6438331007957458,
+      "learning_rate": 6.295338121573085e-05,
+      "loss": 0.3772,
+      "step": 8598
+    },
+    {
+      "epoch": 0.632325906316641,
+      "grad_norm": 0.8033629655838013,
+      "learning_rate": 6.293126070173914e-05,
+      "loss": 0.3315,
+      "step": 8599
+    },
+    {
+      "epoch": 0.6323994411353776,
+      "grad_norm": 0.5071994662284851,
+      "learning_rate": 6.290914229032543e-05,
+      "loss": 0.3018,
+      "step": 8600
+    },
+    {
+      "epoch": 0.6324729759541142,
+      "grad_norm": 0.7013834714889526,
+      "learning_rate": 6.288702598274431e-05,
+      "loss": 0.3334,
+      "step": 8601
+    },
+    {
+      "epoch": 0.632546510772851,
+      "grad_norm": 0.9141992926597595,
+      "learning_rate": 6.286491178025027e-05,
+      "loss": 0.3489,
+      "step": 8602
+    },
+    {
+      "epoch": 0.6326200455915876,
+      "grad_norm": 0.482257217168808,
+      "learning_rate": 6.28427996840976e-05,
+      "loss": 0.3313,
+      "step": 8603
+    },
+    {
+      "epoch": 0.6326935804103243,
+      "grad_norm": 0.6825869083404541,
+      "learning_rate": 6.282068969554057e-05,
+      "loss": 0.3175,
+      "step": 8604
+    },
+    {
+      "epoch": 0.6327671152290609,
+      "grad_norm": 0.5404843688011169,
+      "learning_rate": 6.279858181583324e-05,
+      "loss": 0.3516,
+      "step": 8605
+    },
+    {
+      "epoch": 0.6328406500477977,
+      "grad_norm": 0.6534824967384338,
+      "learning_rate": 6.277647604622965e-05,
+      "loss": 0.3431,
+      "step": 8606
+    },
+    {
+      "epoch": 0.6329141848665343,
+      "grad_norm": 0.7009466886520386,
+      "learning_rate": 6.275437238798361e-05,
+      "loss": 0.3522,
+      "step": 8607
+    },
+    {
+      "epoch": 0.632987719685271,
+      "grad_norm": 0.7398335337638855,
+      "learning_rate": 6.273227084234883e-05,
+      "loss": 0.3557,
+      "step": 8608
+    },
+    {
+      "epoch": 0.6330612545040076,
+      "grad_norm": 1.2750904560089111,
+      "learning_rate": 6.2710171410579e-05,
+      "loss": 0.3479,
+      "step": 8609
+    },
+    {
+      "epoch": 0.6331347893227444,
+      "grad_norm": 0.7455692291259766,
+      "learning_rate": 6.26880740939276e-05,
+      "loss": 0.3517,
+      "step": 8610
+    },
+    {
+      "epoch": 0.633208324141481,
+      "grad_norm": 1.485133171081543,
+      "learning_rate": 6.266597889364801e-05,
+      "loss": 0.3526,
+      "step": 8611
+    },
+    {
+      "epoch": 0.6332818589602177,
+      "grad_norm": 0.763263463973999,
+      "learning_rate": 6.264388581099352e-05,
+      "loss": 0.3386,
+      "step": 8612
+    },
+    {
+      "epoch": 0.6333553937789543,
+      "grad_norm": 1.352057933807373,
+      "learning_rate": 6.26217948472172e-05,
+      "loss": 0.3444,
+      "step": 8613
+    },
+    {
+      "epoch": 0.633428928597691,
+      "grad_norm": 0.8768470287322998,
+      "learning_rate": 6.259970600357216e-05,
+      "loss": 0.332,
+      "step": 8614
+    },
+    {
+      "epoch": 0.6335024634164277,
+      "grad_norm": 1.775553822517395,
+      "learning_rate": 6.257761928131128e-05,
+      "loss": 0.3318,
+      "step": 8615
+    },
+    {
+      "epoch": 0.6335759982351643,
+      "grad_norm": 1.7122787237167358,
+      "learning_rate": 6.255553468168729e-05,
+      "loss": 0.3551,
+      "step": 8616
+    },
+    {
+      "epoch": 0.633649533053901,
+      "grad_norm": 0.7942845821380615,
+      "learning_rate": 6.253345220595297e-05,
+      "loss": 0.327,
+      "step": 8617
+    },
+    {
+      "epoch": 0.6337230678726377,
+      "grad_norm": 1.4313294887542725,
+      "learning_rate": 6.25113718553607e-05,
+      "loss": 0.337,
+      "step": 8618
+    },
+    {
+      "epoch": 0.6337966026913744,
+      "grad_norm": 0.6885858774185181,
+      "learning_rate": 6.248929363116301e-05,
+      "loss": 0.3387,
+      "step": 8619
+    },
+    {
+      "epoch": 0.633870137510111,
+      "grad_norm": 0.7664195895195007,
+      "learning_rate": 6.246721753461216e-05,
+      "loss": 0.3321,
+      "step": 8620
+    },
+    {
+      "epoch": 0.6339436723288477,
+      "grad_norm": 0.7564030289649963,
+      "learning_rate": 6.244514356696034e-05,
+      "loss": 0.3731,
+      "step": 8621
+    },
+    {
+      "epoch": 0.6340172071475844,
+      "grad_norm": 0.555818498134613,
+      "learning_rate": 6.242307172945962e-05,
+      "loss": 0.3327,
+      "step": 8622
+    },
+    {
+      "epoch": 0.6340907419663211,
+      "grad_norm": 1.149488091468811,
+      "learning_rate": 6.240100202336191e-05,
+      "loss": 0.359,
+      "step": 8623
+    },
+    {
+      "epoch": 0.6341642767850577,
+      "grad_norm": 0.4592645764350891,
+      "learning_rate": 6.237893444991907e-05,
+      "loss": 0.327,
+      "step": 8624
+    },
+    {
+      "epoch": 0.6342378116037944,
+      "grad_norm": 0.7232848405838013,
+      "learning_rate": 6.235686901038276e-05,
+      "loss": 0.3491,
+      "step": 8625
+    },
+    {
+      "epoch": 0.6343113464225311,
+      "grad_norm": 0.48014840483665466,
+      "learning_rate": 6.233480570600455e-05,
+      "loss": 0.3371,
+      "step": 8626
+    },
+    {
+      "epoch": 0.6343848812412678,
+      "grad_norm": 0.5456093549728394,
+      "learning_rate": 6.231274453803593e-05,
+      "loss": 0.3203,
+      "step": 8627
+    },
+    {
+      "epoch": 0.6344584160600044,
+      "grad_norm": 0.49424153566360474,
+      "learning_rate": 6.22906855077282e-05,
+      "loss": 0.3473,
+      "step": 8628
+    },
+    {
+      "epoch": 0.634531950878741,
+      "grad_norm": 0.6821517944335938,
+      "learning_rate": 6.226862861633263e-05,
+      "loss": 0.3585,
+      "step": 8629
+    },
+    {
+      "epoch": 0.6346054856974778,
+      "grad_norm": 0.4902707040309906,
+      "learning_rate": 6.224657386510024e-05,
+      "loss": 0.3246,
+      "step": 8630
+    },
+    {
+      "epoch": 0.6346790205162144,
+      "grad_norm": 0.5775149464607239,
+      "learning_rate": 6.2224521255282e-05,
+      "loss": 0.3525,
+      "step": 8631
+    },
+    {
+      "epoch": 0.6347525553349511,
+      "grad_norm": 0.44037798047065735,
+      "learning_rate": 6.22024707881288e-05,
+      "loss": 0.3476,
+      "step": 8632
+    },
+    {
+      "epoch": 0.6348260901536877,
+      "grad_norm": 0.4631916284561157,
+      "learning_rate": 6.218042246489133e-05,
+      "loss": 0.3305,
+      "step": 8633
+    },
+    {
+      "epoch": 0.6348996249724245,
+      "grad_norm": 0.4392843544483185,
+      "learning_rate": 6.21583762868202e-05,
+      "loss": 0.3415,
+      "step": 8634
+    },
+    {
+      "epoch": 0.6349731597911611,
+      "grad_norm": 0.46527329087257385,
+      "learning_rate": 6.213633225516593e-05,
+      "loss": 0.344,
+      "step": 8635
+    },
+    {
+      "epoch": 0.6350466946098978,
+      "grad_norm": 0.572043776512146,
+      "learning_rate": 6.211429037117882e-05,
+      "loss": 0.3679,
+      "step": 8636
+    },
+    {
+      "epoch": 0.6351202294286344,
+      "grad_norm": 0.5587449073791504,
+      "learning_rate": 6.209225063610919e-05,
+      "loss": 0.3274,
+      "step": 8637
+    },
+    {
+      "epoch": 0.6351937642473712,
+      "grad_norm": 0.47344931960105896,
+      "learning_rate": 6.207021305120708e-05,
+      "loss": 0.3368,
+      "step": 8638
+    },
+    {
+      "epoch": 0.6352672990661078,
+      "grad_norm": 0.4687710106372833,
+      "learning_rate": 6.20481776177225e-05,
+      "loss": 0.3203,
+      "step": 8639
+    },
+    {
+      "epoch": 0.6353408338848445,
+      "grad_norm": 0.4902970492839813,
+      "learning_rate": 6.202614433690536e-05,
+      "loss": 0.3448,
+      "step": 8640
+    },
+    {
+      "epoch": 0.6354143687035811,
+      "grad_norm": 0.5528318285942078,
+      "learning_rate": 6.200411321000543e-05,
+      "loss": 0.3548,
+      "step": 8641
+    },
+    {
+      "epoch": 0.6354879035223178,
+      "grad_norm": 0.4647836685180664,
+      "learning_rate": 6.198208423827226e-05,
+      "loss": 0.3325,
+      "step": 8642
+    },
+    {
+      "epoch": 0.6355614383410545,
+      "grad_norm": 0.5052602291107178,
+      "learning_rate": 6.196005742295538e-05,
+      "loss": 0.3463,
+      "step": 8643
+    },
+    {
+      "epoch": 0.6356349731597911,
+      "grad_norm": 0.5540039539337158,
+      "learning_rate": 6.193803276530418e-05,
+      "loss": 0.3619,
+      "step": 8644
+    },
+    {
+      "epoch": 0.6357085079785278,
+      "grad_norm": 0.7421672344207764,
+      "learning_rate": 6.191601026656797e-05,
+      "loss": 0.3297,
+      "step": 8645
+    },
+    {
+      "epoch": 0.6357820427972645,
+      "grad_norm": 0.4412699341773987,
+      "learning_rate": 6.189398992799582e-05,
+      "loss": 0.3587,
+      "step": 8646
+    },
+    {
+      "epoch": 0.6358555776160012,
+      "grad_norm": 0.7814032435417175,
+      "learning_rate": 6.187197175083677e-05,
+      "loss": 0.3422,
+      "step": 8647
+    },
+    {
+      "epoch": 0.6359291124347378,
+      "grad_norm": 0.7962145805358887,
+      "learning_rate": 6.184995573633974e-05,
+      "loss": 0.3154,
+      "step": 8648
+    },
+    {
+      "epoch": 0.6360026472534746,
+      "grad_norm": 0.43708091974258423,
+      "learning_rate": 6.182794188575346e-05,
+      "loss": 0.3481,
+      "step": 8649
+    },
+    {
+      "epoch": 0.6360761820722112,
+      "grad_norm": 0.9538200497627258,
+      "learning_rate": 6.180593020032662e-05,
+      "loss": 0.345,
+      "step": 8650
+    },
+    {
+      "epoch": 0.6361497168909479,
+      "grad_norm": 0.9069247245788574,
+      "learning_rate": 6.178392068130772e-05,
+      "loss": 0.3337,
+      "step": 8651
+    },
+    {
+      "epoch": 0.6362232517096845,
+      "grad_norm": 0.452641099691391,
+      "learning_rate": 6.176191332994521e-05,
+      "loss": 0.3148,
+      "step": 8652
+    },
+    {
+      "epoch": 0.6362967865284213,
+      "grad_norm": 0.7601242065429688,
+      "learning_rate": 6.173990814748729e-05,
+      "loss": 0.3516,
+      "step": 8653
+    },
+    {
+      "epoch": 0.6363703213471579,
+      "grad_norm": 0.4428872764110565,
+      "learning_rate": 6.171790513518214e-05,
+      "loss": 0.3616,
+      "step": 8654
+    },
+    {
+      "epoch": 0.6364438561658946,
+      "grad_norm": 0.5526891946792603,
+      "learning_rate": 6.169590429427784e-05,
+      "loss": 0.328,
+      "step": 8655
+    },
+    {
+      "epoch": 0.6365173909846312,
+      "grad_norm": 0.6776340007781982,
+      "learning_rate": 6.167390562602227e-05,
+      "loss": 0.336,
+      "step": 8656
+    },
+    {
+      "epoch": 0.6365909258033678,
+      "grad_norm": 0.7375097274780273,
+      "learning_rate": 6.16519091316632e-05,
+      "loss": 0.3141,
+      "step": 8657
+    },
+    {
+      "epoch": 0.6366644606221046,
+      "grad_norm": 0.5597546696662903,
+      "learning_rate": 6.162991481244833e-05,
+      "loss": 0.3252,
+      "step": 8658
+    },
+    {
+      "epoch": 0.6367379954408412,
+      "grad_norm": 0.9202846884727478,
+      "learning_rate": 6.160792266962517e-05,
+      "loss": 0.3434,
+      "step": 8659
+    },
+    {
+      "epoch": 0.6368115302595779,
+      "grad_norm": 0.4904326796531677,
+      "learning_rate": 6.158593270444117e-05,
+      "loss": 0.3167,
+      "step": 8660
+    },
+    {
+      "epoch": 0.6368850650783145,
+      "grad_norm": 0.6774338483810425,
+      "learning_rate": 6.156394491814362e-05,
+      "loss": 0.3085,
+      "step": 8661
+    },
+    {
+      "epoch": 0.6369585998970513,
+      "grad_norm": 0.8612265586853027,
+      "learning_rate": 6.154195931197964e-05,
+      "loss": 0.3477,
+      "step": 8662
+    },
+    {
+      "epoch": 0.6370321347157879,
+      "grad_norm": 0.42301061749458313,
+      "learning_rate": 6.151997588719633e-05,
+      "loss": 0.3243,
+      "step": 8663
+    },
+    {
+      "epoch": 0.6371056695345246,
+      "grad_norm": 0.7362720370292664,
+      "learning_rate": 6.149799464504063e-05,
+      "loss": 0.3282,
+      "step": 8664
+    },
+    {
+      "epoch": 0.6371792043532613,
+      "grad_norm": 0.41994529962539673,
+      "learning_rate": 6.147601558675928e-05,
+      "loss": 0.3284,
+      "step": 8665
+    },
+    {
+      "epoch": 0.637252739171998,
+      "grad_norm": 0.8701484799385071,
+      "learning_rate": 6.145403871359899e-05,
+      "loss": 0.3715,
+      "step": 8666
+    },
+    {
+      "epoch": 0.6373262739907346,
+      "grad_norm": 0.5445263385772705,
+      "learning_rate": 6.143206402680628e-05,
+      "loss": 0.3387,
+      "step": 8667
+    },
+    {
+      "epoch": 0.6373998088094713,
+      "grad_norm": 0.6568160653114319,
+      "learning_rate": 6.141009152762762e-05,
+      "loss": 0.3632,
+      "step": 8668
+    },
+    {
+      "epoch": 0.637473343628208,
+      "grad_norm": 0.6711809039115906,
+      "learning_rate": 6.13881212173093e-05,
+      "loss": 0.3303,
+      "step": 8669
+    },
+    {
+      "epoch": 0.6375468784469446,
+      "grad_norm": 0.5065097212791443,
+      "learning_rate": 6.136615309709747e-05,
+      "loss": 0.3311,
+      "step": 8670
+    },
+    {
+      "epoch": 0.6376204132656813,
+      "grad_norm": 0.6537981033325195,
+      "learning_rate": 6.134418716823823e-05,
+      "loss": 0.3647,
+      "step": 8671
+    },
+    {
+      "epoch": 0.6376939480844179,
+      "grad_norm": 0.444328248500824,
+      "learning_rate": 6.132222343197745e-05,
+      "loss": 0.3362,
+      "step": 8672
+    },
+    {
+      "epoch": 0.6377674829031547,
+      "grad_norm": 0.5979285836219788,
+      "learning_rate": 6.130026188956102e-05,
+      "loss": 0.3239,
+      "step": 8673
+    },
+    {
+      "epoch": 0.6378410177218913,
+      "grad_norm": 0.4292072653770447,
+      "learning_rate": 6.127830254223457e-05,
+      "loss": 0.3393,
+      "step": 8674
+    },
+    {
+      "epoch": 0.637914552540628,
+      "grad_norm": 0.6690128445625305,
+      "learning_rate": 6.125634539124363e-05,
+      "loss": 0.3268,
+      "step": 8675
+    },
+    {
+      "epoch": 0.6379880873593646,
+      "grad_norm": 0.4534841477870941,
+      "learning_rate": 6.123439043783375e-05,
+      "loss": 0.3728,
+      "step": 8676
+    },
+    {
+      "epoch": 0.6380616221781014,
+      "grad_norm": 0.9023979902267456,
+      "learning_rate": 6.121243768325007e-05,
+      "loss": 0.3454,
+      "step": 8677
+    },
+    {
+      "epoch": 0.638135156996838,
+      "grad_norm": 0.7521181106567383,
+      "learning_rate": 6.119048712873789e-05,
+      "loss": 0.3474,
+      "step": 8678
+    },
+    {
+      "epoch": 0.6382086918155747,
+      "grad_norm": 0.6248573660850525,
+      "learning_rate": 6.116853877554225e-05,
+      "loss": 0.343,
+      "step": 8679
+    },
+    {
+      "epoch": 0.6382822266343113,
+      "grad_norm": 1.0840555429458618,
+      "learning_rate": 6.114659262490804e-05,
+      "loss": 0.3322,
+      "step": 8680
+    },
+    {
+      "epoch": 0.6383557614530481,
+      "grad_norm": 0.8282070755958557,
+      "learning_rate": 6.112464867808012e-05,
+      "loss": 0.3334,
+      "step": 8681
+    },
+    {
+      "epoch": 0.6384292962717847,
+      "grad_norm": 0.694075345993042,
+      "learning_rate": 6.110270693630313e-05,
+      "loss": 0.3309,
+      "step": 8682
+    },
+    {
+      "epoch": 0.6385028310905214,
+      "grad_norm": 1.1778515577316284,
+      "learning_rate": 6.108076740082169e-05,
+      "loss": 0.3217,
+      "step": 8683
+    },
+    {
+      "epoch": 0.638576365909258,
+      "grad_norm": 0.627206027507782,
+      "learning_rate": 6.105883007288018e-05,
+      "loss": 0.341,
+      "step": 8684
+    },
+    {
+      "epoch": 0.6386499007279947,
+      "grad_norm": 1.3259307146072388,
+      "learning_rate": 6.10368949537229e-05,
+      "loss": 0.3463,
+      "step": 8685
+    },
+    {
+      "epoch": 0.6387234355467314,
+      "grad_norm": 0.6099632978439331,
+      "learning_rate": 6.101496204459409e-05,
+      "loss": 0.3551,
+      "step": 8686
+    },
+    {
+      "epoch": 0.638796970365468,
+      "grad_norm": 1.4396555423736572,
+      "learning_rate": 6.09930313467378e-05,
+      "loss": 0.3622,
+      "step": 8687
+    },
+    {
+      "epoch": 0.6388705051842047,
+      "grad_norm": 0.6882794499397278,
+      "learning_rate": 6.09711028613979e-05,
+      "loss": 0.3508,
+      "step": 8688
+    },
+    {
+      "epoch": 0.6389440400029414,
+      "grad_norm": 1.0993090867996216,
+      "learning_rate": 6.0949176589818244e-05,
+      "loss": 0.3362,
+      "step": 8689
+    },
+    {
+      "epoch": 0.6390175748216781,
+      "grad_norm": 0.5981147885322571,
+      "learning_rate": 6.0927252533242474e-05,
+      "loss": 0.3358,
+      "step": 8690
+    },
+    {
+      "epoch": 0.6390911096404147,
+      "grad_norm": 0.8727046251296997,
+      "learning_rate": 6.090533069291421e-05,
+      "loss": 0.354,
+      "step": 8691
+    },
+    {
+      "epoch": 0.6391646444591514,
+      "grad_norm": 0.5073428153991699,
+      "learning_rate": 6.088341107007685e-05,
+      "loss": 0.3327,
+      "step": 8692
+    },
+    {
+      "epoch": 0.6392381792778881,
+      "grad_norm": 0.6651918888092041,
+      "learning_rate": 6.086149366597366e-05,
+      "loss": 0.3527,
+      "step": 8693
+    },
+    {
+      "epoch": 0.6393117140966248,
+      "grad_norm": 0.6110631823539734,
+      "learning_rate": 6.083957848184788e-05,
+      "loss": 0.3504,
+      "step": 8694
+    },
+    {
+      "epoch": 0.6393852489153614,
+      "grad_norm": 0.4717518985271454,
+      "learning_rate": 6.081766551894251e-05,
+      "loss": 0.3188,
+      "step": 8695
+    },
+    {
+      "epoch": 0.6394587837340981,
+      "grad_norm": 0.6424765586853027,
+      "learning_rate": 6.0795754778500534e-05,
+      "loss": 0.3528,
+      "step": 8696
+    },
+    {
+      "epoch": 0.6395323185528348,
+      "grad_norm": 0.48179763555526733,
+      "learning_rate": 6.077384626176472e-05,
+      "loss": 0.3626,
+      "step": 8697
+    },
+    {
+      "epoch": 0.6396058533715714,
+      "grad_norm": 0.4457365572452545,
+      "learning_rate": 6.075193996997772e-05,
+      "loss": 0.3318,
+      "step": 8698
+    },
+    {
+      "epoch": 0.6396793881903081,
+      "grad_norm": 0.4241383671760559,
+      "learning_rate": 6.073003590438215e-05,
+      "loss": 0.3272,
+      "step": 8699
+    },
+    {
+      "epoch": 0.6397529230090447,
+      "grad_norm": 0.48266321420669556,
+      "learning_rate": 6.0708134066220353e-05,
+      "loss": 0.3286,
+      "step": 8700
+    },
+    {
+      "epoch": 0.6398264578277815,
+      "grad_norm": 0.521031379699707,
+      "learning_rate": 6.068623445673467e-05,
+      "loss": 0.3249,
+      "step": 8701
+    },
+    {
+      "epoch": 0.6398999926465181,
+      "grad_norm": 0.5186377763748169,
+      "learning_rate": 6.066433707716725e-05,
+      "loss": 0.3275,
+      "step": 8702
+    },
+    {
+      "epoch": 0.6399735274652548,
+      "grad_norm": 0.4693365693092346,
+      "learning_rate": 6.064244192876013e-05,
+      "loss": 0.3313,
+      "step": 8703
+    },
+    {
+      "epoch": 0.6400470622839914,
+      "grad_norm": 0.5220229625701904,
+      "learning_rate": 6.062054901275526e-05,
+      "loss": 0.3125,
+      "step": 8704
+    },
+    {
+      "epoch": 0.6401205971027282,
+      "grad_norm": 0.5214353799819946,
+      "learning_rate": 6.0598658330394374e-05,
+      "loss": 0.3446,
+      "step": 8705
+    },
+    {
+      "epoch": 0.6401941319214648,
+      "grad_norm": 0.6873384714126587,
+      "learning_rate": 6.057676988291921e-05,
+      "loss": 0.3668,
+      "step": 8706
+    },
+    {
+      "epoch": 0.6402676667402015,
+      "grad_norm": 0.4710334241390228,
+      "learning_rate": 6.055488367157125e-05,
+      "loss": 0.3216,
+      "step": 8707
+    },
+    {
+      "epoch": 0.6403412015589381,
+      "grad_norm": 0.6445189714431763,
+      "learning_rate": 6.053299969759191e-05,
+      "loss": 0.3614,
+      "step": 8708
+    },
+    {
+      "epoch": 0.6404147363776749,
+      "grad_norm": 0.5251979231834412,
+      "learning_rate": 6.0511117962222486e-05,
+      "loss": 0.3494,
+      "step": 8709
+    },
+    {
+      "epoch": 0.6404882711964115,
+      "grad_norm": 0.482692152261734,
+      "learning_rate": 6.0489238466704137e-05,
+      "loss": 0.3432,
+      "step": 8710
+    },
+    {
+      "epoch": 0.6405618060151482,
+      "grad_norm": 0.45067885518074036,
+      "learning_rate": 6.0467361212277885e-05,
+      "loss": 0.3365,
+      "step": 8711
+    },
+    {
+      "epoch": 0.6406353408338848,
+      "grad_norm": 0.5094512104988098,
+      "learning_rate": 6.0445486200184607e-05,
+      "loss": 0.3488,
+      "step": 8712
+    },
+    {
+      "epoch": 0.6407088756526215,
+      "grad_norm": 0.5219285488128662,
+      "learning_rate": 6.0423613431665084e-05,
+      "loss": 0.3286,
+      "step": 8713
+    },
+    {
+      "epoch": 0.6407824104713582,
+      "grad_norm": 0.46207770705223083,
+      "learning_rate": 6.040174290795999e-05,
+      "loss": 0.3424,
+      "step": 8714
+    },
+    {
+      "epoch": 0.6408559452900948,
+      "grad_norm": 0.4554549753665924,
+      "learning_rate": 6.037987463030983e-05,
+      "loss": 0.3518,
+      "step": 8715
+    },
+    {
+      "epoch": 0.6409294801088316,
+      "grad_norm": 0.6285716891288757,
+      "learning_rate": 6.035800859995498e-05,
+      "loss": 0.3598,
+      "step": 8716
+    },
+    {
+      "epoch": 0.6410030149275682,
+      "grad_norm": 0.47043052315711975,
+      "learning_rate": 6.033614481813572e-05,
+      "loss": 0.3558,
+      "step": 8717
+    },
+    {
+      "epoch": 0.6410765497463049,
+      "grad_norm": 0.7113535404205322,
+      "learning_rate": 6.031428328609217e-05,
+      "loss": 0.3278,
+      "step": 8718
+    },
+    {
+      "epoch": 0.6411500845650415,
+      "grad_norm": 0.4149990379810333,
+      "learning_rate": 6.029242400506439e-05,
+      "loss": 0.3562,
+      "step": 8719
+    },
+    {
+      "epoch": 0.6412236193837783,
+      "grad_norm": 0.9634997248649597,
+      "learning_rate": 6.027056697629221e-05,
+      "loss": 0.3351,
+      "step": 8720
+    },
+    {
+      "epoch": 0.6412971542025149,
+      "grad_norm": 0.8941571712493896,
+      "learning_rate": 6.0248712201015376e-05,
+      "loss": 0.3224,
+      "step": 8721
+    },
+    {
+      "epoch": 0.6413706890212516,
+      "grad_norm": 0.49902328848838806,
+      "learning_rate": 6.0226859680473566e-05,
+      "loss": 0.3444,
+      "step": 8722
+    },
+    {
+      "epoch": 0.6414442238399882,
+      "grad_norm": 0.807862401008606,
+      "learning_rate": 6.0205009415906264e-05,
+      "loss": 0.3271,
+      "step": 8723
+    },
+    {
+      "epoch": 0.641517758658725,
+      "grad_norm": 0.68709397315979,
+      "learning_rate": 6.018316140855281e-05,
+      "loss": 0.3424,
+      "step": 8724
+    },
+    {
+      "epoch": 0.6415912934774616,
+      "grad_norm": 0.5026355385780334,
+      "learning_rate": 6.0161315659652465e-05,
+      "loss": 0.3562,
+      "step": 8725
+    },
+    {
+      "epoch": 0.6416648282961982,
+      "grad_norm": 0.6886836290359497,
+      "learning_rate": 6.013947217044431e-05,
+      "loss": 0.3335,
+      "step": 8726
+    },
+    {
+      "epoch": 0.6417383631149349,
+      "grad_norm": 0.46080562472343445,
+      "learning_rate": 6.011763094216738e-05,
+      "loss": 0.3335,
+      "step": 8727
+    },
+    {
+      "epoch": 0.6418118979336715,
+      "grad_norm": 0.6029114723205566,
+      "learning_rate": 6.00957919760605e-05,
+      "loss": 0.3506,
+      "step": 8728
+    },
+    {
+      "epoch": 0.6418854327524083,
+      "grad_norm": 0.4346432685852051,
+      "learning_rate": 6.007395527336244e-05,
+      "loss": 0.348,
+      "step": 8729
+    },
+    {
+      "epoch": 0.6419589675711449,
+      "grad_norm": 0.4767722189426422,
+      "learning_rate": 6.0052120835311755e-05,
+      "loss": 0.3245,
+      "step": 8730
+    },
+    {
+      "epoch": 0.6420325023898816,
+      "grad_norm": 0.4134998917579651,
+      "learning_rate": 6.003028866314691e-05,
+      "loss": 0.3353,
+      "step": 8731
+    },
+    {
+      "epoch": 0.6421060372086183,
+      "grad_norm": 0.5121638774871826,
+      "learning_rate": 6.000845875810631e-05,
+      "loss": 0.3192,
+      "step": 8732
+    },
+    {
+      "epoch": 0.642179572027355,
+      "grad_norm": 0.46951714158058167,
+      "learning_rate": 5.998663112142813e-05,
+      "loss": 0.3551,
+      "step": 8733
+    },
+    {
+      "epoch": 0.6422531068460916,
+      "grad_norm": 0.4016415774822235,
+      "learning_rate": 5.9964805754350486e-05,
+      "loss": 0.3469,
+      "step": 8734
+    },
+    {
+      "epoch": 0.6423266416648283,
+      "grad_norm": 0.48030996322631836,
+      "learning_rate": 5.994298265811128e-05,
+      "loss": 0.3416,
+      "step": 8735
+    },
+    {
+      "epoch": 0.642400176483565,
+      "grad_norm": 0.44147658348083496,
+      "learning_rate": 5.992116183394836e-05,
+      "loss": 0.3842,
+      "step": 8736
+    },
+    {
+      "epoch": 0.6424737113023017,
+      "grad_norm": 0.4404730498790741,
+      "learning_rate": 5.989934328309945e-05,
+      "loss": 0.346,
+      "step": 8737
+    },
+    {
+      "epoch": 0.6425472461210383,
+      "grad_norm": 0.430340051651001,
+      "learning_rate": 5.987752700680211e-05,
+      "loss": 0.3327,
+      "step": 8738
+    },
+    {
+      "epoch": 0.642620780939775,
+      "grad_norm": 0.5162236094474792,
+      "learning_rate": 5.985571300629376e-05,
+      "loss": 0.3317,
+      "step": 8739
+    },
+    {
+      "epoch": 0.6426943157585117,
+      "grad_norm": 0.4910590648651123,
+      "learning_rate": 5.9833901282811764e-05,
+      "loss": 0.338,
+      "step": 8740
+    },
+    {
+      "epoch": 0.6427678505772483,
+      "grad_norm": 0.4738474190235138,
+      "learning_rate": 5.9812091837593244e-05,
+      "loss": 0.3752,
+      "step": 8741
+    },
+    {
+      "epoch": 0.642841385395985,
+      "grad_norm": 0.43797174096107483,
+      "learning_rate": 5.979028467187531e-05,
+      "loss": 0.3389,
+      "step": 8742
+    },
+    {
+      "epoch": 0.6429149202147216,
+      "grad_norm": 0.4200631380081177,
+      "learning_rate": 5.976847978689487e-05,
+      "loss": 0.3283,
+      "step": 8743
+    },
+    {
+      "epoch": 0.6429884550334584,
+      "grad_norm": 0.5424063205718994,
+      "learning_rate": 5.974667718388869e-05,
+      "loss": 0.3432,
+      "step": 8744
+    },
+    {
+      "epoch": 0.643061989852195,
+      "grad_norm": 0.4350639581680298,
+      "learning_rate": 5.972487686409348e-05,
+      "loss": 0.3192,
+      "step": 8745
+    },
+    {
+      "epoch": 0.6431355246709317,
+      "grad_norm": 0.5608150362968445,
+      "learning_rate": 5.9703078828745785e-05,
+      "loss": 0.3276,
+      "step": 8746
+    },
+    {
+      "epoch": 0.6432090594896683,
+      "grad_norm": 0.4876312017440796,
+      "learning_rate": 5.9681283079081965e-05,
+      "loss": 0.3594,
+      "step": 8747
+    },
+    {
+      "epoch": 0.6432825943084051,
+      "grad_norm": 0.5564594864845276,
+      "learning_rate": 5.965948961633832e-05,
+      "loss": 0.3442,
+      "step": 8748
+    },
+    {
+      "epoch": 0.6433561291271417,
+      "grad_norm": 0.5048115253448486,
+      "learning_rate": 5.963769844175095e-05,
+      "loss": 0.3297,
+      "step": 8749
+    },
+    {
+      "epoch": 0.6434296639458784,
+      "grad_norm": 0.825630784034729,
+      "learning_rate": 5.9615909556555986e-05,
+      "loss": 0.3291,
+      "step": 8750
+    },
+    {
+      "epoch": 0.643503198764615,
+      "grad_norm": 0.6693244576454163,
+      "learning_rate": 5.9594122961989195e-05,
+      "loss": 0.3384,
+      "step": 8751
+    },
+    {
+      "epoch": 0.6435767335833518,
+      "grad_norm": 0.6084843277931213,
+      "learning_rate": 5.957233865928642e-05,
+      "loss": 0.3415,
+      "step": 8752
+    },
+    {
+      "epoch": 0.6436502684020884,
+      "grad_norm": 0.653131365776062,
+      "learning_rate": 5.9550556649683275e-05,
+      "loss": 0.3453,
+      "step": 8753
+    },
+    {
+      "epoch": 0.643723803220825,
+      "grad_norm": 0.6298924684524536,
+      "learning_rate": 5.952877693441521e-05,
+      "loss": 0.3604,
+      "step": 8754
+    },
+    {
+      "epoch": 0.6437973380395617,
+      "grad_norm": 0.7335695624351501,
+      "learning_rate": 5.950699951471765e-05,
+      "loss": 0.3502,
+      "step": 8755
+    },
+    {
+      "epoch": 0.6438708728582984,
+      "grad_norm": 0.848233163356781,
+      "learning_rate": 5.9485224391825823e-05,
+      "loss": 0.3554,
+      "step": 8756
+    },
+    {
+      "epoch": 0.6439444076770351,
+      "grad_norm": 0.7591545581817627,
+      "learning_rate": 5.9463451566974795e-05,
+      "loss": 0.3531,
+      "step": 8757
+    },
+    {
+      "epoch": 0.6440179424957717,
+      "grad_norm": 0.5447898507118225,
+      "learning_rate": 5.944168104139964e-05,
+      "loss": 0.3306,
+      "step": 8758
+    },
+    {
+      "epoch": 0.6440914773145084,
+      "grad_norm": 0.8188866376876831,
+      "learning_rate": 5.941991281633508e-05,
+      "loss": 0.3343,
+      "step": 8759
+    },
+    {
+      "epoch": 0.6441650121332451,
+      "grad_norm": 0.5622565746307373,
+      "learning_rate": 5.93981468930159e-05,
+      "loss": 0.3316,
+      "step": 8760
+    },
+    {
+      "epoch": 0.6442385469519818,
+      "grad_norm": 0.9202976822853088,
+      "learning_rate": 5.937638327267668e-05,
+      "loss": 0.3431,
+      "step": 8761
+    },
+    {
+      "epoch": 0.6443120817707184,
+      "grad_norm": 0.6508467197418213,
+      "learning_rate": 5.935462195655186e-05,
+      "loss": 0.341,
+      "step": 8762
+    },
+    {
+      "epoch": 0.6443856165894551,
+      "grad_norm": 0.761972188949585,
+      "learning_rate": 5.933286294587579e-05,
+      "loss": 0.3379,
+      "step": 8763
+    },
+    {
+      "epoch": 0.6444591514081918,
+      "grad_norm": 0.7612439393997192,
+      "learning_rate": 5.9311106241882644e-05,
+      "loss": 0.3498,
+      "step": 8764
+    },
+    {
+      "epoch": 0.6445326862269285,
+      "grad_norm": 0.6324072480201721,
+      "learning_rate": 5.9289351845806506e-05,
+      "loss": 0.3513,
+      "step": 8765
+    },
+    {
+      "epoch": 0.6446062210456651,
+      "grad_norm": 1.2771495580673218,
+      "learning_rate": 5.9267599758881285e-05,
+      "loss": 0.3872,
+      "step": 8766
+    },
+    {
+      "epoch": 0.6446797558644018,
+      "grad_norm": 0.49184930324554443,
+      "learning_rate": 5.924584998234077e-05,
+      "loss": 0.3443,
+      "step": 8767
+    },
+    {
+      "epoch": 0.6447532906831385,
+      "grad_norm": 1.194439172744751,
+      "learning_rate": 5.92241025174187e-05,
+      "loss": 0.3748,
+      "step": 8768
+    },
+    {
+      "epoch": 0.6448268255018751,
+      "grad_norm": 0.44466596841812134,
+      "learning_rate": 5.920235736534856e-05,
+      "loss": 0.3373,
+      "step": 8769
+    },
+    {
+      "epoch": 0.6449003603206118,
+      "grad_norm": 0.9143692851066589,
+      "learning_rate": 5.918061452736376e-05,
+      "loss": 0.3426,
+      "step": 8770
+    },
+    {
+      "epoch": 0.6449738951393484,
+      "grad_norm": 0.5903466939926147,
+      "learning_rate": 5.9158874004697586e-05,
+      "loss": 0.3349,
+      "step": 8771
+    },
+    {
+      "epoch": 0.6450474299580852,
+      "grad_norm": 0.7614386677742004,
+      "learning_rate": 5.913713579858315e-05,
+      "loss": 0.3302,
+      "step": 8772
+    },
+    {
+      "epoch": 0.6451209647768218,
+      "grad_norm": 0.7714390754699707,
+      "learning_rate": 5.911539991025352e-05,
+      "loss": 0.3465,
+      "step": 8773
+    },
+    {
+      "epoch": 0.6451944995955585,
+      "grad_norm": 0.6166185736656189,
+      "learning_rate": 5.9093666340941536e-05,
+      "loss": 0.3236,
+      "step": 8774
+    },
+    {
+      "epoch": 0.6452680344142951,
+      "grad_norm": 1.3717399835586548,
+      "learning_rate": 5.907193509187998e-05,
+      "loss": 0.3606,
+      "step": 8775
+    },
+    {
+      "epoch": 0.6453415692330319,
+      "grad_norm": 0.4941730201244354,
+      "learning_rate": 5.9050206164301456e-05,
+      "loss": 0.3501,
+      "step": 8776
+    },
+    {
+      "epoch": 0.6454151040517685,
+      "grad_norm": 1.3776518106460571,
+      "learning_rate": 5.902847955943844e-05,
+      "loss": 0.3793,
+      "step": 8777
+    },
+    {
+      "epoch": 0.6454886388705052,
+      "grad_norm": 0.4541395306587219,
+      "learning_rate": 5.900675527852333e-05,
+      "loss": 0.3391,
+      "step": 8778
+    },
+    {
+      "epoch": 0.6455621736892418,
+      "grad_norm": 0.6447547674179077,
+      "learning_rate": 5.898503332278833e-05,
+      "loss": 0.3249,
+      "step": 8779
+    },
+    {
+      "epoch": 0.6456357085079786,
+      "grad_norm": 0.4802902638912201,
+      "learning_rate": 5.896331369346548e-05,
+      "loss": 0.3397,
+      "step": 8780
+    },
+    {
+      "epoch": 0.6457092433267152,
+      "grad_norm": 0.8745157122612,
+      "learning_rate": 5.894159639178687e-05,
+      "loss": 0.3392,
+      "step": 8781
+    },
+    {
+      "epoch": 0.6457827781454518,
+      "grad_norm": 0.5087623000144958,
+      "learning_rate": 5.891988141898417e-05,
+      "loss": 0.3209,
+      "step": 8782
+    },
+    {
+      "epoch": 0.6458563129641886,
+      "grad_norm": 0.8471574783325195,
+      "learning_rate": 5.889816877628915e-05,
+      "loss": 0.3119,
+      "step": 8783
+    },
+    {
+      "epoch": 0.6459298477829252,
+      "grad_norm": 0.9823768734931946,
+      "learning_rate": 5.8876458464933406e-05,
+      "loss": 0.3622,
+      "step": 8784
+    },
+    {
+      "epoch": 0.6460033826016619,
+      "grad_norm": 0.7960575819015503,
+      "learning_rate": 5.88547504861483e-05,
+      "loss": 0.3399,
+      "step": 8785
+    },
+    {
+      "epoch": 0.6460769174203985,
+      "grad_norm": 1.2454854249954224,
+      "learning_rate": 5.883304484116519e-05,
+      "loss": 0.3312,
+      "step": 8786
+    },
+    {
+      "epoch": 0.6461504522391353,
+      "grad_norm": 0.7292595505714417,
+      "learning_rate": 5.881134153121521e-05,
+      "loss": 0.3231,
+      "step": 8787
+    },
+    {
+      "epoch": 0.6462239870578719,
+      "grad_norm": 1.212473750114441,
+      "learning_rate": 5.8789640557529425e-05,
+      "loss": 0.3314,
+      "step": 8788
+    },
+    {
+      "epoch": 0.6462975218766086,
+      "grad_norm": 0.6359857320785522,
+      "learning_rate": 5.876794192133871e-05,
+      "loss": 0.3394,
+      "step": 8789
+    },
+    {
+      "epoch": 0.6463710566953452,
+      "grad_norm": 0.8536525964736938,
+      "learning_rate": 5.874624562387383e-05,
+      "loss": 0.356,
+      "step": 8790
+    },
+    {
+      "epoch": 0.646444591514082,
+      "grad_norm": 0.6393150091171265,
+      "learning_rate": 5.872455166636546e-05,
+      "loss": 0.3322,
+      "step": 8791
+    },
+    {
+      "epoch": 0.6465181263328186,
+      "grad_norm": 1.2294557094573975,
+      "learning_rate": 5.870286005004406e-05,
+      "loss": 0.3302,
+      "step": 8792
+    },
+    {
+      "epoch": 0.6465916611515553,
+      "grad_norm": 0.5068145394325256,
+      "learning_rate": 5.868117077614006e-05,
+      "loss": 0.3408,
+      "step": 8793
+    },
+    {
+      "epoch": 0.6466651959702919,
+      "grad_norm": 1.2147737741470337,
+      "learning_rate": 5.865948384588362e-05,
+      "loss": 0.333,
+      "step": 8794
+    },
+    {
+      "epoch": 0.6467387307890287,
+      "grad_norm": 0.5022394061088562,
+      "learning_rate": 5.8637799260504875e-05,
+      "loss": 0.3493,
+      "step": 8795
+    },
+    {
+      "epoch": 0.6468122656077653,
+      "grad_norm": 1.2144463062286377,
+      "learning_rate": 5.861611702123384e-05,
+      "loss": 0.3205,
+      "step": 8796
+    },
+    {
+      "epoch": 0.6468858004265019,
+      "grad_norm": 0.5641620755195618,
+      "learning_rate": 5.859443712930031e-05,
+      "loss": 0.3353,
+      "step": 8797
+    },
+    {
+      "epoch": 0.6469593352452386,
+      "grad_norm": 1.431090235710144,
+      "learning_rate": 5.857275958593398e-05,
+      "loss": 0.333,
+      "step": 8798
+    },
+    {
+      "epoch": 0.6470328700639753,
+      "grad_norm": 0.7521259188652039,
+      "learning_rate": 5.855108439236448e-05,
+      "loss": 0.3406,
+      "step": 8799
+    },
+    {
+      "epoch": 0.647106404882712,
+      "grad_norm": 1.4022005796432495,
+      "learning_rate": 5.852941154982117e-05,
+      "loss": 0.3315,
+      "step": 8800
+    },
+    {
+      "epoch": 0.6471799397014486,
+      "grad_norm": 1.0637414455413818,
+      "learning_rate": 5.850774105953345e-05,
+      "loss": 0.317,
+      "step": 8801
+    },
+    {
+      "epoch": 0.6472534745201853,
+      "grad_norm": 1.3926827907562256,
+      "learning_rate": 5.848607292273043e-05,
+      "loss": 0.3196,
+      "step": 8802
+    },
+    {
+      "epoch": 0.647327009338922,
+      "grad_norm": 1.4027336835861206,
+      "learning_rate": 5.846440714064114e-05,
+      "loss": 0.3232,
+      "step": 8803
+    },
+    {
+      "epoch": 0.6474005441576587,
+      "grad_norm": 1.123592734336853,
+      "learning_rate": 5.8442743714494585e-05,
+      "loss": 0.3992,
+      "step": 8804
+    },
+    {
+      "epoch": 0.6474740789763953,
+      "grad_norm": 0.8696422576904297,
+      "learning_rate": 5.84210826455194e-05,
+      "loss": 0.3449,
+      "step": 8805
+    },
+    {
+      "epoch": 0.647547613795132,
+      "grad_norm": 0.8173090815544128,
+      "learning_rate": 5.8399423934944306e-05,
+      "loss": 0.3435,
+      "step": 8806
+    },
+    {
+      "epoch": 0.6476211486138687,
+      "grad_norm": 1.026296854019165,
+      "learning_rate": 5.837776758399777e-05,
+      "loss": 0.3346,
+      "step": 8807
+    },
+    {
+      "epoch": 0.6476946834326054,
+      "grad_norm": 1.208998680114746,
+      "learning_rate": 5.835611359390822e-05,
+      "loss": 0.3346,
+      "step": 8808
+    },
+    {
+      "epoch": 0.647768218251342,
+      "grad_norm": 0.8659824728965759,
+      "learning_rate": 5.833446196590382e-05,
+      "loss": 0.3372,
+      "step": 8809
+    },
+    {
+      "epoch": 0.6478417530700786,
+      "grad_norm": 0.962339460849762,
+      "learning_rate": 5.83128127012127e-05,
+      "loss": 0.3054,
+      "step": 8810
+    },
+    {
+      "epoch": 0.6479152878888154,
+      "grad_norm": 1.0175743103027344,
+      "learning_rate": 5.829116580106288e-05,
+      "loss": 0.3532,
+      "step": 8811
+    },
+    {
+      "epoch": 0.647988822707552,
+      "grad_norm": 1.199742317199707,
+      "learning_rate": 5.826952126668212e-05,
+      "loss": 0.3681,
+      "step": 8812
+    },
+    {
+      "epoch": 0.6480623575262887,
+      "grad_norm": 1.3745979070663452,
+      "learning_rate": 5.824787909929815e-05,
+      "loss": 0.3543,
+      "step": 8813
+    },
+    {
+      "epoch": 0.6481358923450253,
+      "grad_norm": 1.2116116285324097,
+      "learning_rate": 5.8226239300138585e-05,
+      "loss": 0.3617,
+      "step": 8814
+    },
+    {
+      "epoch": 0.6482094271637621,
+      "grad_norm": 0.9171216487884521,
+      "learning_rate": 5.820460187043077e-05,
+      "loss": 0.3398,
+      "step": 8815
+    },
+    {
+      "epoch": 0.6482829619824987,
+      "grad_norm": 0.9172824025154114,
+      "learning_rate": 5.8182966811402075e-05,
+      "loss": 0.3262,
+      "step": 8816
+    },
+    {
+      "epoch": 0.6483564968012354,
+      "grad_norm": 0.9558387398719788,
+      "learning_rate": 5.816133412427961e-05,
+      "loss": 0.3162,
+      "step": 8817
+    },
+    {
+      "epoch": 0.648430031619972,
+      "grad_norm": 0.9697593450546265,
+      "learning_rate": 5.813970381029046e-05,
+      "loss": 0.3217,
+      "step": 8818
+    },
+    {
+      "epoch": 0.6485035664387088,
+      "grad_norm": 0.9768808484077454,
+      "learning_rate": 5.8118075870661446e-05,
+      "loss": 0.3156,
+      "step": 8819
+    },
+    {
+      "epoch": 0.6485771012574454,
+      "grad_norm": 1.496112585067749,
+      "learning_rate": 5.8096450306619366e-05,
+      "loss": 0.3356,
+      "step": 8820
+    },
+    {
+      "epoch": 0.6486506360761821,
+      "grad_norm": 0.8144456148147583,
+      "learning_rate": 5.8074827119390895e-05,
+      "loss": 0.3449,
+      "step": 8821
+    },
+    {
+      "epoch": 0.6487241708949187,
+      "grad_norm": 1.1250865459442139,
+      "learning_rate": 5.8053206310202423e-05,
+      "loss": 0.3324,
+      "step": 8822
+    },
+    {
+      "epoch": 0.6487977057136555,
+      "grad_norm": 0.6794196963310242,
+      "learning_rate": 5.803158788028037e-05,
+      "loss": 0.3551,
+      "step": 8823
+    },
+    {
+      "epoch": 0.6488712405323921,
+      "grad_norm": 0.8583236932754517,
+      "learning_rate": 5.8009971830850975e-05,
+      "loss": 0.3173,
+      "step": 8824
+    },
+    {
+      "epoch": 0.6489447753511287,
+      "grad_norm": 0.4636106491088867,
+      "learning_rate": 5.7988358163140254e-05,
+      "loss": 0.3741,
+      "step": 8825
+    },
+    {
+      "epoch": 0.6490183101698654,
+      "grad_norm": 0.8675144910812378,
+      "learning_rate": 5.7966746878374204e-05,
+      "loss": 0.3443,
+      "step": 8826
+    },
+    {
+      "epoch": 0.6490918449886021,
+      "grad_norm": 0.4090771973133087,
+      "learning_rate": 5.794513797777865e-05,
+      "loss": 0.3363,
+      "step": 8827
+    },
+    {
+      "epoch": 0.6491653798073388,
+      "grad_norm": 0.835857629776001,
+      "learning_rate": 5.7923531462579275e-05,
+      "loss": 0.3296,
+      "step": 8828
+    },
+    {
+      "epoch": 0.6492389146260754,
+      "grad_norm": 0.4148204028606415,
+      "learning_rate": 5.7901927334001546e-05,
+      "loss": 0.3092,
+      "step": 8829
+    },
+    {
+      "epoch": 0.6493124494448121,
+      "grad_norm": 0.881401002407074,
+      "learning_rate": 5.7880325593270926e-05,
+      "loss": 0.3577,
+      "step": 8830
+    },
+    {
+      "epoch": 0.6493859842635488,
+      "grad_norm": 0.6102088689804077,
+      "learning_rate": 5.7858726241612726e-05,
+      "loss": 0.3322,
+      "step": 8831
+    },
+    {
+      "epoch": 0.6494595190822855,
+      "grad_norm": 0.5548239350318909,
+      "learning_rate": 5.783712928025202e-05,
+      "loss": 0.358,
+      "step": 8832
+    },
+    {
+      "epoch": 0.6495330539010221,
+      "grad_norm": 0.6174277663230896,
+      "learning_rate": 5.781553471041383e-05,
+      "loss": 0.3631,
+      "step": 8833
+    },
+    {
+      "epoch": 0.6496065887197588,
+      "grad_norm": 0.41466525197029114,
+      "learning_rate": 5.779394253332308e-05,
+      "loss": 0.3207,
+      "step": 8834
+    },
+    {
+      "epoch": 0.6496801235384955,
+      "grad_norm": 0.8494731783866882,
+      "learning_rate": 5.7772352750204394e-05,
+      "loss": 0.3342,
+      "step": 8835
+    },
+    {
+      "epoch": 0.6497536583572322,
+      "grad_norm": 0.4867028594017029,
+      "learning_rate": 5.7750765362282436e-05,
+      "loss": 0.3232,
+      "step": 8836
+    },
+    {
+      "epoch": 0.6498271931759688,
+      "grad_norm": 0.8394287824630737,
+      "learning_rate": 5.77291803707817e-05,
+      "loss": 0.3828,
+      "step": 8837
+    },
+    {
+      "epoch": 0.6499007279947054,
+      "grad_norm": 0.7464196681976318,
+      "learning_rate": 5.77075977769264e-05,
+      "loss": 0.3422,
+      "step": 8838
+    },
+    {
+      "epoch": 0.6499742628134422,
+      "grad_norm": 0.48318204283714294,
+      "learning_rate": 5.768601758194086e-05,
+      "loss": 0.3364,
+      "step": 8839
+    },
+    {
+      "epoch": 0.6500477976321788,
+      "grad_norm": 0.7527589797973633,
+      "learning_rate": 5.7664439787048984e-05,
+      "loss": 0.3426,
+      "step": 8840
+    },
+    {
+      "epoch": 0.6501213324509155,
+      "grad_norm": 0.49119001626968384,
+      "learning_rate": 5.7642864393474815e-05,
+      "loss": 0.3432,
+      "step": 8841
+    },
+    {
+      "epoch": 0.6501948672696521,
+      "grad_norm": 0.46275538206100464,
+      "learning_rate": 5.7621291402442055e-05,
+      "loss": 0.3726,
+      "step": 8842
+    },
+    {
+      "epoch": 0.6502684020883889,
+      "grad_norm": 0.48731622099876404,
+      "learning_rate": 5.759972081517435e-05,
+      "loss": 0.3317,
+      "step": 8843
+    },
+    {
+      "epoch": 0.6503419369071255,
+      "grad_norm": 0.6994304060935974,
+      "learning_rate": 5.757815263289528e-05,
+      "loss": 0.3475,
+      "step": 8844
+    },
+    {
+      "epoch": 0.6504154717258622,
+      "grad_norm": 0.7189375758171082,
+      "learning_rate": 5.755658685682812e-05,
+      "loss": 0.3284,
+      "step": 8845
+    },
+    {
+      "epoch": 0.6504890065445988,
+      "grad_norm": 0.44530126452445984,
+      "learning_rate": 5.753502348819614e-05,
+      "loss": 0.3464,
+      "step": 8846
+    },
+    {
+      "epoch": 0.6505625413633356,
+      "grad_norm": 0.7976187467575073,
+      "learning_rate": 5.7513462528222486e-05,
+      "loss": 0.3053,
+      "step": 8847
+    },
+    {
+      "epoch": 0.6506360761820722,
+      "grad_norm": 0.49551352858543396,
+      "learning_rate": 5.749190397813004e-05,
+      "loss": 0.3299,
+      "step": 8848
+    },
+    {
+      "epoch": 0.6507096110008089,
+      "grad_norm": 0.6203333735466003,
+      "learning_rate": 5.747034783914165e-05,
+      "loss": 0.3517,
+      "step": 8849
+    },
+    {
+      "epoch": 0.6507831458195455,
+      "grad_norm": 0.42952191829681396,
+      "learning_rate": 5.7448794112480054e-05,
+      "loss": 0.3667,
+      "step": 8850
+    },
+    {
+      "epoch": 0.6508566806382823,
+      "grad_norm": 0.5836290717124939,
+      "learning_rate": 5.7427242799367774e-05,
+      "loss": 0.3201,
+      "step": 8851
+    },
+    {
+      "epoch": 0.6509302154570189,
+      "grad_norm": 0.6757875084877014,
+      "learning_rate": 5.740569390102717e-05,
+      "loss": 0.3378,
+      "step": 8852
+    },
+    {
+      "epoch": 0.6510037502757555,
+      "grad_norm": 0.5208500623703003,
+      "learning_rate": 5.7384147418680546e-05,
+      "loss": 0.3147,
+      "step": 8853
+    },
+    {
+      "epoch": 0.6510772850944923,
+      "grad_norm": 0.7658019065856934,
+      "learning_rate": 5.73626033535501e-05,
+      "loss": 0.3503,
+      "step": 8854
+    },
+    {
+      "epoch": 0.6511508199132289,
+      "grad_norm": 0.4873826801776886,
+      "learning_rate": 5.7341061706857755e-05,
+      "loss": 0.3589,
+      "step": 8855
+    },
+    {
+      "epoch": 0.6512243547319656,
+      "grad_norm": 0.7992836236953735,
+      "learning_rate": 5.7319522479825414e-05,
+      "loss": 0.3362,
+      "step": 8856
+    },
+    {
+      "epoch": 0.6512978895507022,
+      "grad_norm": 0.47449609637260437,
+      "learning_rate": 5.729798567367484e-05,
+      "loss": 0.331,
+      "step": 8857
+    },
+    {
+      "epoch": 0.651371424369439,
+      "grad_norm": 0.942480206489563,
+      "learning_rate": 5.727645128962753e-05,
+      "loss": 0.3456,
+      "step": 8858
+    },
+    {
+      "epoch": 0.6514449591881756,
+      "grad_norm": 0.5957028269767761,
+      "learning_rate": 5.7254919328905e-05,
+      "loss": 0.3347,
+      "step": 8859
+    },
+    {
+      "epoch": 0.6515184940069123,
+      "grad_norm": 0.7637529373168945,
+      "learning_rate": 5.7233389792728584e-05,
+      "loss": 0.3343,
+      "step": 8860
+    },
+    {
+      "epoch": 0.6515920288256489,
+      "grad_norm": 1.0370227098464966,
+      "learning_rate": 5.721186268231942e-05,
+      "loss": 0.3655,
+      "step": 8861
+    },
+    {
+      "epoch": 0.6516655636443857,
+      "grad_norm": 0.7225240468978882,
+      "learning_rate": 5.7190337998898536e-05,
+      "loss": 0.3438,
+      "step": 8862
+    },
+    {
+      "epoch": 0.6517390984631223,
+      "grad_norm": 0.5192568898200989,
+      "learning_rate": 5.716881574368691e-05,
+      "loss": 0.3602,
+      "step": 8863
+    },
+    {
+      "epoch": 0.651812633281859,
+      "grad_norm": 0.5848924517631531,
+      "learning_rate": 5.7147295917905245e-05,
+      "loss": 0.3493,
+      "step": 8864
+    },
+    {
+      "epoch": 0.6518861681005956,
+      "grad_norm": 0.6096547245979309,
+      "learning_rate": 5.7125778522774153e-05,
+      "loss": 0.371,
+      "step": 8865
+    },
+    {
+      "epoch": 0.6519597029193323,
+      "grad_norm": 0.5046768188476562,
+      "learning_rate": 5.7104263559514146e-05,
+      "loss": 0.3427,
+      "step": 8866
+    },
+    {
+      "epoch": 0.652033237738069,
+      "grad_norm": 0.5149642825126648,
+      "learning_rate": 5.708275102934561e-05,
+      "loss": 0.3326,
+      "step": 8867
+    },
+    {
+      "epoch": 0.6521067725568056,
+      "grad_norm": 0.47995755076408386,
+      "learning_rate": 5.7061240933488704e-05,
+      "loss": 0.3084,
+      "step": 8868
+    },
+    {
+      "epoch": 0.6521803073755423,
+      "grad_norm": 0.44972458481788635,
+      "learning_rate": 5.7039733273163523e-05,
+      "loss": 0.3406,
+      "step": 8869
+    },
+    {
+      "epoch": 0.652253842194279,
+      "grad_norm": 0.4482592046260834,
+      "learning_rate": 5.701822804959005e-05,
+      "loss": 0.318,
+      "step": 8870
+    },
+    {
+      "epoch": 0.6523273770130157,
+      "grad_norm": 0.46273866295814514,
+      "learning_rate": 5.6996725263987994e-05,
+      "loss": 0.3284,
+      "step": 8871
+    },
+    {
+      "epoch": 0.6524009118317523,
+      "grad_norm": 0.5163241028785706,
+      "learning_rate": 5.697522491757707e-05,
+      "loss": 0.3417,
+      "step": 8872
+    },
+    {
+      "epoch": 0.652474446650489,
+      "grad_norm": 0.42066386342048645,
+      "learning_rate": 5.695372701157683e-05,
+      "loss": 0.352,
+      "step": 8873
+    },
+    {
+      "epoch": 0.6525479814692257,
+      "grad_norm": 0.5187553763389587,
+      "learning_rate": 5.693223154720663e-05,
+      "loss": 0.3707,
+      "step": 8874
+    },
+    {
+      "epoch": 0.6526215162879624,
+      "grad_norm": 0.517228901386261,
+      "learning_rate": 5.6910738525685657e-05,
+      "loss": 0.3706,
+      "step": 8875
+    },
+    {
+      "epoch": 0.652695051106699,
+      "grad_norm": 0.6945897340774536,
+      "learning_rate": 5.688924794823307e-05,
+      "loss": 0.3405,
+      "step": 8876
+    },
+    {
+      "epoch": 0.6527685859254357,
+      "grad_norm": 0.4933670461177826,
+      "learning_rate": 5.686775981606789e-05,
+      "loss": 0.3516,
+      "step": 8877
+    },
+    {
+      "epoch": 0.6528421207441724,
+      "grad_norm": 0.5865617990493774,
+      "learning_rate": 5.684627413040884e-05,
+      "loss": 0.3428,
+      "step": 8878
+    },
+    {
+      "epoch": 0.6529156555629091,
+      "grad_norm": 0.46442219614982605,
+      "learning_rate": 5.682479089247468e-05,
+      "loss": 0.3552,
+      "step": 8879
+    },
+    {
+      "epoch": 0.6529891903816457,
+      "grad_norm": 0.5948463678359985,
+      "learning_rate": 5.6803310103483966e-05,
+      "loss": 0.3622,
+      "step": 8880
+    },
+    {
+      "epoch": 0.6530627252003823,
+      "grad_norm": 0.5443723201751709,
+      "learning_rate": 5.678183176465506e-05,
+      "loss": 0.3177,
+      "step": 8881
+    },
+    {
+      "epoch": 0.6531362600191191,
+      "grad_norm": 0.7218431830406189,
+      "learning_rate": 5.676035587720625e-05,
+      "loss": 0.348,
+      "step": 8882
+    },
+    {
+      "epoch": 0.6532097948378557,
+      "grad_norm": 0.8444048762321472,
+      "learning_rate": 5.6738882442355744e-05,
+      "loss": 0.3511,
+      "step": 8883
+    },
+    {
+      "epoch": 0.6532833296565924,
+      "grad_norm": 0.6575563549995422,
+      "learning_rate": 5.6717411461321435e-05,
+      "loss": 0.3519,
+      "step": 8884
+    },
+    {
+      "epoch": 0.653356864475329,
+      "grad_norm": 0.6374346017837524,
+      "learning_rate": 5.669594293532123e-05,
+      "loss": 0.3416,
+      "step": 8885
+    },
+    {
+      "epoch": 0.6534303992940658,
+      "grad_norm": 0.49256283044815063,
+      "learning_rate": 5.667447686557288e-05,
+      "loss": 0.3542,
+      "step": 8886
+    },
+    {
+      "epoch": 0.6535039341128024,
+      "grad_norm": 0.8250295519828796,
+      "learning_rate": 5.665301325329393e-05,
+      "loss": 0.3206,
+      "step": 8887
+    },
+    {
+      "epoch": 0.6535774689315391,
+      "grad_norm": 0.615329384803772,
+      "learning_rate": 5.6631552099701765e-05,
+      "loss": 0.325,
+      "step": 8888
+    },
+    {
+      "epoch": 0.6536510037502757,
+      "grad_norm": 0.7883737683296204,
+      "learning_rate": 5.661009340601373e-05,
+      "loss": 0.3557,
+      "step": 8889
+    },
+    {
+      "epoch": 0.6537245385690125,
+      "grad_norm": 0.7361504435539246,
+      "learning_rate": 5.658863717344705e-05,
+      "loss": 0.352,
+      "step": 8890
+    },
+    {
+      "epoch": 0.6537980733877491,
+      "grad_norm": 0.47302085161209106,
+      "learning_rate": 5.6567183403218626e-05,
+      "loss": 0.3364,
+      "step": 8891
+    },
+    {
+      "epoch": 0.6538716082064858,
+      "grad_norm": 0.7221549153327942,
+      "learning_rate": 5.65457320965454e-05,
+      "loss": 0.3302,
+      "step": 8892
+    },
+    {
+      "epoch": 0.6539451430252224,
+      "grad_norm": 0.49547410011291504,
+      "learning_rate": 5.652428325464414e-05,
+      "loss": 0.333,
+      "step": 8893
+    },
+    {
+      "epoch": 0.6540186778439591,
+      "grad_norm": 0.5057852864265442,
+      "learning_rate": 5.650283687873139e-05,
+      "loss": 0.3498,
+      "step": 8894
+    },
+    {
+      "epoch": 0.6540922126626958,
+      "grad_norm": 0.4154070019721985,
+      "learning_rate": 5.648139297002364e-05,
+      "loss": 0.3471,
+      "step": 8895
+    },
+    {
+      "epoch": 0.6541657474814324,
+      "grad_norm": 0.5977425575256348,
+      "learning_rate": 5.645995152973723e-05,
+      "loss": 0.3443,
+      "step": 8896
+    },
+    {
+      "epoch": 0.6542392823001691,
+      "grad_norm": 0.8068264722824097,
+      "learning_rate": 5.643851255908828e-05,
+      "loss": 0.3752,
+      "step": 8897
+    },
+    {
+      "epoch": 0.6543128171189058,
+      "grad_norm": 0.9938722252845764,
+      "learning_rate": 5.6417076059292916e-05,
+      "loss": 0.3332,
+      "step": 8898
+    },
+    {
+      "epoch": 0.6543863519376425,
+      "grad_norm": 0.6637061834335327,
+      "learning_rate": 5.6395642031566954e-05,
+      "loss": 0.3498,
+      "step": 8899
+    },
+    {
+      "epoch": 0.6544598867563791,
+      "grad_norm": 0.6948564648628235,
+      "learning_rate": 5.637421047712623e-05,
+      "loss": 0.3389,
+      "step": 8900
+    },
+    {
+      "epoch": 0.6545334215751158,
+      "grad_norm": 0.7808278799057007,
+      "learning_rate": 5.6352781397186293e-05,
+      "loss": 0.3439,
+      "step": 8901
+    },
+    {
+      "epoch": 0.6546069563938525,
+      "grad_norm": 0.5644370913505554,
+      "learning_rate": 5.633135479296266e-05,
+      "loss": 0.3509,
+      "step": 8902
+    },
+    {
+      "epoch": 0.6546804912125892,
+      "grad_norm": 0.9042781591415405,
+      "learning_rate": 5.630993066567071e-05,
+      "loss": 0.3188,
+      "step": 8903
+    },
+    {
+      "epoch": 0.6547540260313258,
+      "grad_norm": 0.4146818518638611,
+      "learning_rate": 5.628850901652556e-05,
+      "loss": 0.3676,
+      "step": 8904
+    },
+    {
+      "epoch": 0.6548275608500626,
+      "grad_norm": 1.0216302871704102,
+      "learning_rate": 5.626708984674229e-05,
+      "loss": 0.3441,
+      "step": 8905
+    },
+    {
+      "epoch": 0.6549010956687992,
+      "grad_norm": 0.4905248284339905,
+      "learning_rate": 5.6245673157535906e-05,
+      "loss": 0.3318,
+      "step": 8906
+    },
+    {
+      "epoch": 0.6549746304875359,
+      "grad_norm": 0.6184876561164856,
+      "learning_rate": 5.622425895012105e-05,
+      "loss": 0.3378,
+      "step": 8907
+    },
+    {
+      "epoch": 0.6550481653062725,
+      "grad_norm": 0.6306099891662598,
+      "learning_rate": 5.620284722571244e-05,
+      "loss": 0.3394,
+      "step": 8908
+    },
+    {
+      "epoch": 0.6551217001250091,
+      "grad_norm": 0.7167635560035706,
+      "learning_rate": 5.61814379855246e-05,
+      "loss": 0.3362,
+      "step": 8909
+    },
+    {
+      "epoch": 0.6551952349437459,
+      "grad_norm": 0.6107304692268372,
+      "learning_rate": 5.6160031230771825e-05,
+      "loss": 0.3157,
+      "step": 8910
+    },
+    {
+      "epoch": 0.6552687697624825,
+      "grad_norm": 0.6437468528747559,
+      "learning_rate": 5.6138626962668316e-05,
+      "loss": 0.3264,
+      "step": 8911
+    },
+    {
+      "epoch": 0.6553423045812192,
+      "grad_norm": 0.5289298892021179,
+      "learning_rate": 5.611722518242817e-05,
+      "loss": 0.3529,
+      "step": 8912
+    },
+    {
+      "epoch": 0.6554158393999558,
+      "grad_norm": 0.6182288527488708,
+      "learning_rate": 5.609582589126536e-05,
+      "loss": 0.3614,
+      "step": 8913
+    },
+    {
+      "epoch": 0.6554893742186926,
+      "grad_norm": 0.5588524341583252,
+      "learning_rate": 5.607442909039361e-05,
+      "loss": 0.3547,
+      "step": 8914
+    },
+    {
+      "epoch": 0.6555629090374292,
+      "grad_norm": 0.6578432321548462,
+      "learning_rate": 5.605303478102659e-05,
+      "loss": 0.3377,
+      "step": 8915
+    },
+    {
+      "epoch": 0.6556364438561659,
+      "grad_norm": 0.49485188722610474,
+      "learning_rate": 5.6031642964377865e-05,
+      "loss": 0.308,
+      "step": 8916
+    },
+    {
+      "epoch": 0.6557099786749025,
+      "grad_norm": 0.4708307981491089,
+      "learning_rate": 5.60102536416607e-05,
+      "loss": 0.3434,
+      "step": 8917
+    },
+    {
+      "epoch": 0.6557835134936393,
+      "grad_norm": 0.5991211533546448,
+      "learning_rate": 5.598886681408838e-05,
+      "loss": 0.3709,
+      "step": 8918
+    },
+    {
+      "epoch": 0.6558570483123759,
+      "grad_norm": 0.44052016735076904,
+      "learning_rate": 5.596748248287402e-05,
+      "loss": 0.3473,
+      "step": 8919
+    },
+    {
+      "epoch": 0.6559305831311126,
+      "grad_norm": 0.46854284405708313,
+      "learning_rate": 5.5946100649230496e-05,
+      "loss": 0.3179,
+      "step": 8920
+    },
+    {
+      "epoch": 0.6560041179498493,
+      "grad_norm": 0.4756245017051697,
+      "learning_rate": 5.5924721314370655e-05,
+      "loss": 0.34,
+      "step": 8921
+    },
+    {
+      "epoch": 0.6560776527685859,
+      "grad_norm": 0.4803054928779602,
+      "learning_rate": 5.5903344479507104e-05,
+      "loss": 0.3477,
+      "step": 8922
+    },
+    {
+      "epoch": 0.6561511875873226,
+      "grad_norm": 0.63947594165802,
+      "learning_rate": 5.588197014585242e-05,
+      "loss": 0.3538,
+      "step": 8923
+    },
+    {
+      "epoch": 0.6562247224060592,
+      "grad_norm": 0.4952636659145355,
+      "learning_rate": 5.5860598314618916e-05,
+      "loss": 0.3222,
+      "step": 8924
+    },
+    {
+      "epoch": 0.656298257224796,
+      "grad_norm": 0.6652458906173706,
+      "learning_rate": 5.583922898701888e-05,
+      "loss": 0.3352,
+      "step": 8925
+    },
+    {
+      "epoch": 0.6563717920435326,
+      "grad_norm": 0.9156947731971741,
+      "learning_rate": 5.581786216426439e-05,
+      "loss": 0.3154,
+      "step": 8926
+    },
+    {
+      "epoch": 0.6564453268622693,
+      "grad_norm": 0.5469861030578613,
+      "learning_rate": 5.5796497847567356e-05,
+      "loss": 0.3215,
+      "step": 8927
+    },
+    {
+      "epoch": 0.6565188616810059,
+      "grad_norm": 0.7771610617637634,
+      "learning_rate": 5.5775136038139616e-05,
+      "loss": 0.3184,
+      "step": 8928
+    },
+    {
+      "epoch": 0.6565923964997427,
+      "grad_norm": 0.5189828872680664,
+      "learning_rate": 5.575377673719286e-05,
+      "loss": 0.334,
+      "step": 8929
+    },
+    {
+      "epoch": 0.6566659313184793,
+      "grad_norm": 0.801257848739624,
+      "learning_rate": 5.5732419945938566e-05,
+      "loss": 0.326,
+      "step": 8930
+    },
+    {
+      "epoch": 0.656739466137216,
+      "grad_norm": 0.9380092024803162,
+      "learning_rate": 5.5711065665588116e-05,
+      "loss": 0.3503,
+      "step": 8931
+    },
+    {
+      "epoch": 0.6568130009559526,
+      "grad_norm": 0.5378051400184631,
+      "learning_rate": 5.56897138973528e-05,
+      "loss": 0.3595,
+      "step": 8932
+    },
+    {
+      "epoch": 0.6568865357746894,
+      "grad_norm": 0.9609537124633789,
+      "learning_rate": 5.5668364642443684e-05,
+      "loss": 0.3356,
+      "step": 8933
+    },
+    {
+      "epoch": 0.656960070593426,
+      "grad_norm": 0.5115134716033936,
+      "learning_rate": 5.5647017902071655e-05,
+      "loss": 0.3403,
+      "step": 8934
+    },
+    {
+      "epoch": 0.6570336054121627,
+      "grad_norm": 0.9006178379058838,
+      "learning_rate": 5.56256736774476e-05,
+      "loss": 0.4004,
+      "step": 8935
+    },
+    {
+      "epoch": 0.6571071402308993,
+      "grad_norm": 0.5625331997871399,
+      "learning_rate": 5.560433196978219e-05,
+      "loss": 0.3349,
+      "step": 8936
+    },
+    {
+      "epoch": 0.657180675049636,
+      "grad_norm": 0.4241753816604614,
+      "learning_rate": 5.558299278028587e-05,
+      "loss": 0.3463,
+      "step": 8937
+    },
+    {
+      "epoch": 0.6572542098683727,
+      "grad_norm": 0.6122838258743286,
+      "learning_rate": 5.55616561101691e-05,
+      "loss": 0.3142,
+      "step": 8938
+    },
+    {
+      "epoch": 0.6573277446871093,
+      "grad_norm": 0.5739588737487793,
+      "learning_rate": 5.5540321960642114e-05,
+      "loss": 0.3286,
+      "step": 8939
+    },
+    {
+      "epoch": 0.657401279505846,
+      "grad_norm": 0.7942254543304443,
+      "learning_rate": 5.551899033291496e-05,
+      "loss": 0.3414,
+      "step": 8940
+    },
+    {
+      "epoch": 0.6574748143245827,
+      "grad_norm": 0.5160457491874695,
+      "learning_rate": 5.5497661228197604e-05,
+      "loss": 0.3203,
+      "step": 8941
+    },
+    {
+      "epoch": 0.6575483491433194,
+      "grad_norm": 0.7081305384635925,
+      "learning_rate": 5.547633464769992e-05,
+      "loss": 0.32,
+      "step": 8942
+    },
+    {
+      "epoch": 0.657621883962056,
+      "grad_norm": 0.6106747388839722,
+      "learning_rate": 5.545501059263147e-05,
+      "loss": 0.3423,
+      "step": 8943
+    },
+    {
+      "epoch": 0.6576954187807927,
+      "grad_norm": 0.4887669086456299,
+      "learning_rate": 5.5433689064201846e-05,
+      "loss": 0.3344,
+      "step": 8944
+    },
+    {
+      "epoch": 0.6577689535995294,
+      "grad_norm": 0.7648841738700867,
+      "learning_rate": 5.5412370063620444e-05,
+      "loss": 0.3319,
+      "step": 8945
+    },
+    {
+      "epoch": 0.6578424884182661,
+      "grad_norm": 0.45820820331573486,
+      "learning_rate": 5.5391053592096464e-05,
+      "loss": 0.3264,
+      "step": 8946
+    },
+    {
+      "epoch": 0.6579160232370027,
+      "grad_norm": 0.6837165355682373,
+      "learning_rate": 5.536973965083896e-05,
+      "loss": 0.3271,
+      "step": 8947
+    },
+    {
+      "epoch": 0.6579895580557394,
+      "grad_norm": 0.5210690498352051,
+      "learning_rate": 5.534842824105693e-05,
+      "loss": 0.321,
+      "step": 8948
+    },
+    {
+      "epoch": 0.6580630928744761,
+      "grad_norm": 0.48744332790374756,
+      "learning_rate": 5.5327119363959215e-05,
+      "loss": 0.3336,
+      "step": 8949
+    },
+    {
+      "epoch": 0.6581366276932127,
+      "grad_norm": 0.539983332157135,
+      "learning_rate": 5.5305813020754385e-05,
+      "loss": 0.3474,
+      "step": 8950
+    },
+    {
+      "epoch": 0.6582101625119494,
+      "grad_norm": 0.6480966210365295,
+      "learning_rate": 5.5284509212651017e-05,
+      "loss": 0.3282,
+      "step": 8951
+    },
+    {
+      "epoch": 0.658283697330686,
+      "grad_norm": 0.4749270975589752,
+      "learning_rate": 5.526320794085751e-05,
+      "loss": 0.3368,
+      "step": 8952
+    },
+    {
+      "epoch": 0.6583572321494228,
+      "grad_norm": 0.6340076327323914,
+      "learning_rate": 5.524190920658202e-05,
+      "loss": 0.3417,
+      "step": 8953
+    },
+    {
+      "epoch": 0.6584307669681594,
+      "grad_norm": 0.5758188962936401,
+      "learning_rate": 5.522061301103267e-05,
+      "loss": 0.3789,
+      "step": 8954
+    },
+    {
+      "epoch": 0.6585043017868961,
+      "grad_norm": 0.47970864176750183,
+      "learning_rate": 5.519931935541747e-05,
+      "loss": 0.3384,
+      "step": 8955
+    },
+    {
+      "epoch": 0.6585778366056327,
+      "grad_norm": 0.9275467395782471,
+      "learning_rate": 5.5178028240944135e-05,
+      "loss": 0.3363,
+      "step": 8956
+    },
+    {
+      "epoch": 0.6586513714243695,
+      "grad_norm": 0.6690807938575745,
+      "learning_rate": 5.515673966882031e-05,
+      "loss": 0.3166,
+      "step": 8957
+    },
+    {
+      "epoch": 0.6587249062431061,
+      "grad_norm": 0.5646783113479614,
+      "learning_rate": 5.513545364025353e-05,
+      "loss": 0.3458,
+      "step": 8958
+    },
+    {
+      "epoch": 0.6587984410618428,
+      "grad_norm": 0.6714746952056885,
+      "learning_rate": 5.511417015645122e-05,
+      "loss": 0.3206,
+      "step": 8959
+    },
+    {
+      "epoch": 0.6588719758805794,
+      "grad_norm": 0.5775771737098694,
+      "learning_rate": 5.509288921862049e-05,
+      "loss": 0.3374,
+      "step": 8960
+    },
+    {
+      "epoch": 0.6589455106993162,
+      "grad_norm": 0.4839058816432953,
+      "learning_rate": 5.507161082796849e-05,
+      "loss": 0.3352,
+      "step": 8961
+    },
+    {
+      "epoch": 0.6590190455180528,
+      "grad_norm": 0.5009884834289551,
+      "learning_rate": 5.505033498570218e-05,
+      "loss": 0.3348,
+      "step": 8962
+    },
+    {
+      "epoch": 0.6590925803367895,
+      "grad_norm": 0.4344617426395416,
+      "learning_rate": 5.502906169302826e-05,
+      "loss": 0.3266,
+      "step": 8963
+    },
+    {
+      "epoch": 0.6591661151555261,
+      "grad_norm": 0.5510485172271729,
+      "learning_rate": 5.500779095115341e-05,
+      "loss": 0.3471,
+      "step": 8964
+    },
+    {
+      "epoch": 0.6592396499742628,
+      "grad_norm": 0.5602208971977234,
+      "learning_rate": 5.49865227612842e-05,
+      "loss": 0.3173,
+      "step": 8965
+    },
+    {
+      "epoch": 0.6593131847929995,
+      "grad_norm": 0.4418964087963104,
+      "learning_rate": 5.496525712462687e-05,
+      "loss": 0.3265,
+      "step": 8966
+    },
+    {
+      "epoch": 0.6593867196117361,
+      "grad_norm": 0.48034000396728516,
+      "learning_rate": 5.494399404238768e-05,
+      "loss": 0.352,
+      "step": 8967
+    },
+    {
+      "epoch": 0.6594602544304728,
+      "grad_norm": 0.6321499347686768,
+      "learning_rate": 5.492273351577275e-05,
+      "loss": 0.3501,
+      "step": 8968
+    },
+    {
+      "epoch": 0.6595337892492095,
+      "grad_norm": 0.4463176727294922,
+      "learning_rate": 5.4901475545987945e-05,
+      "loss": 0.3107,
+      "step": 8969
+    },
+    {
+      "epoch": 0.6596073240679462,
+      "grad_norm": 0.6728103756904602,
+      "learning_rate": 5.488022013423899e-05,
+      "loss": 0.3443,
+      "step": 8970
+    },
+    {
+      "epoch": 0.6596808588866828,
+      "grad_norm": 0.43073004484176636,
+      "learning_rate": 5.4858967281731575e-05,
+      "loss": 0.323,
+      "step": 8971
+    },
+    {
+      "epoch": 0.6597543937054196,
+      "grad_norm": 0.6240332126617432,
+      "learning_rate": 5.483771698967122e-05,
+      "loss": 0.3265,
+      "step": 8972
+    },
+    {
+      "epoch": 0.6598279285241562,
+      "grad_norm": 0.6825792193412781,
+      "learning_rate": 5.481646925926317e-05,
+      "loss": 0.3353,
+      "step": 8973
+    },
+    {
+      "epoch": 0.6599014633428929,
+      "grad_norm": 0.6142512559890747,
+      "learning_rate": 5.4795224091712674e-05,
+      "loss": 0.3342,
+      "step": 8974
+    },
+    {
+      "epoch": 0.6599749981616295,
+      "grad_norm": 0.8672756552696228,
+      "learning_rate": 5.477398148822481e-05,
+      "loss": 0.3316,
+      "step": 8975
+    },
+    {
+      "epoch": 0.6600485329803663,
+      "grad_norm": 0.40508782863616943,
+      "learning_rate": 5.4752741450004394e-05,
+      "loss": 0.3421,
+      "step": 8976
+    },
+    {
+      "epoch": 0.6601220677991029,
+      "grad_norm": 0.6268506646156311,
+      "learning_rate": 5.473150397825626e-05,
+      "loss": 0.3249,
+      "step": 8977
+    },
+    {
+      "epoch": 0.6601956026178395,
+      "grad_norm": 0.5498319864273071,
+      "learning_rate": 5.4710269074185015e-05,
+      "loss": 0.3118,
+      "step": 8978
+    },
+    {
+      "epoch": 0.6602691374365762,
+      "grad_norm": 0.560508131980896,
+      "learning_rate": 5.468903673899507e-05,
+      "loss": 0.3218,
+      "step": 8979
+    },
+    {
+      "epoch": 0.6603426722553128,
+      "grad_norm": 0.6557292938232422,
+      "learning_rate": 5.4667806973890824e-05,
+      "loss": 0.3182,
+      "step": 8980
+    },
+    {
+      "epoch": 0.6604162070740496,
+      "grad_norm": 0.8113361597061157,
+      "learning_rate": 5.464657978007636e-05,
+      "loss": 0.3364,
+      "step": 8981
+    },
+    {
+      "epoch": 0.6604897418927862,
+      "grad_norm": 1.1392184495925903,
+      "learning_rate": 5.46253551587558e-05,
+      "loss": 0.3029,
+      "step": 8982
+    },
+    {
+      "epoch": 0.6605632767115229,
+      "grad_norm": 0.6277981400489807,
+      "learning_rate": 5.460413311113293e-05,
+      "loss": 0.3449,
+      "step": 8983
+    },
+    {
+      "epoch": 0.6606368115302595,
+      "grad_norm": 1.2526192665100098,
+      "learning_rate": 5.458291363841155e-05,
+      "loss": 0.3156,
+      "step": 8984
+    },
+    {
+      "epoch": 0.6607103463489963,
+      "grad_norm": 0.564647912979126,
+      "learning_rate": 5.456169674179526e-05,
+      "loss": 0.356,
+      "step": 8985
+    },
+    {
+      "epoch": 0.6607838811677329,
+      "grad_norm": 1.4406375885009766,
+      "learning_rate": 5.454048242248746e-05,
+      "loss": 0.3565,
+      "step": 8986
+    },
+    {
+      "epoch": 0.6608574159864696,
+      "grad_norm": 0.592749297618866,
+      "learning_rate": 5.451927068169147e-05,
+      "loss": 0.3522,
+      "step": 8987
+    },
+    {
+      "epoch": 0.6609309508052063,
+      "grad_norm": 1.1480262279510498,
+      "learning_rate": 5.44980615206105e-05,
+      "loss": 0.3256,
+      "step": 8988
+    },
+    {
+      "epoch": 0.661004485623943,
+      "grad_norm": 0.5801066160202026,
+      "learning_rate": 5.447685494044743e-05,
+      "loss": 0.3325,
+      "step": 8989
+    },
+    {
+      "epoch": 0.6610780204426796,
+      "grad_norm": 1.3644534349441528,
+      "learning_rate": 5.445565094240521e-05,
+      "loss": 0.3245,
+      "step": 8990
+    },
+    {
+      "epoch": 0.6611515552614163,
+      "grad_norm": 0.40025702118873596,
+      "learning_rate": 5.4434449527686573e-05,
+      "loss": 0.3337,
+      "step": 8991
+    },
+    {
+      "epoch": 0.661225090080153,
+      "grad_norm": 1.1345428228378296,
+      "learning_rate": 5.4413250697494044e-05,
+      "loss": 0.3326,
+      "step": 8992
+    },
+    {
+      "epoch": 0.6612986248988896,
+      "grad_norm": 0.49898993968963623,
+      "learning_rate": 5.4392054453030007e-05,
+      "loss": 0.3526,
+      "step": 8993
+    },
+    {
+      "epoch": 0.6613721597176263,
+      "grad_norm": 0.9873189926147461,
+      "learning_rate": 5.437086079549676e-05,
+      "loss": 0.3368,
+      "step": 8994
+    },
+    {
+      "epoch": 0.6614456945363629,
+      "grad_norm": 1.0701957941055298,
+      "learning_rate": 5.43496697260965e-05,
+      "loss": 0.3503,
+      "step": 8995
+    },
+    {
+      "epoch": 0.6615192293550997,
+      "grad_norm": 0.534438967704773,
+      "learning_rate": 5.43284812460311e-05,
+      "loss": 0.3581,
+      "step": 8996
+    },
+    {
+      "epoch": 0.6615927641738363,
+      "grad_norm": 0.6869476437568665,
+      "learning_rate": 5.430729535650245e-05,
+      "loss": 0.323,
+      "step": 8997
+    },
+    {
+      "epoch": 0.661666298992573,
+      "grad_norm": 0.6213985681533813,
+      "learning_rate": 5.428611205871226e-05,
+      "loss": 0.349,
+      "step": 8998
+    },
+    {
+      "epoch": 0.6617398338113096,
+      "grad_norm": 0.49467939138412476,
+      "learning_rate": 5.4264931353862e-05,
+      "loss": 0.3179,
+      "step": 8999
+    },
+    {
+      "epoch": 0.6618133686300464,
+      "grad_norm": 0.5148028135299683,
+      "learning_rate": 5.424375324315311e-05,
+      "loss": 0.3502,
+      "step": 9000
+    },
+    {
+      "epoch": 0.661886903448783,
+      "grad_norm": 0.7300476431846619,
+      "learning_rate": 5.422257772778685e-05,
+      "loss": 0.3412,
+      "step": 9001
+    },
+    {
+      "epoch": 0.6619604382675197,
+      "grad_norm": 0.40457063913345337,
+      "learning_rate": 5.420140480896426e-05,
+      "loss": 0.3313,
+      "step": 9002
+    },
+    {
+      "epoch": 0.6620339730862563,
+      "grad_norm": 0.5417361855506897,
+      "learning_rate": 5.418023448788636e-05,
+      "loss": 0.3514,
+      "step": 9003
+    },
+    {
+      "epoch": 0.6621075079049931,
+      "grad_norm": 0.5143250226974487,
+      "learning_rate": 5.415906676575388e-05,
+      "loss": 0.3316,
+      "step": 9004
+    },
+    {
+      "epoch": 0.6621810427237297,
+      "grad_norm": 0.5877178311347961,
+      "learning_rate": 5.4137901643767556e-05,
+      "loss": 0.3683,
+      "step": 9005
+    },
+    {
+      "epoch": 0.6622545775424663,
+      "grad_norm": 0.677971363067627,
+      "learning_rate": 5.41167391231278e-05,
+      "loss": 0.3373,
+      "step": 9006
+    },
+    {
+      "epoch": 0.662328112361203,
+      "grad_norm": 0.5378010869026184,
+      "learning_rate": 5.409557920503502e-05,
+      "loss": 0.3282,
+      "step": 9007
+    },
+    {
+      "epoch": 0.6624016471799397,
+      "grad_norm": 0.664073646068573,
+      "learning_rate": 5.4074421890689475e-05,
+      "loss": 0.3537,
+      "step": 9008
+    },
+    {
+      "epoch": 0.6624751819986764,
+      "grad_norm": 1.0838494300842285,
+      "learning_rate": 5.4053267181291156e-05,
+      "loss": 0.3581,
+      "step": 9009
+    },
+    {
+      "epoch": 0.662548716817413,
+      "grad_norm": 0.5066359043121338,
+      "learning_rate": 5.403211507804e-05,
+      "loss": 0.3583,
+      "step": 9010
+    },
+    {
+      "epoch": 0.6626222516361497,
+      "grad_norm": 1.175115942955017,
+      "learning_rate": 5.401096558213584e-05,
+      "loss": 0.3442,
+      "step": 9011
+    },
+    {
+      "epoch": 0.6626957864548864,
+      "grad_norm": 0.5336934924125671,
+      "learning_rate": 5.3989818694778204e-05,
+      "loss": 0.3322,
+      "step": 9012
+    },
+    {
+      "epoch": 0.6627693212736231,
+      "grad_norm": 0.8527644276618958,
+      "learning_rate": 5.39686744171666e-05,
+      "loss": 0.33,
+      "step": 9013
+    },
+    {
+      "epoch": 0.6628428560923597,
+      "grad_norm": 0.7540430426597595,
+      "learning_rate": 5.3947532750500416e-05,
+      "loss": 0.3709,
+      "step": 9014
+    },
+    {
+      "epoch": 0.6629163909110964,
+      "grad_norm": 0.5777547955513,
+      "learning_rate": 5.392639369597876e-05,
+      "loss": 0.3408,
+      "step": 9015
+    },
+    {
+      "epoch": 0.6629899257298331,
+      "grad_norm": 0.8585905432701111,
+      "learning_rate": 5.3905257254800646e-05,
+      "loss": 0.3788,
+      "step": 9016
+    },
+    {
+      "epoch": 0.6630634605485698,
+      "grad_norm": 0.5297351479530334,
+      "learning_rate": 5.388412342816498e-05,
+      "loss": 0.3538,
+      "step": 9017
+    },
+    {
+      "epoch": 0.6631369953673064,
+      "grad_norm": 0.6117640137672424,
+      "learning_rate": 5.386299221727056e-05,
+      "loss": 0.3511,
+      "step": 9018
+    },
+    {
+      "epoch": 0.6632105301860431,
+      "grad_norm": 0.5433455109596252,
+      "learning_rate": 5.384186362331585e-05,
+      "loss": 0.3625,
+      "step": 9019
+    },
+    {
+      "epoch": 0.6632840650047798,
+      "grad_norm": 0.5279272198677063,
+      "learning_rate": 5.382073764749935e-05,
+      "loss": 0.3307,
+      "step": 9020
+    },
+    {
+      "epoch": 0.6633575998235164,
+      "grad_norm": 0.5868209004402161,
+      "learning_rate": 5.3799614291019396e-05,
+      "loss": 0.3345,
+      "step": 9021
+    },
+    {
+      "epoch": 0.6634311346422531,
+      "grad_norm": 0.7286995053291321,
+      "learning_rate": 5.377849355507403e-05,
+      "loss": 0.3698,
+      "step": 9022
+    },
+    {
+      "epoch": 0.6635046694609897,
+      "grad_norm": 0.5504971146583557,
+      "learning_rate": 5.375737544086129e-05,
+      "loss": 0.3366,
+      "step": 9023
+    },
+    {
+      "epoch": 0.6635782042797265,
+      "grad_norm": 0.6938560009002686,
+      "learning_rate": 5.373625994957906e-05,
+      "loss": 0.3551,
+      "step": 9024
+    },
+    {
+      "epoch": 0.6636517390984631,
+      "grad_norm": 0.7555395364761353,
+      "learning_rate": 5.3715147082424944e-05,
+      "loss": 0.3606,
+      "step": 9025
+    },
+    {
+      "epoch": 0.6637252739171998,
+      "grad_norm": 0.4948424994945526,
+      "learning_rate": 5.369403684059657e-05,
+      "loss": 0.3241,
+      "step": 9026
+    },
+    {
+      "epoch": 0.6637988087359364,
+      "grad_norm": 0.47408992052078247,
+      "learning_rate": 5.367292922529126e-05,
+      "loss": 0.3354,
+      "step": 9027
+    },
+    {
+      "epoch": 0.6638723435546732,
+      "grad_norm": 0.4390876293182373,
+      "learning_rate": 5.365182423770634e-05,
+      "loss": 0.3606,
+      "step": 9028
+    },
+    {
+      "epoch": 0.6639458783734098,
+      "grad_norm": 0.5207401514053345,
+      "learning_rate": 5.363072187903882e-05,
+      "loss": 0.3653,
+      "step": 9029
+    },
+    {
+      "epoch": 0.6640194131921465,
+      "grad_norm": 0.4247838854789734,
+      "learning_rate": 5.3609622150485685e-05,
+      "loss": 0.3673,
+      "step": 9030
+    },
+    {
+      "epoch": 0.6640929480108831,
+      "grad_norm": 0.5824205279350281,
+      "learning_rate": 5.3588525053243785e-05,
+      "loss": 0.3321,
+      "step": 9031
+    },
+    {
+      "epoch": 0.6641664828296199,
+      "grad_norm": 0.5858168601989746,
+      "learning_rate": 5.356743058850968e-05,
+      "loss": 0.3597,
+      "step": 9032
+    },
+    {
+      "epoch": 0.6642400176483565,
+      "grad_norm": 0.49212533235549927,
+      "learning_rate": 5.354633875747992e-05,
+      "loss": 0.3247,
+      "step": 9033
+    },
+    {
+      "epoch": 0.6643135524670931,
+      "grad_norm": 0.5873421430587769,
+      "learning_rate": 5.352524956135087e-05,
+      "loss": 0.3571,
+      "step": 9034
+    },
+    {
+      "epoch": 0.6643870872858298,
+      "grad_norm": 0.446135014295578,
+      "learning_rate": 5.3504163001318695e-05,
+      "loss": 0.3593,
+      "step": 9035
+    },
+    {
+      "epoch": 0.6644606221045665,
+      "grad_norm": 0.614727258682251,
+      "learning_rate": 5.3483079078579455e-05,
+      "loss": 0.3791,
+      "step": 9036
+    },
+    {
+      "epoch": 0.6645341569233032,
+      "grad_norm": 0.5380789041519165,
+      "learning_rate": 5.3461997794329096e-05,
+      "loss": 0.3718,
+      "step": 9037
+    },
+    {
+      "epoch": 0.6646076917420398,
+      "grad_norm": 0.8356353640556335,
+      "learning_rate": 5.3440919149763344e-05,
+      "loss": 0.383,
+      "step": 9038
+    },
+    {
+      "epoch": 0.6646812265607766,
+      "grad_norm": 0.9215155839920044,
+      "learning_rate": 5.3419843146077754e-05,
+      "loss": 0.3335,
+      "step": 9039
+    },
+    {
+      "epoch": 0.6647547613795132,
+      "grad_norm": 0.44171571731567383,
+      "learning_rate": 5.339876978446782e-05,
+      "loss": 0.3297,
+      "step": 9040
+    },
+    {
+      "epoch": 0.6648282961982499,
+      "grad_norm": 0.6840935945510864,
+      "learning_rate": 5.3377699066128885e-05,
+      "loss": 0.3429,
+      "step": 9041
+    },
+    {
+      "epoch": 0.6649018310169865,
+      "grad_norm": 0.5977712273597717,
+      "learning_rate": 5.3356630992256006e-05,
+      "loss": 0.3637,
+      "step": 9042
+    },
+    {
+      "epoch": 0.6649753658357233,
+      "grad_norm": 0.6553040146827698,
+      "learning_rate": 5.3335565564044256e-05,
+      "loss": 0.3578,
+      "step": 9043
+    },
+    {
+      "epoch": 0.6650489006544599,
+      "grad_norm": 0.7114161252975464,
+      "learning_rate": 5.33145027826885e-05,
+      "loss": 0.3405,
+      "step": 9044
+    },
+    {
+      "epoch": 0.6651224354731966,
+      "grad_norm": 0.397790789604187,
+      "learning_rate": 5.3293442649383395e-05,
+      "loss": 0.3151,
+      "step": 9045
+    },
+    {
+      "epoch": 0.6651959702919332,
+      "grad_norm": 0.6493392586708069,
+      "learning_rate": 5.327238516532351e-05,
+      "loss": 0.3427,
+      "step": 9046
+    },
+    {
+      "epoch": 0.66526950511067,
+      "grad_norm": 0.4530346691608429,
+      "learning_rate": 5.3251330331703284e-05,
+      "loss": 0.3431,
+      "step": 9047
+    },
+    {
+      "epoch": 0.6653430399294066,
+      "grad_norm": 0.5681231021881104,
+      "learning_rate": 5.3230278149716906e-05,
+      "loss": 0.3683,
+      "step": 9048
+    },
+    {
+      "epoch": 0.6654165747481432,
+      "grad_norm": 0.8552966117858887,
+      "learning_rate": 5.320922862055849e-05,
+      "loss": 0.357,
+      "step": 9049
+    },
+    {
+      "epoch": 0.6654901095668799,
+      "grad_norm": 0.4860028624534607,
+      "learning_rate": 5.318818174542206e-05,
+      "loss": 0.3949,
+      "step": 9050
+    },
+    {
+      "epoch": 0.6655636443856165,
+      "grad_norm": 0.5521372556686401,
+      "learning_rate": 5.316713752550138e-05,
+      "loss": 0.3543,
+      "step": 9051
+    },
+    {
+      "epoch": 0.6656371792043533,
+      "grad_norm": 0.4808638095855713,
+      "learning_rate": 5.314609596199002e-05,
+      "loss": 0.3452,
+      "step": 9052
+    },
+    {
+      "epoch": 0.6657107140230899,
+      "grad_norm": 0.4433295726776123,
+      "learning_rate": 5.312505705608155e-05,
+      "loss": 0.3613,
+      "step": 9053
+    },
+    {
+      "epoch": 0.6657842488418266,
+      "grad_norm": 0.6036554574966431,
+      "learning_rate": 5.310402080896936e-05,
+      "loss": 0.3612,
+      "step": 9054
+    },
+    {
+      "epoch": 0.6658577836605633,
+      "grad_norm": 0.39353638887405396,
+      "learning_rate": 5.3082987221846546e-05,
+      "loss": 0.3521,
+      "step": 9055
+    },
+    {
+      "epoch": 0.6659313184793,
+      "grad_norm": 0.6472161412239075,
+      "learning_rate": 5.3061956295906224e-05,
+      "loss": 0.3624,
+      "step": 9056
+    },
+    {
+      "epoch": 0.6660048532980366,
+      "grad_norm": 0.3938266932964325,
+      "learning_rate": 5.3040928032341316e-05,
+      "loss": 0.3554,
+      "step": 9057
+    },
+    {
+      "epoch": 0.6660783881167733,
+      "grad_norm": 0.7054498195648193,
+      "learning_rate": 5.301990243234449e-05,
+      "loss": 0.3435,
+      "step": 9058
+    },
+    {
+      "epoch": 0.66615192293551,
+      "grad_norm": 0.583750307559967,
+      "learning_rate": 5.299887949710838e-05,
+      "loss": 0.3632,
+      "step": 9059
+    },
+    {
+      "epoch": 0.6662254577542467,
+      "grad_norm": 0.6245880126953125,
+      "learning_rate": 5.297785922782546e-05,
+      "loss": 0.3653,
+      "step": 9060
+    },
+    {
+      "epoch": 0.6662989925729833,
+      "grad_norm": 0.6210644245147705,
+      "learning_rate": 5.295684162568798e-05,
+      "loss": 0.3629,
+      "step": 9061
+    },
+    {
+      "epoch": 0.6663725273917199,
+      "grad_norm": 0.49368733167648315,
+      "learning_rate": 5.2935826691888055e-05,
+      "loss": 0.3551,
+      "step": 9062
+    },
+    {
+      "epoch": 0.6664460622104567,
+      "grad_norm": 0.5267891883850098,
+      "learning_rate": 5.291481442761772e-05,
+      "loss": 0.3592,
+      "step": 9063
+    },
+    {
+      "epoch": 0.6665195970291933,
+      "grad_norm": 0.5470593571662903,
+      "learning_rate": 5.2893804834068826e-05,
+      "loss": 0.3807,
+      "step": 9064
+    },
+    {
+      "epoch": 0.66659313184793,
+      "grad_norm": 0.6644781231880188,
+      "learning_rate": 5.2872797912433e-05,
+      "loss": 0.3335,
+      "step": 9065
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.7985317707061768,
+      "learning_rate": 5.285179366390183e-05,
+      "loss": 0.3528,
+      "step": 9066
+    },
+    {
+      "epoch": 0.6667402014854034,
+      "grad_norm": 0.5807579755783081,
+      "learning_rate": 5.2830792089666705e-05,
+      "loss": 0.3437,
+      "step": 9067
+    },
+    {
+      "epoch": 0.66681373630414,
+      "grad_norm": 0.8439132571220398,
+      "learning_rate": 5.2809793190918785e-05,
+      "loss": 0.3437,
+      "step": 9068
+    },
+    {
+      "epoch": 0.6668872711228767,
+      "grad_norm": 0.5521168112754822,
+      "learning_rate": 5.278879696884921e-05,
+      "loss": 0.3667,
+      "step": 9069
+    },
+    {
+      "epoch": 0.6669608059416133,
+      "grad_norm": 0.8222416043281555,
+      "learning_rate": 5.276780342464893e-05,
+      "loss": 0.3431,
+      "step": 9070
+    },
+    {
+      "epoch": 0.6670343407603501,
+      "grad_norm": 0.6538490653038025,
+      "learning_rate": 5.274681255950864e-05,
+      "loss": 0.3673,
+      "step": 9071
+    },
+    {
+      "epoch": 0.6671078755790867,
+      "grad_norm": 0.7115405201911926,
+      "learning_rate": 5.272582437461902e-05,
+      "loss": 0.318,
+      "step": 9072
+    },
+    {
+      "epoch": 0.6671814103978234,
+      "grad_norm": 0.7003092169761658,
+      "learning_rate": 5.270483887117057e-05,
+      "loss": 0.3521,
+      "step": 9073
+    },
+    {
+      "epoch": 0.66725494521656,
+      "grad_norm": 0.7057807445526123,
+      "learning_rate": 5.268385605035357e-05,
+      "loss": 0.3923,
+      "step": 9074
+    },
+    {
+      "epoch": 0.6673284800352968,
+      "grad_norm": 1.0178412199020386,
+      "learning_rate": 5.2662875913358146e-05,
+      "loss": 0.343,
+      "step": 9075
+    },
+    {
+      "epoch": 0.6674020148540334,
+      "grad_norm": 0.49745509028434753,
+      "learning_rate": 5.264189846137437e-05,
+      "loss": 0.3626,
+      "step": 9076
+    },
+    {
+      "epoch": 0.66747554967277,
+      "grad_norm": 0.8197179436683655,
+      "learning_rate": 5.2620923695592116e-05,
+      "loss": 0.3501,
+      "step": 9077
+    },
+    {
+      "epoch": 0.6675490844915067,
+      "grad_norm": 0.550167977809906,
+      "learning_rate": 5.259995161720104e-05,
+      "loss": 0.3533,
+      "step": 9078
+    },
+    {
+      "epoch": 0.6676226193102434,
+      "grad_norm": 0.5464334487915039,
+      "learning_rate": 5.257898222739073e-05,
+      "loss": 0.3452,
+      "step": 9079
+    },
+    {
+      "epoch": 0.6676961541289801,
+      "grad_norm": 0.6096771955490112,
+      "learning_rate": 5.2558015527350626e-05,
+      "loss": 0.3369,
+      "step": 9080
+    },
+    {
+      "epoch": 0.6677696889477167,
+      "grad_norm": 0.5185221433639526,
+      "learning_rate": 5.253705151826992e-05,
+      "loss": 0.3661,
+      "step": 9081
+    },
+    {
+      "epoch": 0.6678432237664534,
+      "grad_norm": 0.49026668071746826,
+      "learning_rate": 5.251609020133774e-05,
+      "loss": 0.3497,
+      "step": 9082
+    },
+    {
+      "epoch": 0.6679167585851901,
+      "grad_norm": 0.45178863406181335,
+      "learning_rate": 5.2495131577743063e-05,
+      "loss": 0.3378,
+      "step": 9083
+    },
+    {
+      "epoch": 0.6679902934039268,
+      "grad_norm": 0.48347628116607666,
+      "learning_rate": 5.247417564867463e-05,
+      "loss": 0.3548,
+      "step": 9084
+    },
+    {
+      "epoch": 0.6680638282226634,
+      "grad_norm": 0.5355552434921265,
+      "learning_rate": 5.245322241532116e-05,
+      "loss": 0.3383,
+      "step": 9085
+    },
+    {
+      "epoch": 0.6681373630414001,
+      "grad_norm": 0.526624858379364,
+      "learning_rate": 5.243227187887103e-05,
+      "loss": 0.3418,
+      "step": 9086
+    },
+    {
+      "epoch": 0.6682108978601368,
+      "grad_norm": 0.6488314867019653,
+      "learning_rate": 5.241132404051269e-05,
+      "loss": 0.3252,
+      "step": 9087
+    },
+    {
+      "epoch": 0.6682844326788735,
+      "grad_norm": 0.5360782146453857,
+      "learning_rate": 5.2390378901434233e-05,
+      "loss": 0.3539,
+      "step": 9088
+    },
+    {
+      "epoch": 0.6683579674976101,
+      "grad_norm": 0.5337024331092834,
+      "learning_rate": 5.236943646282373e-05,
+      "loss": 0.3672,
+      "step": 9089
+    },
+    {
+      "epoch": 0.6684315023163467,
+      "grad_norm": 0.4782045781612396,
+      "learning_rate": 5.234849672586909e-05,
+      "loss": 0.3725,
+      "step": 9090
+    },
+    {
+      "epoch": 0.6685050371350835,
+      "grad_norm": 0.5059740543365479,
+      "learning_rate": 5.232755969175797e-05,
+      "loss": 0.36,
+      "step": 9091
+    },
+    {
+      "epoch": 0.6685785719538201,
+      "grad_norm": 0.4224645495414734,
+      "learning_rate": 5.230662536167797e-05,
+      "loss": 0.3439,
+      "step": 9092
+    },
+    {
+      "epoch": 0.6686521067725568,
+      "grad_norm": 0.43376508355140686,
+      "learning_rate": 5.228569373681655e-05,
+      "loss": 0.3528,
+      "step": 9093
+    },
+    {
+      "epoch": 0.6687256415912934,
+      "grad_norm": 0.5350832343101501,
+      "learning_rate": 5.226476481836091e-05,
+      "loss": 0.3575,
+      "step": 9094
+    },
+    {
+      "epoch": 0.6687991764100302,
+      "grad_norm": 0.5511617660522461,
+      "learning_rate": 5.2243838607498176e-05,
+      "loss": 0.3474,
+      "step": 9095
+    },
+    {
+      "epoch": 0.6688727112287668,
+      "grad_norm": 0.6100373864173889,
+      "learning_rate": 5.222291510541537e-05,
+      "loss": 0.3258,
+      "step": 9096
+    },
+    {
+      "epoch": 0.6689462460475035,
+      "grad_norm": 0.49237605929374695,
+      "learning_rate": 5.220199431329922e-05,
+      "loss": 0.3622,
+      "step": 9097
+    },
+    {
+      "epoch": 0.6690197808662401,
+      "grad_norm": 0.5537406802177429,
+      "learning_rate": 5.218107623233637e-05,
+      "loss": 0.3494,
+      "step": 9098
+    },
+    {
+      "epoch": 0.6690933156849769,
+      "grad_norm": 0.7974762320518494,
+      "learning_rate": 5.2160160863713336e-05,
+      "loss": 0.3437,
+      "step": 9099
+    },
+    {
+      "epoch": 0.6691668505037135,
+      "grad_norm": 0.7785638570785522,
+      "learning_rate": 5.213924820861651e-05,
+      "loss": 0.3542,
+      "step": 9100
+    },
+    {
+      "epoch": 0.6692403853224502,
+      "grad_norm": 0.6120486855506897,
+      "learning_rate": 5.211833826823198e-05,
+      "loss": 0.3469,
+      "step": 9101
+    },
+    {
+      "epoch": 0.6693139201411868,
+      "grad_norm": 0.5469079613685608,
+      "learning_rate": 5.209743104374585e-05,
+      "loss": 0.343,
+      "step": 9102
+    },
+    {
+      "epoch": 0.6693874549599236,
+      "grad_norm": 0.5131022334098816,
+      "learning_rate": 5.207652653634402e-05,
+      "loss": 0.333,
+      "step": 9103
+    },
+    {
+      "epoch": 0.6694609897786602,
+      "grad_norm": 0.5169462561607361,
+      "learning_rate": 5.2055624747212136e-05,
+      "loss": 0.3468,
+      "step": 9104
+    },
+    {
+      "epoch": 0.6695345245973968,
+      "grad_norm": 0.5003489255905151,
+      "learning_rate": 5.203472567753582e-05,
+      "loss": 0.3498,
+      "step": 9105
+    },
+    {
+      "epoch": 0.6696080594161335,
+      "grad_norm": 0.671402096748352,
+      "learning_rate": 5.201382932850052e-05,
+      "loss": 0.3524,
+      "step": 9106
+    },
+    {
+      "epoch": 0.6696815942348702,
+      "grad_norm": 0.4753052890300751,
+      "learning_rate": 5.1992935701291426e-05,
+      "loss": 0.3517,
+      "step": 9107
+    },
+    {
+      "epoch": 0.6697551290536069,
+      "grad_norm": 0.7441774010658264,
+      "learning_rate": 5.1972044797093725e-05,
+      "loss": 0.337,
+      "step": 9108
+    },
+    {
+      "epoch": 0.6698286638723435,
+      "grad_norm": 0.5064038038253784,
+      "learning_rate": 5.195115661709229e-05,
+      "loss": 0.3456,
+      "step": 9109
+    },
+    {
+      "epoch": 0.6699021986910803,
+      "grad_norm": 0.7473894357681274,
+      "learning_rate": 5.1930271162471985e-05,
+      "loss": 0.356,
+      "step": 9110
+    },
+    {
+      "epoch": 0.6699757335098169,
+      "grad_norm": 0.5534399151802063,
+      "learning_rate": 5.1909388434417397e-05,
+      "loss": 0.3464,
+      "step": 9111
+    },
+    {
+      "epoch": 0.6700492683285536,
+      "grad_norm": 0.9921430349349976,
+      "learning_rate": 5.188850843411306e-05,
+      "loss": 0.3647,
+      "step": 9112
+    },
+    {
+      "epoch": 0.6701228031472902,
+      "grad_norm": 0.6931642889976501,
+      "learning_rate": 5.186763116274332e-05,
+      "loss": 0.3344,
+      "step": 9113
+    },
+    {
+      "epoch": 0.670196337966027,
+      "grad_norm": 0.7508934736251831,
+      "learning_rate": 5.184675662149232e-05,
+      "loss": 0.3714,
+      "step": 9114
+    },
+    {
+      "epoch": 0.6702698727847636,
+      "grad_norm": 0.9772274494171143,
+      "learning_rate": 5.182588481154409e-05,
+      "loss": 0.3671,
+      "step": 9115
+    },
+    {
+      "epoch": 0.6703434076035003,
+      "grad_norm": 0.5330337882041931,
+      "learning_rate": 5.180501573408254e-05,
+      "loss": 0.3719,
+      "step": 9116
+    },
+    {
+      "epoch": 0.6704169424222369,
+      "grad_norm": 0.6764013171195984,
+      "learning_rate": 5.1784149390291334e-05,
+      "loss": 0.3757,
+      "step": 9117
+    },
+    {
+      "epoch": 0.6704904772409735,
+      "grad_norm": 0.5046889185905457,
+      "learning_rate": 5.176328578135405e-05,
+      "loss": 0.3532,
+      "step": 9118
+    },
+    {
+      "epoch": 0.6705640120597103,
+      "grad_norm": 0.6079386472702026,
+      "learning_rate": 5.174242490845416e-05,
+      "loss": 0.333,
+      "step": 9119
+    },
+    {
+      "epoch": 0.6706375468784469,
+      "grad_norm": 0.606886625289917,
+      "learning_rate": 5.1721566772774846e-05,
+      "loss": 0.3409,
+      "step": 9120
+    },
+    {
+      "epoch": 0.6707110816971836,
+      "grad_norm": 0.6533175110816956,
+      "learning_rate": 5.170071137549918e-05,
+      "loss": 0.3571,
+      "step": 9121
+    },
+    {
+      "epoch": 0.6707846165159203,
+      "grad_norm": 0.5158531665802002,
+      "learning_rate": 5.1679858717810134e-05,
+      "loss": 0.3444,
+      "step": 9122
+    },
+    {
+      "epoch": 0.670858151334657,
+      "grad_norm": 0.772067129611969,
+      "learning_rate": 5.165900880089054e-05,
+      "loss": 0.3663,
+      "step": 9123
+    },
+    {
+      "epoch": 0.6709316861533936,
+      "grad_norm": 0.49545982480049133,
+      "learning_rate": 5.1638161625922934e-05,
+      "loss": 0.3704,
+      "step": 9124
+    },
+    {
+      "epoch": 0.6710052209721303,
+      "grad_norm": 0.7756413221359253,
+      "learning_rate": 5.1617317194089846e-05,
+      "loss": 0.3476,
+      "step": 9125
+    },
+    {
+      "epoch": 0.671078755790867,
+      "grad_norm": 0.5851352214813232,
+      "learning_rate": 5.159647550657363e-05,
+      "loss": 0.3777,
+      "step": 9126
+    },
+    {
+      "epoch": 0.6711522906096037,
+      "grad_norm": 0.5338357090950012,
+      "learning_rate": 5.157563656455635e-05,
+      "loss": 0.3472,
+      "step": 9127
+    },
+    {
+      "epoch": 0.6712258254283403,
+      "grad_norm": 0.7941222190856934,
+      "learning_rate": 5.155480036922008e-05,
+      "loss": 0.3534,
+      "step": 9128
+    },
+    {
+      "epoch": 0.671299360247077,
+      "grad_norm": 0.9932007193565369,
+      "learning_rate": 5.15339669217467e-05,
+      "loss": 0.3527,
+      "step": 9129
+    },
+    {
+      "epoch": 0.6713728950658137,
+      "grad_norm": 0.5260328054428101,
+      "learning_rate": 5.1513136223317816e-05,
+      "loss": 0.3561,
+      "step": 9130
+    },
+    {
+      "epoch": 0.6714464298845504,
+      "grad_norm": 1.2769746780395508,
+      "learning_rate": 5.1492308275115065e-05,
+      "loss": 0.3356,
+      "step": 9131
+    },
+    {
+      "epoch": 0.671519964703287,
+      "grad_norm": 0.6628523468971252,
+      "learning_rate": 5.147148307831974e-05,
+      "loss": 0.352,
+      "step": 9132
+    },
+    {
+      "epoch": 0.6715934995220236,
+      "grad_norm": 1.1739606857299805,
+      "learning_rate": 5.1450660634113146e-05,
+      "loss": 0.3511,
+      "step": 9133
+    },
+    {
+      "epoch": 0.6716670343407604,
+      "grad_norm": 0.6106473207473755,
+      "learning_rate": 5.1429840943676265e-05,
+      "loss": 0.3592,
+      "step": 9134
+    },
+    {
+      "epoch": 0.671740569159497,
+      "grad_norm": 1.1467218399047852,
+      "learning_rate": 5.1409024008190084e-05,
+      "loss": 0.3378,
+      "step": 9135
+    },
+    {
+      "epoch": 0.6718141039782337,
+      "grad_norm": 0.49355265498161316,
+      "learning_rate": 5.1388209828835374e-05,
+      "loss": 0.3472,
+      "step": 9136
+    },
+    {
+      "epoch": 0.6718876387969703,
+      "grad_norm": 1.1108330488204956,
+      "learning_rate": 5.136739840679267e-05,
+      "loss": 0.3369,
+      "step": 9137
+    },
+    {
+      "epoch": 0.6719611736157071,
+      "grad_norm": 0.4994361698627472,
+      "learning_rate": 5.134658974324245e-05,
+      "loss": 0.3613,
+      "step": 9138
+    },
+    {
+      "epoch": 0.6720347084344437,
+      "grad_norm": 1.2831515073776245,
+      "learning_rate": 5.132578383936504e-05,
+      "loss": 0.3435,
+      "step": 9139
+    },
+    {
+      "epoch": 0.6721082432531804,
+      "grad_norm": 0.5916924476623535,
+      "learning_rate": 5.130498069634049e-05,
+      "loss": 0.3543,
+      "step": 9140
+    },
+    {
+      "epoch": 0.672181778071917,
+      "grad_norm": 1.1730097532272339,
+      "learning_rate": 5.128418031534884e-05,
+      "loss": 0.3447,
+      "step": 9141
+    },
+    {
+      "epoch": 0.6722553128906538,
+      "grad_norm": 0.5837717056274414,
+      "learning_rate": 5.126338269756993e-05,
+      "loss": 0.3634,
+      "step": 9142
+    },
+    {
+      "epoch": 0.6723288477093904,
+      "grad_norm": 1.0217790603637695,
+      "learning_rate": 5.124258784418339e-05,
+      "loss": 0.3825,
+      "step": 9143
+    },
+    {
+      "epoch": 0.6724023825281271,
+      "grad_norm": 0.6228481531143188,
+      "learning_rate": 5.122179575636867e-05,
+      "loss": 0.3519,
+      "step": 9144
+    },
+    {
+      "epoch": 0.6724759173468637,
+      "grad_norm": 0.824905514717102,
+      "learning_rate": 5.120100643530519e-05,
+      "loss": 0.373,
+      "step": 9145
+    },
+    {
+      "epoch": 0.6725494521656004,
+      "grad_norm": 0.919682502746582,
+      "learning_rate": 5.1180219882172164e-05,
+      "loss": 0.3386,
+      "step": 9146
+    },
+    {
+      "epoch": 0.6726229869843371,
+      "grad_norm": 0.7107966542243958,
+      "learning_rate": 5.115943609814853e-05,
+      "loss": 0.3487,
+      "step": 9147
+    },
+    {
+      "epoch": 0.6726965218030737,
+      "grad_norm": 0.8226402997970581,
+      "learning_rate": 5.1138655084413246e-05,
+      "loss": 0.3644,
+      "step": 9148
+    },
+    {
+      "epoch": 0.6727700566218104,
+      "grad_norm": 0.4723314642906189,
+      "learning_rate": 5.111787684214504e-05,
+      "loss": 0.3512,
+      "step": 9149
+    },
+    {
+      "epoch": 0.6728435914405471,
+      "grad_norm": 0.5347824096679688,
+      "learning_rate": 5.1097101372522405e-05,
+      "loss": 0.3405,
+      "step": 9150
+    },
+    {
+      "epoch": 0.6729171262592838,
+      "grad_norm": 0.7331796884536743,
+      "learning_rate": 5.107632867672378e-05,
+      "loss": 0.3761,
+      "step": 9151
+    },
+    {
+      "epoch": 0.6729906610780204,
+      "grad_norm": 0.7319709062576294,
+      "learning_rate": 5.105555875592749e-05,
+      "loss": 0.3457,
+      "step": 9152
+    },
+    {
+      "epoch": 0.6730641958967571,
+      "grad_norm": 0.5966716408729553,
+      "learning_rate": 5.103479161131151e-05,
+      "loss": 0.3571,
+      "step": 9153
+    },
+    {
+      "epoch": 0.6731377307154938,
+      "grad_norm": 0.8328314423561096,
+      "learning_rate": 5.101402724405382e-05,
+      "loss": 0.3589,
+      "step": 9154
+    },
+    {
+      "epoch": 0.6732112655342305,
+      "grad_norm": 0.4983055591583252,
+      "learning_rate": 5.0993265655332237e-05,
+      "loss": 0.3474,
+      "step": 9155
+    },
+    {
+      "epoch": 0.6732848003529671,
+      "grad_norm": 0.6110027432441711,
+      "learning_rate": 5.097250684632435e-05,
+      "loss": 0.3623,
+      "step": 9156
+    },
+    {
+      "epoch": 0.6733583351717038,
+      "grad_norm": 0.5487142205238342,
+      "learning_rate": 5.0951750818207575e-05,
+      "loss": 0.3405,
+      "step": 9157
+    },
+    {
+      "epoch": 0.6734318699904405,
+      "grad_norm": 0.5328032374382019,
+      "learning_rate": 5.093099757215926e-05,
+      "loss": 0.3445,
+      "step": 9158
+    },
+    {
+      "epoch": 0.6735054048091772,
+      "grad_norm": 0.9016329646110535,
+      "learning_rate": 5.0910247109356576e-05,
+      "loss": 0.3992,
+      "step": 9159
+    },
+    {
+      "epoch": 0.6735789396279138,
+      "grad_norm": 0.4461432099342346,
+      "learning_rate": 5.0889499430976454e-05,
+      "loss": 0.3627,
+      "step": 9160
+    },
+    {
+      "epoch": 0.6736524744466504,
+      "grad_norm": 0.516396701335907,
+      "learning_rate": 5.0868754538195754e-05,
+      "loss": 0.3392,
+      "step": 9161
+    },
+    {
+      "epoch": 0.6737260092653872,
+      "grad_norm": 0.4211825728416443,
+      "learning_rate": 5.084801243219117e-05,
+      "loss": 0.3351,
+      "step": 9162
+    },
+    {
+      "epoch": 0.6737995440841238,
+      "grad_norm": 0.42536938190460205,
+      "learning_rate": 5.0827273114139173e-05,
+      "loss": 0.3329,
+      "step": 9163
+    },
+    {
+      "epoch": 0.6738730789028605,
+      "grad_norm": 0.6765327453613281,
+      "learning_rate": 5.080653658521613e-05,
+      "loss": 0.3447,
+      "step": 9164
+    },
+    {
+      "epoch": 0.6739466137215971,
+      "grad_norm": 0.47629064321517944,
+      "learning_rate": 5.078580284659829e-05,
+      "loss": 0.3552,
+      "step": 9165
+    },
+    {
+      "epoch": 0.6740201485403339,
+      "grad_norm": 0.5830737352371216,
+      "learning_rate": 5.0765071899461603e-05,
+      "loss": 0.3373,
+      "step": 9166
+    },
+    {
+      "epoch": 0.6740936833590705,
+      "grad_norm": 0.5778400897979736,
+      "learning_rate": 5.074434374498205e-05,
+      "loss": 0.3544,
+      "step": 9167
+    },
+    {
+      "epoch": 0.6741672181778072,
+      "grad_norm": 0.4746284782886505,
+      "learning_rate": 5.0723618384335256e-05,
+      "loss": 0.3467,
+      "step": 9168
+    },
+    {
+      "epoch": 0.6742407529965438,
+      "grad_norm": 0.5475237369537354,
+      "learning_rate": 5.0702895818696875e-05,
+      "loss": 0.3301,
+      "step": 9169
+    },
+    {
+      "epoch": 0.6743142878152806,
+      "grad_norm": 0.45131635665893555,
+      "learning_rate": 5.0682176049242215e-05,
+      "loss": 0.3434,
+      "step": 9170
+    },
+    {
+      "epoch": 0.6743878226340172,
+      "grad_norm": 0.5072926878929138,
+      "learning_rate": 5.066145907714659e-05,
+      "loss": 0.3461,
+      "step": 9171
+    },
+    {
+      "epoch": 0.6744613574527539,
+      "grad_norm": 0.48946765065193176,
+      "learning_rate": 5.06407449035851e-05,
+      "loss": 0.3737,
+      "step": 9172
+    },
+    {
+      "epoch": 0.6745348922714905,
+      "grad_norm": 0.47150281071662903,
+      "learning_rate": 5.0620033529732627e-05,
+      "loss": 0.3341,
+      "step": 9173
+    },
+    {
+      "epoch": 0.6746084270902272,
+      "grad_norm": 0.49534323811531067,
+      "learning_rate": 5.059932495676396e-05,
+      "loss": 0.3585,
+      "step": 9174
+    },
+    {
+      "epoch": 0.6746819619089639,
+      "grad_norm": 0.44053879380226135,
+      "learning_rate": 5.057861918585375e-05,
+      "loss": 0.3346,
+      "step": 9175
+    },
+    {
+      "epoch": 0.6747554967277005,
+      "grad_norm": 0.4377676546573639,
+      "learning_rate": 5.0557916218176385e-05,
+      "loss": 0.3338,
+      "step": 9176
+    },
+    {
+      "epoch": 0.6748290315464373,
+      "grad_norm": 0.41956454515457153,
+      "learning_rate": 5.053721605490618e-05,
+      "loss": 0.4085,
+      "step": 9177
+    },
+    {
+      "epoch": 0.6749025663651739,
+      "grad_norm": 0.4047725200653076,
+      "learning_rate": 5.051651869721733e-05,
+      "loss": 0.3409,
+      "step": 9178
+    },
+    {
+      "epoch": 0.6749761011839106,
+      "grad_norm": 0.42136096954345703,
+      "learning_rate": 5.049582414628375e-05,
+      "loss": 0.343,
+      "step": 9179
+    },
+    {
+      "epoch": 0.6750496360026472,
+      "grad_norm": 0.5853437781333923,
+      "learning_rate": 5.047513240327923e-05,
+      "loss": 0.3512,
+      "step": 9180
+    },
+    {
+      "epoch": 0.675123170821384,
+      "grad_norm": 0.516459047794342,
+      "learning_rate": 5.045444346937749e-05,
+      "loss": 0.335,
+      "step": 9181
+    },
+    {
+      "epoch": 0.6751967056401206,
+      "grad_norm": 0.4359839856624603,
+      "learning_rate": 5.043375734575202e-05,
+      "loss": 0.3634,
+      "step": 9182
+    },
+    {
+      "epoch": 0.6752702404588573,
+      "grad_norm": 0.5395663976669312,
+      "learning_rate": 5.041307403357611e-05,
+      "loss": 0.3683,
+      "step": 9183
+    },
+    {
+      "epoch": 0.6753437752775939,
+      "grad_norm": 0.4857313930988312,
+      "learning_rate": 5.039239353402296e-05,
+      "loss": 0.3655,
+      "step": 9184
+    },
+    {
+      "epoch": 0.6754173100963307,
+      "grad_norm": 0.48797497153282166,
+      "learning_rate": 5.0371715848265657e-05,
+      "loss": 0.3499,
+      "step": 9185
+    },
+    {
+      "epoch": 0.6754908449150673,
+      "grad_norm": 0.4655068516731262,
+      "learning_rate": 5.035104097747696e-05,
+      "loss": 0.3663,
+      "step": 9186
+    },
+    {
+      "epoch": 0.675564379733804,
+      "grad_norm": 0.5279086232185364,
+      "learning_rate": 5.033036892282962e-05,
+      "loss": 0.3649,
+      "step": 9187
+    },
+    {
+      "epoch": 0.6756379145525406,
+      "grad_norm": 0.6693716645240784,
+      "learning_rate": 5.0309699685496217e-05,
+      "loss": 0.3574,
+      "step": 9188
+    },
+    {
+      "epoch": 0.6757114493712773,
+      "grad_norm": 0.4546281397342682,
+      "learning_rate": 5.028903326664903e-05,
+      "loss": 0.3343,
+      "step": 9189
+    },
+    {
+      "epoch": 0.675784984190014,
+      "grad_norm": 0.5276642441749573,
+      "learning_rate": 5.0268369667460394e-05,
+      "loss": 0.3919,
+      "step": 9190
+    },
+    {
+      "epoch": 0.6758585190087506,
+      "grad_norm": 0.41294699907302856,
+      "learning_rate": 5.024770888910226e-05,
+      "loss": 0.3725,
+      "step": 9191
+    },
+    {
+      "epoch": 0.6759320538274873,
+      "grad_norm": 0.4731881618499756,
+      "learning_rate": 5.0227050932746646e-05,
+      "loss": 0.3341,
+      "step": 9192
+    },
+    {
+      "epoch": 0.676005588646224,
+      "grad_norm": 0.4926137328147888,
+      "learning_rate": 5.020639579956517e-05,
+      "loss": 0.3387,
+      "step": 9193
+    },
+    {
+      "epoch": 0.6760791234649607,
+      "grad_norm": 0.5178477168083191,
+      "learning_rate": 5.018574349072949e-05,
+      "loss": 0.3693,
+      "step": 9194
+    },
+    {
+      "epoch": 0.6761526582836973,
+      "grad_norm": 0.6052430272102356,
+      "learning_rate": 5.0165094007411025e-05,
+      "loss": 0.3467,
+      "step": 9195
+    },
+    {
+      "epoch": 0.676226193102434,
+      "grad_norm": 0.4509792625904083,
+      "learning_rate": 5.014444735078101e-05,
+      "loss": 0.3567,
+      "step": 9196
+    },
+    {
+      "epoch": 0.6762997279211707,
+      "grad_norm": 0.5323389172554016,
+      "learning_rate": 5.012380352201054e-05,
+      "loss": 0.3602,
+      "step": 9197
+    },
+    {
+      "epoch": 0.6763732627399074,
+      "grad_norm": 0.45209479331970215,
+      "learning_rate": 5.010316252227061e-05,
+      "loss": 0.3584,
+      "step": 9198
+    },
+    {
+      "epoch": 0.676446797558644,
+      "grad_norm": 0.5782949924468994,
+      "learning_rate": 5.008252435273193e-05,
+      "loss": 0.3451,
+      "step": 9199
+    },
+    {
+      "epoch": 0.6765203323773807,
+      "grad_norm": 0.5385561585426331,
+      "learning_rate": 5.0061889014565144e-05,
+      "loss": 0.3245,
+      "step": 9200
+    },
+    {
+      "epoch": 0.6765938671961174,
+      "grad_norm": 0.510797917842865,
+      "learning_rate": 5.0041256508940706e-05,
+      "loss": 0.354,
+      "step": 9201
+    },
+    {
+      "epoch": 0.676667402014854,
+      "grad_norm": 0.5987036228179932,
+      "learning_rate": 5.0020626837029005e-05,
+      "loss": 0.3381,
+      "step": 9202
+    },
+    {
+      "epoch": 0.6767409368335907,
+      "grad_norm": 0.5135964751243591,
+      "learning_rate": 5.000000000000002e-05,
+      "loss": 0.3502,
+      "step": 9203
+    },
+    {
+      "epoch": 0.6768144716523273,
+      "grad_norm": 0.6540167331695557,
+      "learning_rate": 4.9979375999023804e-05,
+      "loss": 0.3301,
+      "step": 9204
+    },
+    {
+      "epoch": 0.6768880064710641,
+      "grad_norm": 0.45795777440071106,
+      "learning_rate": 4.99587548352702e-05,
+      "loss": 0.3942,
+      "step": 9205
+    },
+    {
+      "epoch": 0.6769615412898007,
+      "grad_norm": 0.6248285174369812,
+      "learning_rate": 4.993813650990881e-05,
+      "loss": 0.3686,
+      "step": 9206
+    },
+    {
+      "epoch": 0.6770350761085374,
+      "grad_norm": 0.6636595726013184,
+      "learning_rate": 4.9917521024109135e-05,
+      "loss": 0.346,
+      "step": 9207
+    },
+    {
+      "epoch": 0.677108610927274,
+      "grad_norm": 0.45447632670402527,
+      "learning_rate": 4.9896908379040564e-05,
+      "loss": 0.3457,
+      "step": 9208
+    },
+    {
+      "epoch": 0.6771821457460108,
+      "grad_norm": 0.5769398808479309,
+      "learning_rate": 4.987629857587219e-05,
+      "loss": 0.3542,
+      "step": 9209
+    },
+    {
+      "epoch": 0.6772556805647474,
+      "grad_norm": 0.601531982421875,
+      "learning_rate": 4.985569161577306e-05,
+      "loss": 0.3587,
+      "step": 9210
+    },
+    {
+      "epoch": 0.6773292153834841,
+      "grad_norm": 0.44370830059051514,
+      "learning_rate": 4.983508749991206e-05,
+      "loss": 0.3734,
+      "step": 9211
+    },
+    {
+      "epoch": 0.6774027502022207,
+      "grad_norm": 0.6808139681816101,
+      "learning_rate": 4.9814486229457794e-05,
+      "loss": 0.3323,
+      "step": 9212
+    },
+    {
+      "epoch": 0.6774762850209575,
+      "grad_norm": 0.4772658944129944,
+      "learning_rate": 4.979388780557888e-05,
+      "loss": 0.3545,
+      "step": 9213
+    },
+    {
+      "epoch": 0.6775498198396941,
+      "grad_norm": 0.727008044719696,
+      "learning_rate": 4.977329222944358e-05,
+      "loss": 0.363,
+      "step": 9214
+    },
+    {
+      "epoch": 0.6776233546584308,
+      "grad_norm": 0.561408519744873,
+      "learning_rate": 4.97526995022202e-05,
+      "loss": 0.3543,
+      "step": 9215
+    },
+    {
+      "epoch": 0.6776968894771674,
+      "grad_norm": 0.43456777930259705,
+      "learning_rate": 4.973210962507669e-05,
+      "loss": 0.3794,
+      "step": 9216
+    },
+    {
+      "epoch": 0.6777704242959041,
+      "grad_norm": 0.40223437547683716,
+      "learning_rate": 4.971152259918097e-05,
+      "loss": 0.3505,
+      "step": 9217
+    },
+    {
+      "epoch": 0.6778439591146408,
+      "grad_norm": 0.4362450838088989,
+      "learning_rate": 4.969093842570081e-05,
+      "loss": 0.3204,
+      "step": 9218
+    },
+    {
+      "epoch": 0.6779174939333774,
+      "grad_norm": 0.6654127836227417,
+      "learning_rate": 4.9670357105803665e-05,
+      "loss": 0.3506,
+      "step": 9219
+    },
+    {
+      "epoch": 0.6779910287521141,
+      "grad_norm": 0.48708340525627136,
+      "learning_rate": 4.964977864065699e-05,
+      "loss": 0.3384,
+      "step": 9220
+    },
+    {
+      "epoch": 0.6780645635708508,
+      "grad_norm": 0.5846834778785706,
+      "learning_rate": 4.962920303142803e-05,
+      "loss": 0.3584,
+      "step": 9221
+    },
+    {
+      "epoch": 0.6781380983895875,
+      "grad_norm": 0.5075253248214722,
+      "learning_rate": 4.9608630279283805e-05,
+      "loss": 0.3281,
+      "step": 9222
+    },
+    {
+      "epoch": 0.6782116332083241,
+      "grad_norm": 0.4328802227973938,
+      "learning_rate": 4.958806038539125e-05,
+      "loss": 0.3599,
+      "step": 9223
+    },
+    {
+      "epoch": 0.6782851680270608,
+      "grad_norm": 0.800054669380188,
+      "learning_rate": 4.9567493350917114e-05,
+      "loss": 0.3534,
+      "step": 9224
+    },
+    {
+      "epoch": 0.6783587028457975,
+      "grad_norm": 0.49771827459335327,
+      "learning_rate": 4.954692917702805e-05,
+      "loss": 0.3399,
+      "step": 9225
+    },
+    {
+      "epoch": 0.6784322376645342,
+      "grad_norm": 0.8276044726371765,
+      "learning_rate": 4.952636786489033e-05,
+      "loss": 0.3546,
+      "step": 9226
+    },
+    {
+      "epoch": 0.6785057724832708,
+      "grad_norm": 0.5701926946640015,
+      "learning_rate": 4.95058094156703e-05,
+      "loss": 0.3505,
+      "step": 9227
+    },
+    {
+      "epoch": 0.6785793073020076,
+      "grad_norm": 0.8430971503257751,
+      "learning_rate": 4.9485253830534084e-05,
+      "loss": 0.3466,
+      "step": 9228
+    },
+    {
+      "epoch": 0.6786528421207442,
+      "grad_norm": 0.6065070629119873,
+      "learning_rate": 4.946470111064754e-05,
+      "loss": 0.3595,
+      "step": 9229
+    },
+    {
+      "epoch": 0.6787263769394808,
+      "grad_norm": 0.6445738077163696,
+      "learning_rate": 4.9444151257176485e-05,
+      "loss": 0.3361,
+      "step": 9230
+    },
+    {
+      "epoch": 0.6787999117582175,
+      "grad_norm": 0.5739269852638245,
+      "learning_rate": 4.9423604271286564e-05,
+      "loss": 0.3424,
+      "step": 9231
+    },
+    {
+      "epoch": 0.6788734465769541,
+      "grad_norm": 0.7058626413345337,
+      "learning_rate": 4.9403060154143136e-05,
+      "loss": 0.3494,
+      "step": 9232
+    },
+    {
+      "epoch": 0.6789469813956909,
+      "grad_norm": 0.5888022184371948,
+      "learning_rate": 4.9382518906911535e-05,
+      "loss": 0.3665,
+      "step": 9233
+    },
+    {
+      "epoch": 0.6790205162144275,
+      "grad_norm": 0.5873419642448425,
+      "learning_rate": 4.936198053075692e-05,
+      "loss": 0.3642,
+      "step": 9234
+    },
+    {
+      "epoch": 0.6790940510331642,
+      "grad_norm": 0.6009774208068848,
+      "learning_rate": 4.934144502684418e-05,
+      "loss": 0.3409,
+      "step": 9235
+    },
+    {
+      "epoch": 0.6791675858519008,
+      "grad_norm": 0.505914032459259,
+      "learning_rate": 4.932091239633814e-05,
+      "loss": 0.3429,
+      "step": 9236
+    },
+    {
+      "epoch": 0.6792411206706376,
+      "grad_norm": 0.5633138418197632,
+      "learning_rate": 4.9300382640403465e-05,
+      "loss": 0.3587,
+      "step": 9237
+    },
+    {
+      "epoch": 0.6793146554893742,
+      "grad_norm": 0.6625150442123413,
+      "learning_rate": 4.927985576020459e-05,
+      "loss": 0.3622,
+      "step": 9238
+    },
+    {
+      "epoch": 0.6793881903081109,
+      "grad_norm": 0.6396831274032593,
+      "learning_rate": 4.925933175690578e-05,
+      "loss": 0.3808,
+      "step": 9239
+    },
+    {
+      "epoch": 0.6794617251268475,
+      "grad_norm": 0.635872483253479,
+      "learning_rate": 4.923881063167123e-05,
+      "loss": 0.3635,
+      "step": 9240
+    },
+    {
+      "epoch": 0.6795352599455843,
+      "grad_norm": 1.0092602968215942,
+      "learning_rate": 4.921829238566496e-05,
+      "loss": 0.3637,
+      "step": 9241
+    },
+    {
+      "epoch": 0.6796087947643209,
+      "grad_norm": 0.5368188619613647,
+      "learning_rate": 4.919777702005068e-05,
+      "loss": 0.37,
+      "step": 9242
+    },
+    {
+      "epoch": 0.6796823295830576,
+      "grad_norm": 0.8835824728012085,
+      "learning_rate": 4.9177264535992104e-05,
+      "loss": 0.373,
+      "step": 9243
+    },
+    {
+      "epoch": 0.6797558644017943,
+      "grad_norm": 0.5732256770133972,
+      "learning_rate": 4.915675493465277e-05,
+      "loss": 0.3588,
+      "step": 9244
+    },
+    {
+      "epoch": 0.6798293992205309,
+      "grad_norm": 1.058273196220398,
+      "learning_rate": 4.91362482171959e-05,
+      "loss": 0.3649,
+      "step": 9245
+    },
+    {
+      "epoch": 0.6799029340392676,
+      "grad_norm": 0.5095304846763611,
+      "learning_rate": 4.911574438478472e-05,
+      "loss": 0.3499,
+      "step": 9246
+    },
+    {
+      "epoch": 0.6799764688580042,
+      "grad_norm": 1.1720000505447388,
+      "learning_rate": 4.9095243438582264e-05,
+      "loss": 0.3624,
+      "step": 9247
+    },
+    {
+      "epoch": 0.680050003676741,
+      "grad_norm": 0.6370082497596741,
+      "learning_rate": 4.907474537975131e-05,
+      "loss": 0.3439,
+      "step": 9248
+    },
+    {
+      "epoch": 0.6801235384954776,
+      "grad_norm": 1.0352239608764648,
+      "learning_rate": 4.90542502094545e-05,
+      "loss": 0.3429,
+      "step": 9249
+    },
+    {
+      "epoch": 0.6801970733142143,
+      "grad_norm": 0.6835673451423645,
+      "learning_rate": 4.90337579288544e-05,
+      "loss": 0.3374,
+      "step": 9250
+    },
+    {
+      "epoch": 0.6802706081329509,
+      "grad_norm": 0.77922523021698,
+      "learning_rate": 4.901326853911336e-05,
+      "loss": 0.3631,
+      "step": 9251
+    },
+    {
+      "epoch": 0.6803441429516877,
+      "grad_norm": 1.002801775932312,
+      "learning_rate": 4.8992782041393504e-05,
+      "loss": 0.3685,
+      "step": 9252
+    },
+    {
+      "epoch": 0.6804176777704243,
+      "grad_norm": 0.5694381594657898,
+      "learning_rate": 4.8972298436856877e-05,
+      "loss": 0.3329,
+      "step": 9253
+    },
+    {
+      "epoch": 0.680491212589161,
+      "grad_norm": 0.7233249545097351,
+      "learning_rate": 4.895181772666537e-05,
+      "loss": 0.3524,
+      "step": 9254
+    },
+    {
+      "epoch": 0.6805647474078976,
+      "grad_norm": 0.8471522331237793,
+      "learning_rate": 4.89313399119806e-05,
+      "loss": 0.37,
+      "step": 9255
+    },
+    {
+      "epoch": 0.6806382822266344,
+      "grad_norm": 0.6720590591430664,
+      "learning_rate": 4.8910864993964135e-05,
+      "loss": 0.3575,
+      "step": 9256
+    },
+    {
+      "epoch": 0.680711817045371,
+      "grad_norm": 0.8570370674133301,
+      "learning_rate": 4.8890392973777354e-05,
+      "loss": 0.3498,
+      "step": 9257
+    },
+    {
+      "epoch": 0.6807853518641076,
+      "grad_norm": 1.0003255605697632,
+      "learning_rate": 4.886992385258138e-05,
+      "loss": 0.3587,
+      "step": 9258
+    },
+    {
+      "epoch": 0.6808588866828443,
+      "grad_norm": 0.503968358039856,
+      "learning_rate": 4.8849457631537285e-05,
+      "loss": 0.3658,
+      "step": 9259
+    },
+    {
+      "epoch": 0.680932421501581,
+      "grad_norm": 0.823877215385437,
+      "learning_rate": 4.8828994311805974e-05,
+      "loss": 0.3556,
+      "step": 9260
+    },
+    {
+      "epoch": 0.6810059563203177,
+      "grad_norm": 0.500329315662384,
+      "learning_rate": 4.8808533894548117e-05,
+      "loss": 0.357,
+      "step": 9261
+    },
+    {
+      "epoch": 0.6810794911390543,
+      "grad_norm": 0.5475634932518005,
+      "learning_rate": 4.87880763809242e-05,
+      "loss": 0.3444,
+      "step": 9262
+    },
+    {
+      "epoch": 0.681153025957791,
+      "grad_norm": 0.5239691734313965,
+      "learning_rate": 4.876762177209463e-05,
+      "loss": 0.3581,
+      "step": 9263
+    },
+    {
+      "epoch": 0.6812265607765277,
+      "grad_norm": 1.1009445190429688,
+      "learning_rate": 4.874717006921966e-05,
+      "loss": 0.3589,
+      "step": 9264
+    },
+    {
+      "epoch": 0.6813000955952644,
+      "grad_norm": 0.5671507716178894,
+      "learning_rate": 4.872672127345926e-05,
+      "loss": 0.3554,
+      "step": 9265
+    },
+    {
+      "epoch": 0.681373630414001,
+      "grad_norm": 1.347057819366455,
+      "learning_rate": 4.870627538597334e-05,
+      "loss": 0.3656,
+      "step": 9266
+    },
+    {
+      "epoch": 0.6814471652327377,
+      "grad_norm": 0.6511924862861633,
+      "learning_rate": 4.868583240792165e-05,
+      "loss": 0.413,
+      "step": 9267
+    },
+    {
+      "epoch": 0.6815207000514744,
+      "grad_norm": 0.8186678886413574,
+      "learning_rate": 4.8665392340463655e-05,
+      "loss": 0.362,
+      "step": 9268
+    },
+    {
+      "epoch": 0.6815942348702111,
+      "grad_norm": 1.061375617980957,
+      "learning_rate": 4.864495518475879e-05,
+      "loss": 0.3668,
+      "step": 9269
+    },
+    {
+      "epoch": 0.6816677696889477,
+      "grad_norm": 1.141419768333435,
+      "learning_rate": 4.86245209419663e-05,
+      "loss": 0.3326,
+      "step": 9270
+    },
+    {
+      "epoch": 0.6817413045076844,
+      "grad_norm": 0.9563266038894653,
+      "learning_rate": 4.860408961324516e-05,
+      "loss": 0.3689,
+      "step": 9271
+    },
+    {
+      "epoch": 0.6818148393264211,
+      "grad_norm": 1.1126505136489868,
+      "learning_rate": 4.858366119975434e-05,
+      "loss": 0.3564,
+      "step": 9272
+    },
+    {
+      "epoch": 0.6818883741451577,
+      "grad_norm": 1.198561191558838,
+      "learning_rate": 4.856323570265247e-05,
+      "loss": 0.3755,
+      "step": 9273
+    },
+    {
+      "epoch": 0.6819619089638944,
+      "grad_norm": 1.036733865737915,
+      "learning_rate": 4.854281312309821e-05,
+      "loss": 0.3687,
+      "step": 9274
+    },
+    {
+      "epoch": 0.682035443782631,
+      "grad_norm": 1.0437469482421875,
+      "learning_rate": 4.852239346224985e-05,
+      "loss": 0.3527,
+      "step": 9275
+    },
+    {
+      "epoch": 0.6821089786013678,
+      "grad_norm": 1.18838632106781,
+      "learning_rate": 4.850197672126565e-05,
+      "loss": 0.366,
+      "step": 9276
+    },
+    {
+      "epoch": 0.6821825134201044,
+      "grad_norm": 0.8843088746070862,
+      "learning_rate": 4.8481562901303733e-05,
+      "loss": 0.3451,
+      "step": 9277
+    },
+    {
+      "epoch": 0.6822560482388411,
+      "grad_norm": 0.7144618630409241,
+      "learning_rate": 4.846115200352188e-05,
+      "loss": 0.3572,
+      "step": 9278
+    },
+    {
+      "epoch": 0.6823295830575777,
+      "grad_norm": 1.268159031867981,
+      "learning_rate": 4.844074402907789e-05,
+      "loss": 0.3424,
+      "step": 9279
+    },
+    {
+      "epoch": 0.6824031178763145,
+      "grad_norm": 0.4581794738769531,
+      "learning_rate": 4.842033897912934e-05,
+      "loss": 0.3661,
+      "step": 9280
+    },
+    {
+      "epoch": 0.6824766526950511,
+      "grad_norm": 1.4515931606292725,
+      "learning_rate": 4.8399936854833574e-05,
+      "loss": 0.4272,
+      "step": 9281
+    },
+    {
+      "epoch": 0.6825501875137878,
+      "grad_norm": 0.8704625964164734,
+      "learning_rate": 4.837953765734784e-05,
+      "loss": 0.3762,
+      "step": 9282
+    },
+    {
+      "epoch": 0.6826237223325244,
+      "grad_norm": 1.4590128660202026,
+      "learning_rate": 4.8359141387829235e-05,
+      "loss": 0.3914,
+      "step": 9283
+    },
+    {
+      "epoch": 0.6826972571512612,
+      "grad_norm": 1.3135226964950562,
+      "learning_rate": 4.8338748047434634e-05,
+      "loss": 0.3871,
+      "step": 9284
+    },
+    {
+      "epoch": 0.6827707919699978,
+      "grad_norm": 1.1378141641616821,
+      "learning_rate": 4.831835763732073e-05,
+      "loss": 0.3476,
+      "step": 9285
+    },
+    {
+      "epoch": 0.6828443267887344,
+      "grad_norm": 1.2121801376342773,
+      "learning_rate": 4.829797015864412e-05,
+      "loss": 0.3546,
+      "step": 9286
+    },
+    {
+      "epoch": 0.6829178616074711,
+      "grad_norm": 0.8651596307754517,
+      "learning_rate": 4.827758561256125e-05,
+      "loss": 0.3703,
+      "step": 9287
+    },
+    {
+      "epoch": 0.6829913964262078,
+      "grad_norm": 1.1279910802841187,
+      "learning_rate": 4.8257204000228266e-05,
+      "loss": 0.3551,
+      "step": 9288
+    },
+    {
+      "epoch": 0.6830649312449445,
+      "grad_norm": 0.6935418844223022,
+      "learning_rate": 4.823682532280127e-05,
+      "loss": 0.3766,
+      "step": 9289
+    },
+    {
+      "epoch": 0.6831384660636811,
+      "grad_norm": 1.0293846130371094,
+      "learning_rate": 4.821644958143622e-05,
+      "loss": 0.3789,
+      "step": 9290
+    },
+    {
+      "epoch": 0.6832120008824178,
+      "grad_norm": 0.6278707385063171,
+      "learning_rate": 4.8196076777288754e-05,
+      "loss": 0.3701,
+      "step": 9291
+    },
+    {
+      "epoch": 0.6832855357011545,
+      "grad_norm": 1.112011432647705,
+      "learning_rate": 4.8175706911514486e-05,
+      "loss": 0.3417,
+      "step": 9292
+    },
+    {
+      "epoch": 0.6833590705198912,
+      "grad_norm": 0.9940866827964783,
+      "learning_rate": 4.815533998526884e-05,
+      "loss": 0.3802,
+      "step": 9293
+    },
+    {
+      "epoch": 0.6834326053386278,
+      "grad_norm": 0.6034048795700073,
+      "learning_rate": 4.813497599970701e-05,
+      "loss": 0.3525,
+      "step": 9294
+    },
+    {
+      "epoch": 0.6835061401573646,
+      "grad_norm": 0.6576797962188721,
+      "learning_rate": 4.811461495598409e-05,
+      "loss": 0.3634,
+      "step": 9295
+    },
+    {
+      "epoch": 0.6835796749761012,
+      "grad_norm": 0.6181974411010742,
+      "learning_rate": 4.809425685525495e-05,
+      "loss": 0.3584,
+      "step": 9296
+    },
+    {
+      "epoch": 0.6836532097948379,
+      "grad_norm": 0.670158326625824,
+      "learning_rate": 4.807390169867437e-05,
+      "loss": 0.3581,
+      "step": 9297
+    },
+    {
+      "epoch": 0.6837267446135745,
+      "grad_norm": 0.5181406140327454,
+      "learning_rate": 4.805354948739684e-05,
+      "loss": 0.387,
+      "step": 9298
+    },
+    {
+      "epoch": 0.6838002794323113,
+      "grad_norm": 0.5671868324279785,
+      "learning_rate": 4.803320022257681e-05,
+      "loss": 0.3646,
+      "step": 9299
+    },
+    {
+      "epoch": 0.6838738142510479,
+      "grad_norm": 0.5563853979110718,
+      "learning_rate": 4.801285390536855e-05,
+      "loss": 0.3384,
+      "step": 9300
+    },
+    {
+      "epoch": 0.6839473490697845,
+      "grad_norm": 0.5679677724838257,
+      "learning_rate": 4.799251053692604e-05,
+      "loss": 0.3572,
+      "step": 9301
+    },
+    {
+      "epoch": 0.6840208838885212,
+      "grad_norm": 0.5725491046905518,
+      "learning_rate": 4.797217011840322e-05,
+      "loss": 0.3644,
+      "step": 9302
+    },
+    {
+      "epoch": 0.6840944187072578,
+      "grad_norm": 0.778260350227356,
+      "learning_rate": 4.7951832650953844e-05,
+      "loss": 0.3785,
+      "step": 9303
+    },
+    {
+      "epoch": 0.6841679535259946,
+      "grad_norm": 0.6267194747924805,
+      "learning_rate": 4.7931498135731414e-05,
+      "loss": 0.363,
+      "step": 9304
+    },
+    {
+      "epoch": 0.6842414883447312,
+      "grad_norm": 1.066162109375,
+      "learning_rate": 4.791116657388937e-05,
+      "loss": 0.3684,
+      "step": 9305
+    },
+    {
+      "epoch": 0.6843150231634679,
+      "grad_norm": 0.5135527849197388,
+      "learning_rate": 4.7890837966580914e-05,
+      "loss": 0.3461,
+      "step": 9306
+    },
+    {
+      "epoch": 0.6843885579822045,
+      "grad_norm": 0.7222369313240051,
+      "learning_rate": 4.78705123149592e-05,
+      "loss": 0.3657,
+      "step": 9307
+    },
+    {
+      "epoch": 0.6844620928009413,
+      "grad_norm": 0.8169330954551697,
+      "learning_rate": 4.785018962017698e-05,
+      "loss": 0.3575,
+      "step": 9308
+    },
+    {
+      "epoch": 0.6845356276196779,
+      "grad_norm": 1.2818348407745361,
+      "learning_rate": 4.7829869883387035e-05,
+      "loss": 0.3719,
+      "step": 9309
+    },
+    {
+      "epoch": 0.6846091624384146,
+      "grad_norm": 0.589993953704834,
+      "learning_rate": 4.780955310574197e-05,
+      "loss": 0.3349,
+      "step": 9310
+    },
+    {
+      "epoch": 0.6846826972571513,
+      "grad_norm": 1.0324839353561401,
+      "learning_rate": 4.7789239288394085e-05,
+      "loss": 0.4063,
+      "step": 9311
+    },
+    {
+      "epoch": 0.684756232075888,
+      "grad_norm": 0.6056477427482605,
+      "learning_rate": 4.776892843249564e-05,
+      "loss": 0.3596,
+      "step": 9312
+    },
+    {
+      "epoch": 0.6848297668946246,
+      "grad_norm": 0.7869539856910706,
+      "learning_rate": 4.774862053919875e-05,
+      "loss": 0.3133,
+      "step": 9313
+    },
+    {
+      "epoch": 0.6849033017133612,
+      "grad_norm": 0.4401235282421112,
+      "learning_rate": 4.7728315609655206e-05,
+      "loss": 0.3619,
+      "step": 9314
+    },
+    {
+      "epoch": 0.684976836532098,
+      "grad_norm": 1.0432552099227905,
+      "learning_rate": 4.7708013645016755e-05,
+      "loss": 0.3642,
+      "step": 9315
+    },
+    {
+      "epoch": 0.6850503713508346,
+      "grad_norm": 0.7330151796340942,
+      "learning_rate": 4.7687714646435e-05,
+      "loss": 0.3825,
+      "step": 9316
+    },
+    {
+      "epoch": 0.6851239061695713,
+      "grad_norm": 0.4609141945838928,
+      "learning_rate": 4.7667418615061245e-05,
+      "loss": 0.337,
+      "step": 9317
+    },
+    {
+      "epoch": 0.6851974409883079,
+      "grad_norm": 0.5906832218170166,
+      "learning_rate": 4.7647125552046766e-05,
+      "loss": 0.361,
+      "step": 9318
+    },
+    {
+      "epoch": 0.6852709758070447,
+      "grad_norm": 0.5032422542572021,
+      "learning_rate": 4.762683545854254e-05,
+      "loss": 0.3699,
+      "step": 9319
+    },
+    {
+      "epoch": 0.6853445106257813,
+      "grad_norm": 0.6335810422897339,
+      "learning_rate": 4.760654833569951e-05,
+      "loss": 0.3388,
+      "step": 9320
+    },
+    {
+      "epoch": 0.685418045444518,
+      "grad_norm": 0.529408872127533,
+      "learning_rate": 4.758626418466832e-05,
+      "loss": 0.3354,
+      "step": 9321
+    },
+    {
+      "epoch": 0.6854915802632546,
+      "grad_norm": 0.4976716935634613,
+      "learning_rate": 4.756598300659954e-05,
+      "loss": 0.3391,
+      "step": 9322
+    },
+    {
+      "epoch": 0.6855651150819914,
+      "grad_norm": 0.4769490361213684,
+      "learning_rate": 4.7545704802643575e-05,
+      "loss": 0.3336,
+      "step": 9323
+    },
+    {
+      "epoch": 0.685638649900728,
+      "grad_norm": 0.5668980479240417,
+      "learning_rate": 4.752542957395055e-05,
+      "loss": 0.3645,
+      "step": 9324
+    },
+    {
+      "epoch": 0.6857121847194647,
+      "grad_norm": 0.5008077621459961,
+      "learning_rate": 4.750515732167054e-05,
+      "loss": 0.3623,
+      "step": 9325
+    },
+    {
+      "epoch": 0.6857857195382013,
+      "grad_norm": 0.4752247631549835,
+      "learning_rate": 4.748488804695346e-05,
+      "loss": 0.3581,
+      "step": 9326
+    },
+    {
+      "epoch": 0.6858592543569381,
+      "grad_norm": 0.5404130220413208,
+      "learning_rate": 4.74646217509489e-05,
+      "loss": 0.3418,
+      "step": 9327
+    },
+    {
+      "epoch": 0.6859327891756747,
+      "grad_norm": 0.47307151556015015,
+      "learning_rate": 4.7444358434806435e-05,
+      "loss": 0.3492,
+      "step": 9328
+    },
+    {
+      "epoch": 0.6860063239944113,
+      "grad_norm": 0.4652186930179596,
+      "learning_rate": 4.742409809967543e-05,
+      "loss": 0.3622,
+      "step": 9329
+    },
+    {
+      "epoch": 0.686079858813148,
+      "grad_norm": 0.4601180851459503,
+      "learning_rate": 4.740384074670513e-05,
+      "loss": 0.3513,
+      "step": 9330
+    },
+    {
+      "epoch": 0.6861533936318847,
+      "grad_norm": 0.5565637350082397,
+      "learning_rate": 4.738358637704443e-05,
+      "loss": 0.4034,
+      "step": 9331
+    },
+    {
+      "epoch": 0.6862269284506214,
+      "grad_norm": 0.39734140038490295,
+      "learning_rate": 4.7363334991842236e-05,
+      "loss": 0.3556,
+      "step": 9332
+    },
+    {
+      "epoch": 0.686300463269358,
+      "grad_norm": 0.48031216859817505,
+      "learning_rate": 4.734308659224725e-05,
+      "loss": 0.3692,
+      "step": 9333
+    },
+    {
+      "epoch": 0.6863739980880947,
+      "grad_norm": 0.5783748030662537,
+      "learning_rate": 4.7322841179407954e-05,
+      "loss": 0.3885,
+      "step": 9334
+    },
+    {
+      "epoch": 0.6864475329068314,
+      "grad_norm": 0.5167304277420044,
+      "learning_rate": 4.730259875447268e-05,
+      "loss": 0.3585,
+      "step": 9335
+    },
+    {
+      "epoch": 0.6865210677255681,
+      "grad_norm": 0.4420221447944641,
+      "learning_rate": 4.7282359318589665e-05,
+      "loss": 0.354,
+      "step": 9336
+    },
+    {
+      "epoch": 0.6865946025443047,
+      "grad_norm": 0.4705166220664978,
+      "learning_rate": 4.7262122872906825e-05,
+      "loss": 0.3439,
+      "step": 9337
+    },
+    {
+      "epoch": 0.6866681373630414,
+      "grad_norm": 0.617626428604126,
+      "learning_rate": 4.7241889418572026e-05,
+      "loss": 0.3403,
+      "step": 9338
+    },
+    {
+      "epoch": 0.6867416721817781,
+      "grad_norm": 0.6594932079315186,
+      "learning_rate": 4.722165895673297e-05,
+      "loss": 0.3588,
+      "step": 9339
+    },
+    {
+      "epoch": 0.6868152070005148,
+      "grad_norm": 0.6511331796646118,
+      "learning_rate": 4.72014314885371e-05,
+      "loss": 0.3524,
+      "step": 9340
+    },
+    {
+      "epoch": 0.6868887418192514,
+      "grad_norm": 0.4457583725452423,
+      "learning_rate": 4.718120701513174e-05,
+      "loss": 0.3753,
+      "step": 9341
+    },
+    {
+      "epoch": 0.686962276637988,
+      "grad_norm": 0.741662859916687,
+      "learning_rate": 4.71609855376641e-05,
+      "loss": 0.3373,
+      "step": 9342
+    },
+    {
+      "epoch": 0.6870358114567248,
+      "grad_norm": 0.5243037343025208,
+      "learning_rate": 4.714076705728113e-05,
+      "loss": 0.3394,
+      "step": 9343
+    },
+    {
+      "epoch": 0.6871093462754614,
+      "grad_norm": 0.6880627274513245,
+      "learning_rate": 4.712055157512959e-05,
+      "loss": 0.3567,
+      "step": 9344
+    },
+    {
+      "epoch": 0.6871828810941981,
+      "grad_norm": 0.661289632320404,
+      "learning_rate": 4.7100339092356184e-05,
+      "loss": 0.3696,
+      "step": 9345
+    },
+    {
+      "epoch": 0.6872564159129347,
+      "grad_norm": 0.5265193581581116,
+      "learning_rate": 4.7080129610107406e-05,
+      "loss": 0.3548,
+      "step": 9346
+    },
+    {
+      "epoch": 0.6873299507316715,
+      "grad_norm": 0.5044072866439819,
+      "learning_rate": 4.7059923129529474e-05,
+      "loss": 0.3406,
+      "step": 9347
+    },
+    {
+      "epoch": 0.6874034855504081,
+      "grad_norm": 0.7032466530799866,
+      "learning_rate": 4.7039719651768585e-05,
+      "loss": 0.3698,
+      "step": 9348
+    },
+    {
+      "epoch": 0.6874770203691448,
+      "grad_norm": 0.4626144468784332,
+      "learning_rate": 4.7019519177970714e-05,
+      "loss": 0.3469,
+      "step": 9349
+    },
+    {
+      "epoch": 0.6875505551878814,
+      "grad_norm": 0.6131424307823181,
+      "learning_rate": 4.6999321709281605e-05,
+      "loss": 0.3357,
+      "step": 9350
+    },
+    {
+      "epoch": 0.6876240900066182,
+      "grad_norm": 0.5976851582527161,
+      "learning_rate": 4.69791272468469e-05,
+      "loss": 0.3516,
+      "step": 9351
+    },
+    {
+      "epoch": 0.6876976248253548,
+      "grad_norm": 0.4462578296661377,
+      "learning_rate": 4.695893579181204e-05,
+      "loss": 0.3464,
+      "step": 9352
+    },
+    {
+      "epoch": 0.6877711596440915,
+      "grad_norm": 0.4681711494922638,
+      "learning_rate": 4.693874734532241e-05,
+      "loss": 0.3429,
+      "step": 9353
+    },
+    {
+      "epoch": 0.6878446944628281,
+      "grad_norm": 0.4431382119655609,
+      "learning_rate": 4.691856190852294e-05,
+      "loss": 0.3461,
+      "step": 9354
+    },
+    {
+      "epoch": 0.6879182292815649,
+      "grad_norm": 0.5711786150932312,
+      "learning_rate": 4.689837948255866e-05,
+      "loss": 0.3429,
+      "step": 9355
+    },
+    {
+      "epoch": 0.6879917641003015,
+      "grad_norm": 0.6051685214042664,
+      "learning_rate": 4.6878200068574386e-05,
+      "loss": 0.3385,
+      "step": 9356
+    },
+    {
+      "epoch": 0.6880652989190381,
+      "grad_norm": 0.5540654063224792,
+      "learning_rate": 4.685802366771461e-05,
+      "loss": 0.3462,
+      "step": 9357
+    },
+    {
+      "epoch": 0.6881388337377748,
+      "grad_norm": 0.5287437438964844,
+      "learning_rate": 4.683785028112383e-05,
+      "loss": 0.324,
+      "step": 9358
+    },
+    {
+      "epoch": 0.6882123685565115,
+      "grad_norm": 0.4439493417739868,
+      "learning_rate": 4.681767990994631e-05,
+      "loss": 0.3435,
+      "step": 9359
+    },
+    {
+      "epoch": 0.6882859033752482,
+      "grad_norm": 0.4821629524230957,
+      "learning_rate": 4.679751255532608e-05,
+      "loss": 0.3207,
+      "step": 9360
+    },
+    {
+      "epoch": 0.6883594381939848,
+      "grad_norm": 0.4405137002468109,
+      "learning_rate": 4.677734821840709e-05,
+      "loss": 0.3473,
+      "step": 9361
+    },
+    {
+      "epoch": 0.6884329730127215,
+      "grad_norm": 0.43476995825767517,
+      "learning_rate": 4.6757186900333114e-05,
+      "loss": 0.3699,
+      "step": 9362
+    },
+    {
+      "epoch": 0.6885065078314582,
+      "grad_norm": 0.4576699137687683,
+      "learning_rate": 4.673702860224765e-05,
+      "loss": 0.366,
+      "step": 9363
+    },
+    {
+      "epoch": 0.6885800426501949,
+      "grad_norm": 0.46221524477005005,
+      "learning_rate": 4.671687332529412e-05,
+      "loss": 0.333,
+      "step": 9364
+    },
+    {
+      "epoch": 0.6886535774689315,
+      "grad_norm": 0.5039784908294678,
+      "learning_rate": 4.669672107061581e-05,
+      "loss": 0.3479,
+      "step": 9365
+    },
+    {
+      "epoch": 0.6887271122876683,
+      "grad_norm": 0.48506271839141846,
+      "learning_rate": 4.6676571839355746e-05,
+      "loss": 0.3292,
+      "step": 9366
+    },
+    {
+      "epoch": 0.6888006471064049,
+      "grad_norm": 0.45630472898483276,
+      "learning_rate": 4.665642563265675e-05,
+      "loss": 0.344,
+      "step": 9367
+    },
+    {
+      "epoch": 0.6888741819251416,
+      "grad_norm": 0.4828987419605255,
+      "learning_rate": 4.663628245166159e-05,
+      "loss": 0.3392,
+      "step": 9368
+    },
+    {
+      "epoch": 0.6889477167438782,
+      "grad_norm": 0.5235334634780884,
+      "learning_rate": 4.661614229751285e-05,
+      "loss": 0.3345,
+      "step": 9369
+    },
+    {
+      "epoch": 0.6890212515626148,
+      "grad_norm": 0.49306684732437134,
+      "learning_rate": 4.659600517135282e-05,
+      "loss": 0.3614,
+      "step": 9370
+    },
+    {
+      "epoch": 0.6890947863813516,
+      "grad_norm": 0.5002691149711609,
+      "learning_rate": 4.657587107432374e-05,
+      "loss": 0.3522,
+      "step": 9371
+    },
+    {
+      "epoch": 0.6891683212000882,
+      "grad_norm": 0.47117549180984497,
+      "learning_rate": 4.655574000756766e-05,
+      "loss": 0.3315,
+      "step": 9372
+    },
+    {
+      "epoch": 0.6892418560188249,
+      "grad_norm": 0.4675918221473694,
+      "learning_rate": 4.653561197222638e-05,
+      "loss": 0.3366,
+      "step": 9373
+    },
+    {
+      "epoch": 0.6893153908375615,
+      "grad_norm": 0.4681224524974823,
+      "learning_rate": 4.65154869694416e-05,
+      "loss": 0.3244,
+      "step": 9374
+    },
+    {
+      "epoch": 0.6893889256562983,
+      "grad_norm": 0.46857091784477234,
+      "learning_rate": 4.649536500035489e-05,
+      "loss": 0.3422,
+      "step": 9375
+    },
+    {
+      "epoch": 0.6894624604750349,
+      "grad_norm": 0.47748643159866333,
+      "learning_rate": 4.6475246066107504e-05,
+      "loss": 0.3698,
+      "step": 9376
+    },
+    {
+      "epoch": 0.6895359952937716,
+      "grad_norm": 0.4441505968570709,
+      "learning_rate": 4.645513016784069e-05,
+      "loss": 0.3592,
+      "step": 9377
+    },
+    {
+      "epoch": 0.6896095301125083,
+      "grad_norm": 0.430897057056427,
+      "learning_rate": 4.643501730669536e-05,
+      "loss": 0.3245,
+      "step": 9378
+    },
+    {
+      "epoch": 0.689683064931245,
+      "grad_norm": 0.5488619208335876,
+      "learning_rate": 4.6414907483812406e-05,
+      "loss": 0.3357,
+      "step": 9379
+    },
+    {
+      "epoch": 0.6897565997499816,
+      "grad_norm": 0.47506463527679443,
+      "learning_rate": 4.639480070033243e-05,
+      "loss": 0.3474,
+      "step": 9380
+    },
+    {
+      "epoch": 0.6898301345687183,
+      "grad_norm": 0.6227326393127441,
+      "learning_rate": 4.637469695739592e-05,
+      "loss": 0.3287,
+      "step": 9381
+    },
+    {
+      "epoch": 0.689903669387455,
+      "grad_norm": 0.5568456053733826,
+      "learning_rate": 4.6354596256143225e-05,
+      "loss": 0.3484,
+      "step": 9382
+    },
+    {
+      "epoch": 0.6899772042061917,
+      "grad_norm": 0.6015021800994873,
+      "learning_rate": 4.633449859771441e-05,
+      "loss": 0.3211,
+      "step": 9383
+    },
+    {
+      "epoch": 0.6900507390249283,
+      "grad_norm": 0.7488493323326111,
+      "learning_rate": 4.631440398324948e-05,
+      "loss": 0.3357,
+      "step": 9384
+    },
+    {
+      "epoch": 0.6901242738436649,
+      "grad_norm": 0.4423559010028839,
+      "learning_rate": 4.6294312413888244e-05,
+      "loss": 0.3411,
+      "step": 9385
+    },
+    {
+      "epoch": 0.6901978086624017,
+      "grad_norm": 0.6635777950286865,
+      "learning_rate": 4.627422389077024e-05,
+      "loss": 0.3125,
+      "step": 9386
+    },
+    {
+      "epoch": 0.6902713434811383,
+      "grad_norm": 0.7283463478088379,
+      "learning_rate": 4.625413841503496e-05,
+      "loss": 0.3553,
+      "step": 9387
+    },
+    {
+      "epoch": 0.690344878299875,
+      "grad_norm": 0.4943459928035736,
+      "learning_rate": 4.623405598782171e-05,
+      "loss": 0.3362,
+      "step": 9388
+    },
+    {
+      "epoch": 0.6904184131186116,
+      "grad_norm": 0.7099090218544006,
+      "learning_rate": 4.621397661026955e-05,
+      "loss": 0.3383,
+      "step": 9389
+    },
+    {
+      "epoch": 0.6904919479373484,
+      "grad_norm": 0.47980210185050964,
+      "learning_rate": 4.6193900283517355e-05,
+      "loss": 0.3311,
+      "step": 9390
+    },
+    {
+      "epoch": 0.690565482756085,
+      "grad_norm": 1.04129159450531,
+      "learning_rate": 4.6173827008703916e-05,
+      "loss": 0.353,
+      "step": 9391
+    },
+    {
+      "epoch": 0.6906390175748217,
+      "grad_norm": 0.5095875859260559,
+      "learning_rate": 4.615375678696786e-05,
+      "loss": 0.3415,
+      "step": 9392
+    },
+    {
+      "epoch": 0.6907125523935583,
+      "grad_norm": 0.8061908483505249,
+      "learning_rate": 4.6133689619447505e-05,
+      "loss": 0.3288,
+      "step": 9393
+    },
+    {
+      "epoch": 0.6907860872122951,
+      "grad_norm": 0.5675178170204163,
+      "learning_rate": 4.6113625507281134e-05,
+      "loss": 0.3424,
+      "step": 9394
+    },
+    {
+      "epoch": 0.6908596220310317,
+      "grad_norm": 0.5159813761711121,
+      "learning_rate": 4.609356445160681e-05,
+      "loss": 0.339,
+      "step": 9395
+    },
+    {
+      "epoch": 0.6909331568497684,
+      "grad_norm": 0.5789747834205627,
+      "learning_rate": 4.6073506453562374e-05,
+      "loss": 0.3251,
+      "step": 9396
+    },
+    {
+      "epoch": 0.691006691668505,
+      "grad_norm": 0.4366048574447632,
+      "learning_rate": 4.605345151428556e-05,
+      "loss": 0.322,
+      "step": 9397
+    },
+    {
+      "epoch": 0.6910802264872417,
+      "grad_norm": 0.5155128240585327,
+      "learning_rate": 4.6033399634913945e-05,
+      "loss": 0.3531,
+      "step": 9398
+    },
+    {
+      "epoch": 0.6911537613059784,
+      "grad_norm": 0.5039729475975037,
+      "learning_rate": 4.601335081658482e-05,
+      "loss": 0.3333,
+      "step": 9399
+    },
+    {
+      "epoch": 0.691227296124715,
+      "grad_norm": 0.5889601707458496,
+      "learning_rate": 4.599330506043544e-05,
+      "loss": 0.3707,
+      "step": 9400
+    },
+    {
+      "epoch": 0.6913008309434517,
+      "grad_norm": 0.4136349856853485,
+      "learning_rate": 4.597326236760278e-05,
+      "loss": 0.3146,
+      "step": 9401
+    },
+    {
+      "epoch": 0.6913743657621884,
+      "grad_norm": 0.7263419032096863,
+      "learning_rate": 4.595322273922371e-05,
+      "loss": 0.3334,
+      "step": 9402
+    },
+    {
+      "epoch": 0.6914479005809251,
+      "grad_norm": 0.4345482289791107,
+      "learning_rate": 4.593318617643485e-05,
+      "loss": 0.3487,
+      "step": 9403
+    },
+    {
+      "epoch": 0.6915214353996617,
+      "grad_norm": 1.1234807968139648,
+      "learning_rate": 4.591315268037273e-05,
+      "loss": 0.3468,
+      "step": 9404
+    },
+    {
+      "epoch": 0.6915949702183984,
+      "grad_norm": 0.4996333420276642,
+      "learning_rate": 4.5893122252173705e-05,
+      "loss": 0.3526,
+      "step": 9405
+    },
+    {
+      "epoch": 0.6916685050371351,
+      "grad_norm": 0.8351877927780151,
+      "learning_rate": 4.5873094892973856e-05,
+      "loss": 0.342,
+      "step": 9406
+    },
+    {
+      "epoch": 0.6917420398558718,
+      "grad_norm": 0.5470802783966064,
+      "learning_rate": 4.585307060390919e-05,
+      "loss": 0.3388,
+      "step": 9407
+    },
+    {
+      "epoch": 0.6918155746746084,
+      "grad_norm": 0.9924602508544922,
+      "learning_rate": 4.583304938611553e-05,
+      "loss": 0.3379,
+      "step": 9408
+    },
+    {
+      "epoch": 0.6918891094933451,
+      "grad_norm": 0.5998031497001648,
+      "learning_rate": 4.5813031240728446e-05,
+      "loss": 0.323,
+      "step": 9409
+    },
+    {
+      "epoch": 0.6919626443120818,
+      "grad_norm": 0.7245424389839172,
+      "learning_rate": 4.5793016168883405e-05,
+      "loss": 0.3397,
+      "step": 9410
+    },
+    {
+      "epoch": 0.6920361791308185,
+      "grad_norm": 0.4622988998889923,
+      "learning_rate": 4.57730041717157e-05,
+      "loss": 0.3443,
+      "step": 9411
+    },
+    {
+      "epoch": 0.6921097139495551,
+      "grad_norm": 0.7246009707450867,
+      "learning_rate": 4.5752995250360496e-05,
+      "loss": 0.3479,
+      "step": 9412
+    },
+    {
+      "epoch": 0.6921832487682917,
+      "grad_norm": 0.5807271599769592,
+      "learning_rate": 4.573298940595259e-05,
+      "loss": 0.3447,
+      "step": 9413
+    },
+    {
+      "epoch": 0.6922567835870285,
+      "grad_norm": 0.6286473870277405,
+      "learning_rate": 4.5712986639626775e-05,
+      "loss": 0.3611,
+      "step": 9414
+    },
+    {
+      "epoch": 0.6923303184057651,
+      "grad_norm": 0.6472976803779602,
+      "learning_rate": 4.56929869525177e-05,
+      "loss": 0.3225,
+      "step": 9415
+    },
+    {
+      "epoch": 0.6924038532245018,
+      "grad_norm": 0.4533155858516693,
+      "learning_rate": 4.567299034575968e-05,
+      "loss": 0.3367,
+      "step": 9416
+    },
+    {
+      "epoch": 0.6924773880432384,
+      "grad_norm": 0.5698635578155518,
+      "learning_rate": 4.565299682048698e-05,
+      "loss": 0.3538,
+      "step": 9417
+    },
+    {
+      "epoch": 0.6925509228619752,
+      "grad_norm": 0.4261224567890167,
+      "learning_rate": 4.563300637783369e-05,
+      "loss": 0.3436,
+      "step": 9418
+    },
+    {
+      "epoch": 0.6926244576807118,
+      "grad_norm": 0.5545915365219116,
+      "learning_rate": 4.561301901893361e-05,
+      "loss": 0.3333,
+      "step": 9419
+    },
+    {
+      "epoch": 0.6926979924994485,
+      "grad_norm": 0.5355130434036255,
+      "learning_rate": 4.5593034744920505e-05,
+      "loss": 0.3631,
+      "step": 9420
+    },
+    {
+      "epoch": 0.6927715273181851,
+      "grad_norm": 0.41957980394363403,
+      "learning_rate": 4.5573053556927914e-05,
+      "loss": 0.3086,
+      "step": 9421
+    },
+    {
+      "epoch": 0.6928450621369219,
+      "grad_norm": 0.5959727764129639,
+      "learning_rate": 4.555307545608912e-05,
+      "loss": 0.3279,
+      "step": 9422
+    },
+    {
+      "epoch": 0.6929185969556585,
+      "grad_norm": 0.4844765365123749,
+      "learning_rate": 4.553310044353736e-05,
+      "loss": 0.343,
+      "step": 9423
+    },
+    {
+      "epoch": 0.6929921317743952,
+      "grad_norm": 0.6758743524551392,
+      "learning_rate": 4.551312852040566e-05,
+      "loss": 0.3265,
+      "step": 9424
+    },
+    {
+      "epoch": 0.6930656665931318,
+      "grad_norm": 0.5676774382591248,
+      "learning_rate": 4.5493159687826804e-05,
+      "loss": 0.3314,
+      "step": 9425
+    },
+    {
+      "epoch": 0.6931392014118685,
+      "grad_norm": 0.7141482830047607,
+      "learning_rate": 4.547319394693342e-05,
+      "loss": 0.3365,
+      "step": 9426
+    },
+    {
+      "epoch": 0.6932127362306052,
+      "grad_norm": 0.5780459642410278,
+      "learning_rate": 4.5453231298858023e-05,
+      "loss": 0.3602,
+      "step": 9427
+    },
+    {
+      "epoch": 0.6932862710493418,
+      "grad_norm": 0.9408038854598999,
+      "learning_rate": 4.543327174473294e-05,
+      "loss": 0.3507,
+      "step": 9428
+    },
+    {
+      "epoch": 0.6933598058680785,
+      "grad_norm": 0.41246497631073,
+      "learning_rate": 4.5413315285690236e-05,
+      "loss": 0.3474,
+      "step": 9429
+    },
+    {
+      "epoch": 0.6934333406868152,
+      "grad_norm": 0.8749226927757263,
+      "learning_rate": 4.539336192286191e-05,
+      "loss": 0.3539,
+      "step": 9430
+    },
+    {
+      "epoch": 0.6935068755055519,
+      "grad_norm": 0.6102595925331116,
+      "learning_rate": 4.537341165737975e-05,
+      "loss": 0.3327,
+      "step": 9431
+    },
+    {
+      "epoch": 0.6935804103242885,
+      "grad_norm": 0.6182093620300293,
+      "learning_rate": 4.53534644903753e-05,
+      "loss": 0.353,
+      "step": 9432
+    },
+    {
+      "epoch": 0.6936539451430253,
+      "grad_norm": 0.6934952735900879,
+      "learning_rate": 4.533352042298e-05,
+      "loss": 0.352,
+      "step": 9433
+    },
+    {
+      "epoch": 0.6937274799617619,
+      "grad_norm": 0.4989652931690216,
+      "learning_rate": 4.531357945632513e-05,
+      "loss": 0.3244,
+      "step": 9434
+    },
+    {
+      "epoch": 0.6938010147804986,
+      "grad_norm": 0.5424145460128784,
+      "learning_rate": 4.529364159154181e-05,
+      "loss": 0.3357,
+      "step": 9435
+    },
+    {
+      "epoch": 0.6938745495992352,
+      "grad_norm": 0.5628585815429688,
+      "learning_rate": 4.527370682976081e-05,
+      "loss": 0.3742,
+      "step": 9436
+    },
+    {
+      "epoch": 0.693948084417972,
+      "grad_norm": 0.4573827087879181,
+      "learning_rate": 4.52537751721129e-05,
+      "loss": 0.3466,
+      "step": 9437
+    },
+    {
+      "epoch": 0.6940216192367086,
+      "grad_norm": 0.6030045747756958,
+      "learning_rate": 4.523384661972868e-05,
+      "loss": 0.3251,
+      "step": 9438
+    },
+    {
+      "epoch": 0.6940951540554453,
+      "grad_norm": 0.47857609391212463,
+      "learning_rate": 4.5213921173738446e-05,
+      "loss": 0.3778,
+      "step": 9439
+    },
+    {
+      "epoch": 0.6941686888741819,
+      "grad_norm": 0.46830612421035767,
+      "learning_rate": 4.519399883527242e-05,
+      "loss": 0.3507,
+      "step": 9440
+    },
+    {
+      "epoch": 0.6942422236929185,
+      "grad_norm": 0.5878310799598694,
+      "learning_rate": 4.517407960546065e-05,
+      "loss": 0.3153,
+      "step": 9441
+    },
+    {
+      "epoch": 0.6943157585116553,
+      "grad_norm": 0.47249191999435425,
+      "learning_rate": 4.515416348543291e-05,
+      "loss": 0.3543,
+      "step": 9442
+    },
+    {
+      "epoch": 0.6943892933303919,
+      "grad_norm": 0.6770315766334534,
+      "learning_rate": 4.51342504763189e-05,
+      "loss": 0.3289,
+      "step": 9443
+    },
+    {
+      "epoch": 0.6944628281491286,
+      "grad_norm": 0.6150310039520264,
+      "learning_rate": 4.5114340579248135e-05,
+      "loss": 0.3533,
+      "step": 9444
+    },
+    {
+      "epoch": 0.6945363629678653,
+      "grad_norm": 0.41761118173599243,
+      "learning_rate": 4.5094433795349867e-05,
+      "loss": 0.3331,
+      "step": 9445
+    },
+    {
+      "epoch": 0.694609897786602,
+      "grad_norm": 0.4758119583129883,
+      "learning_rate": 4.5074530125753255e-05,
+      "loss": 0.3292,
+      "step": 9446
+    },
+    {
+      "epoch": 0.6946834326053386,
+      "grad_norm": 0.4661683142185211,
+      "learning_rate": 4.5054629571587306e-05,
+      "loss": 0.3562,
+      "step": 9447
+    },
+    {
+      "epoch": 0.6947569674240753,
+      "grad_norm": 0.6044057607650757,
+      "learning_rate": 4.5034732133980736e-05,
+      "loss": 0.3664,
+      "step": 9448
+    },
+    {
+      "epoch": 0.694830502242812,
+      "grad_norm": 0.5299770832061768,
+      "learning_rate": 4.501483781406214e-05,
+      "loss": 0.372,
+      "step": 9449
+    },
+    {
+      "epoch": 0.6949040370615487,
+      "grad_norm": 0.47968220710754395,
+      "learning_rate": 4.499494661295998e-05,
+      "loss": 0.3366,
+      "step": 9450
+    },
+    {
+      "epoch": 0.6949775718802853,
+      "grad_norm": 0.5320081114768982,
+      "learning_rate": 4.4975058531802515e-05,
+      "loss": 0.3329,
+      "step": 9451
+    },
+    {
+      "epoch": 0.695051106699022,
+      "grad_norm": 0.45316433906555176,
+      "learning_rate": 4.495517357171777e-05,
+      "loss": 0.3333,
+      "step": 9452
+    },
+    {
+      "epoch": 0.6951246415177587,
+      "grad_norm": 0.5618001222610474,
+      "learning_rate": 4.4935291733833674e-05,
+      "loss": 0.3414,
+      "step": 9453
+    },
+    {
+      "epoch": 0.6951981763364953,
+      "grad_norm": 0.5868658423423767,
+      "learning_rate": 4.4915413019277985e-05,
+      "loss": 0.3286,
+      "step": 9454
+    },
+    {
+      "epoch": 0.695271711155232,
+      "grad_norm": 0.5206711888313293,
+      "learning_rate": 4.4895537429178156e-05,
+      "loss": 0.3391,
+      "step": 9455
+    },
+    {
+      "epoch": 0.6953452459739686,
+      "grad_norm": 0.7400676608085632,
+      "learning_rate": 4.48756649646616e-05,
+      "loss": 0.3461,
+      "step": 9456
+    },
+    {
+      "epoch": 0.6954187807927054,
+      "grad_norm": 0.5749043226242065,
+      "learning_rate": 4.48557956268555e-05,
+      "loss": 0.329,
+      "step": 9457
+    },
+    {
+      "epoch": 0.695492315611442,
+      "grad_norm": 0.48962095379829407,
+      "learning_rate": 4.483592941688691e-05,
+      "loss": 0.3432,
+      "step": 9458
+    },
+    {
+      "epoch": 0.6955658504301787,
+      "grad_norm": 0.5153452157974243,
+      "learning_rate": 4.481606633588262e-05,
+      "loss": 0.3723,
+      "step": 9459
+    },
+    {
+      "epoch": 0.6956393852489153,
+      "grad_norm": 0.7513752579689026,
+      "learning_rate": 4.479620638496925e-05,
+      "loss": 0.3377,
+      "step": 9460
+    },
+    {
+      "epoch": 0.6957129200676521,
+      "grad_norm": 0.5370172262191772,
+      "learning_rate": 4.477634956527335e-05,
+      "loss": 0.3847,
+      "step": 9461
+    },
+    {
+      "epoch": 0.6957864548863887,
+      "grad_norm": 0.49301838874816895,
+      "learning_rate": 4.4756495877921146e-05,
+      "loss": 0.3183,
+      "step": 9462
+    },
+    {
+      "epoch": 0.6958599897051254,
+      "grad_norm": 0.4089479446411133,
+      "learning_rate": 4.4736645324038794e-05,
+      "loss": 0.3194,
+      "step": 9463
+    },
+    {
+      "epoch": 0.695933524523862,
+      "grad_norm": 0.5374164581298828,
+      "learning_rate": 4.471679790475227e-05,
+      "loss": 0.3548,
+      "step": 9464
+    },
+    {
+      "epoch": 0.6960070593425988,
+      "grad_norm": 0.48748213052749634,
+      "learning_rate": 4.4696953621187286e-05,
+      "loss": 0.3256,
+      "step": 9465
+    },
+    {
+      "epoch": 0.6960805941613354,
+      "grad_norm": 0.4621884226799011,
+      "learning_rate": 4.467711247446945e-05,
+      "loss": 0.3283,
+      "step": 9466
+    },
+    {
+      "epoch": 0.6961541289800721,
+      "grad_norm": 0.5193862318992615,
+      "learning_rate": 4.465727446572422e-05,
+      "loss": 0.3518,
+      "step": 9467
+    },
+    {
+      "epoch": 0.6962276637988087,
+      "grad_norm": 0.6988661289215088,
+      "learning_rate": 4.4637439596076744e-05,
+      "loss": 0.3548,
+      "step": 9468
+    },
+    {
+      "epoch": 0.6963011986175454,
+      "grad_norm": 0.659171462059021,
+      "learning_rate": 4.461760786665213e-05,
+      "loss": 0.3401,
+      "step": 9469
+    },
+    {
+      "epoch": 0.6963747334362821,
+      "grad_norm": 0.6651076078414917,
+      "learning_rate": 4.459777927857528e-05,
+      "loss": 0.3472,
+      "step": 9470
+    },
+    {
+      "epoch": 0.6964482682550187,
+      "grad_norm": 0.7161919474601746,
+      "learning_rate": 4.4577953832970856e-05,
+      "loss": 0.354,
+      "step": 9471
+    },
+    {
+      "epoch": 0.6965218030737554,
+      "grad_norm": 0.6348690986633301,
+      "learning_rate": 4.455813153096334e-05,
+      "loss": 0.342,
+      "step": 9472
+    },
+    {
+      "epoch": 0.6965953378924921,
+      "grad_norm": 0.8098137378692627,
+      "learning_rate": 4.4538312373677116e-05,
+      "loss": 0.3814,
+      "step": 9473
+    },
+    {
+      "epoch": 0.6966688727112288,
+      "grad_norm": 0.6382924318313599,
+      "learning_rate": 4.4518496362236384e-05,
+      "loss": 0.3507,
+      "step": 9474
+    },
+    {
+      "epoch": 0.6967424075299654,
+      "grad_norm": 0.5750030279159546,
+      "learning_rate": 4.449868349776506e-05,
+      "loss": 0.3441,
+      "step": 9475
+    },
+    {
+      "epoch": 0.6968159423487021,
+      "grad_norm": 0.5554981231689453,
+      "learning_rate": 4.4478873781386964e-05,
+      "loss": 0.3759,
+      "step": 9476
+    },
+    {
+      "epoch": 0.6968894771674388,
+      "grad_norm": 0.5480430722236633,
+      "learning_rate": 4.445906721422579e-05,
+      "loss": 0.3335,
+      "step": 9477
+    },
+    {
+      "epoch": 0.6969630119861755,
+      "grad_norm": 0.5272908210754395,
+      "learning_rate": 4.443926379740489e-05,
+      "loss": 0.3319,
+      "step": 9478
+    },
+    {
+      "epoch": 0.6970365468049121,
+      "grad_norm": 0.5328624844551086,
+      "learning_rate": 4.44194635320476e-05,
+      "loss": 0.3234,
+      "step": 9479
+    },
+    {
+      "epoch": 0.6971100816236488,
+      "grad_norm": 0.7616665959358215,
+      "learning_rate": 4.439966641927699e-05,
+      "loss": 0.323,
+      "step": 9480
+    },
+    {
+      "epoch": 0.6971836164423855,
+      "grad_norm": 0.5076166987419128,
+      "learning_rate": 4.4379872460216e-05,
+      "loss": 0.3247,
+      "step": 9481
+    },
+    {
+      "epoch": 0.6972571512611221,
+      "grad_norm": 0.7937231659889221,
+      "learning_rate": 4.436008165598736e-05,
+      "loss": 0.359,
+      "step": 9482
+    },
+    {
+      "epoch": 0.6973306860798588,
+      "grad_norm": 0.5043702721595764,
+      "learning_rate": 4.434029400771356e-05,
+      "loss": 0.3234,
+      "step": 9483
+    },
+    {
+      "epoch": 0.6974042208985954,
+      "grad_norm": 0.7981273531913757,
+      "learning_rate": 4.432050951651706e-05,
+      "loss": 0.332,
+      "step": 9484
+    },
+    {
+      "epoch": 0.6974777557173322,
+      "grad_norm": 0.6476008296012878,
+      "learning_rate": 4.430072818351999e-05,
+      "loss": 0.3401,
+      "step": 9485
+    },
+    {
+      "epoch": 0.6975512905360688,
+      "grad_norm": 0.6775811314582825,
+      "learning_rate": 4.42809500098444e-05,
+      "loss": 0.3648,
+      "step": 9486
+    },
+    {
+      "epoch": 0.6976248253548055,
+      "grad_norm": 0.7419989705085754,
+      "learning_rate": 4.426117499661217e-05,
+      "loss": 0.3417,
+      "step": 9487
+    },
+    {
+      "epoch": 0.6976983601735421,
+      "grad_norm": 0.5242445468902588,
+      "learning_rate": 4.424140314494487e-05,
+      "loss": 0.3408,
+      "step": 9488
+    },
+    {
+      "epoch": 0.6977718949922789,
+      "grad_norm": 0.5661635994911194,
+      "learning_rate": 4.422163445596403e-05,
+      "loss": 0.3326,
+      "step": 9489
+    },
+    {
+      "epoch": 0.6978454298110155,
+      "grad_norm": 0.726107656955719,
+      "learning_rate": 4.420186893079099e-05,
+      "loss": 0.3293,
+      "step": 9490
+    },
+    {
+      "epoch": 0.6979189646297522,
+      "grad_norm": 0.5808075070381165,
+      "learning_rate": 4.4182106570546785e-05,
+      "loss": 0.3587,
+      "step": 9491
+    },
+    {
+      "epoch": 0.6979924994484888,
+      "grad_norm": 0.4490525722503662,
+      "learning_rate": 4.416234737635241e-05,
+      "loss": 0.3435,
+      "step": 9492
+    },
+    {
+      "epoch": 0.6980660342672256,
+      "grad_norm": 0.5599502921104431,
+      "learning_rate": 4.414259134932862e-05,
+      "loss": 0.3321,
+      "step": 9493
+    },
+    {
+      "epoch": 0.6981395690859622,
+      "grad_norm": 0.5472418665885925,
+      "learning_rate": 4.412283849059606e-05,
+      "loss": 0.3369,
+      "step": 9494
+    },
+    {
+      "epoch": 0.6982131039046989,
+      "grad_norm": 0.49310481548309326,
+      "learning_rate": 4.4103088801275016e-05,
+      "loss": 0.3265,
+      "step": 9495
+    },
+    {
+      "epoch": 0.6982866387234355,
+      "grad_norm": 0.44733744859695435,
+      "learning_rate": 4.408334228248575e-05,
+      "loss": 0.3445,
+      "step": 9496
+    },
+    {
+      "epoch": 0.6983601735421722,
+      "grad_norm": 0.4592287838459015,
+      "learning_rate": 4.4063598935348374e-05,
+      "loss": 0.3307,
+      "step": 9497
+    },
+    {
+      "epoch": 0.6984337083609089,
+      "grad_norm": 0.4477024972438812,
+      "learning_rate": 4.404385876098264e-05,
+      "loss": 0.3566,
+      "step": 9498
+    },
+    {
+      "epoch": 0.6985072431796455,
+      "grad_norm": 0.4618452787399292,
+      "learning_rate": 4.402412176050831e-05,
+      "loss": 0.3395,
+      "step": 9499
+    },
+    {
+      "epoch": 0.6985807779983823,
+      "grad_norm": 0.7845603823661804,
+      "learning_rate": 4.400438793504489e-05,
+      "loss": 0.3395,
+      "step": 9500
+    },
+    {
+      "epoch": 0.6986543128171189,
+      "grad_norm": 0.7085475921630859,
+      "learning_rate": 4.398465728571165e-05,
+      "loss": 0.3289,
+      "step": 9501
+    },
+    {
+      "epoch": 0.6987278476358556,
+      "grad_norm": 0.627470076084137,
+      "learning_rate": 4.3964929813627766e-05,
+      "loss": 0.3529,
+      "step": 9502
+    },
+    {
+      "epoch": 0.6988013824545922,
+      "grad_norm": 0.761278510093689,
+      "learning_rate": 4.394520551991219e-05,
+      "loss": 0.3494,
+      "step": 9503
+    },
+    {
+      "epoch": 0.698874917273329,
+      "grad_norm": 0.6545718312263489,
+      "learning_rate": 4.392548440568376e-05,
+      "loss": 0.3495,
+      "step": 9504
+    },
+    {
+      "epoch": 0.6989484520920656,
+      "grad_norm": 0.9763182401657104,
+      "learning_rate": 4.3905766472061026e-05,
+      "loss": 0.3851,
+      "step": 9505
+    },
+    {
+      "epoch": 0.6990219869108023,
+      "grad_norm": 0.49489885568618774,
+      "learning_rate": 4.388605172016238e-05,
+      "loss": 0.3614,
+      "step": 9506
+    },
+    {
+      "epoch": 0.6990955217295389,
+      "grad_norm": 0.5143543481826782,
+      "learning_rate": 4.3866340151106144e-05,
+      "loss": 0.3493,
+      "step": 9507
+    },
+    {
+      "epoch": 0.6991690565482757,
+      "grad_norm": 0.506104588508606,
+      "learning_rate": 4.3846631766010295e-05,
+      "loss": 0.3376,
+      "step": 9508
+    },
+    {
+      "epoch": 0.6992425913670123,
+      "grad_norm": 0.6316178441047668,
+      "learning_rate": 4.3826926565992745e-05,
+      "loss": 0.3557,
+      "step": 9509
+    },
+    {
+      "epoch": 0.6993161261857489,
+      "grad_norm": 0.5602684617042542,
+      "learning_rate": 4.3807224552171246e-05,
+      "loss": 0.3374,
+      "step": 9510
+    },
+    {
+      "epoch": 0.6993896610044856,
+      "grad_norm": 0.5462616086006165,
+      "learning_rate": 4.378752572566324e-05,
+      "loss": 0.3575,
+      "step": 9511
+    },
+    {
+      "epoch": 0.6994631958232223,
+      "grad_norm": 0.46482038497924805,
+      "learning_rate": 4.3767830087586093e-05,
+      "loss": 0.3312,
+      "step": 9512
+    },
+    {
+      "epoch": 0.699536730641959,
+      "grad_norm": 0.5931126475334167,
+      "learning_rate": 4.3748137639056994e-05,
+      "loss": 0.3341,
+      "step": 9513
+    },
+    {
+      "epoch": 0.6996102654606956,
+      "grad_norm": 0.5368528962135315,
+      "learning_rate": 4.372844838119288e-05,
+      "loss": 0.3242,
+      "step": 9514
+    },
+    {
+      "epoch": 0.6996838002794323,
+      "grad_norm": 0.48513898253440857,
+      "learning_rate": 4.370876231511054e-05,
+      "loss": 0.3298,
+      "step": 9515
+    },
+    {
+      "epoch": 0.699757335098169,
+      "grad_norm": 0.4980549216270447,
+      "learning_rate": 4.368907944192662e-05,
+      "loss": 0.3515,
+      "step": 9516
+    },
+    {
+      "epoch": 0.6998308699169057,
+      "grad_norm": 0.5750094056129456,
+      "learning_rate": 4.3669399762757604e-05,
+      "loss": 0.3507,
+      "step": 9517
+    },
+    {
+      "epoch": 0.6999044047356423,
+      "grad_norm": 0.5303077697753906,
+      "learning_rate": 4.364972327871961e-05,
+      "loss": 0.338,
+      "step": 9518
+    },
+    {
+      "epoch": 0.699977939554379,
+      "grad_norm": 0.45432770252227783,
+      "learning_rate": 4.363004999092879e-05,
+      "loss": 0.3503,
+      "step": 9519
+    },
+    {
+      "epoch": 0.7000514743731157,
+      "grad_norm": 0.43629467487335205,
+      "learning_rate": 4.3610379900501056e-05,
+      "loss": 0.3449,
+      "step": 9520
+    },
+    {
+      "epoch": 0.7001250091918524,
+      "grad_norm": 0.7642336487770081,
+      "learning_rate": 4.359071300855204e-05,
+      "loss": 0.3635,
+      "step": 9521
+    },
+    {
+      "epoch": 0.700198544010589,
+      "grad_norm": 0.48572009801864624,
+      "learning_rate": 4.357104931619732e-05,
+      "loss": 0.3432,
+      "step": 9522
+    },
+    {
+      "epoch": 0.7002720788293257,
+      "grad_norm": 0.8065558075904846,
+      "learning_rate": 4.355138882455228e-05,
+      "loss": 0.325,
+      "step": 9523
+    },
+    {
+      "epoch": 0.7003456136480624,
+      "grad_norm": 0.8830564618110657,
+      "learning_rate": 4.353173153473199e-05,
+      "loss": 0.3427,
+      "step": 9524
+    },
+    {
+      "epoch": 0.700419148466799,
+      "grad_norm": 0.5497070550918579,
+      "learning_rate": 4.351207744785148e-05,
+      "loss": 0.3421,
+      "step": 9525
+    },
+    {
+      "epoch": 0.7004926832855357,
+      "grad_norm": 0.6770111918449402,
+      "learning_rate": 4.349242656502559e-05,
+      "loss": 0.3526,
+      "step": 9526
+    },
+    {
+      "epoch": 0.7005662181042723,
+      "grad_norm": 0.5949457287788391,
+      "learning_rate": 4.347277888736887e-05,
+      "loss": 0.3348,
+      "step": 9527
+    },
+    {
+      "epoch": 0.7006397529230091,
+      "grad_norm": 0.5866996645927429,
+      "learning_rate": 4.3453134415995775e-05,
+      "loss": 0.3655,
+      "step": 9528
+    },
+    {
+      "epoch": 0.7007132877417457,
+      "grad_norm": 0.4873014986515045,
+      "learning_rate": 4.343349315202062e-05,
+      "loss": 0.3745,
+      "step": 9529
+    },
+    {
+      "epoch": 0.7007868225604824,
+      "grad_norm": 0.48361632227897644,
+      "learning_rate": 4.3413855096557424e-05,
+      "loss": 0.3712,
+      "step": 9530
+    },
+    {
+      "epoch": 0.700860357379219,
+      "grad_norm": 0.532563328742981,
+      "learning_rate": 4.339422025072005e-05,
+      "loss": 0.3723,
+      "step": 9531
+    },
+    {
+      "epoch": 0.7009338921979558,
+      "grad_norm": 0.6483709216117859,
+      "learning_rate": 4.337458861562225e-05,
+      "loss": 0.3548,
+      "step": 9532
+    },
+    {
+      "epoch": 0.7010074270166924,
+      "grad_norm": 0.48652955889701843,
+      "learning_rate": 4.335496019237757e-05,
+      "loss": 0.3351,
+      "step": 9533
+    },
+    {
+      "epoch": 0.7010809618354291,
+      "grad_norm": 0.6242638230323792,
+      "learning_rate": 4.333533498209929e-05,
+      "loss": 0.3387,
+      "step": 9534
+    },
+    {
+      "epoch": 0.7011544966541657,
+      "grad_norm": 0.629062831401825,
+      "learning_rate": 4.331571298590061e-05,
+      "loss": 0.3511,
+      "step": 9535
+    },
+    {
+      "epoch": 0.7012280314729025,
+      "grad_norm": 0.5141100287437439,
+      "learning_rate": 4.329609420489456e-05,
+      "loss": 0.354,
+      "step": 9536
+    },
+    {
+      "epoch": 0.7013015662916391,
+      "grad_norm": 0.6554779410362244,
+      "learning_rate": 4.327647864019383e-05,
+      "loss": 0.3448,
+      "step": 9537
+    },
+    {
+      "epoch": 0.7013751011103757,
+      "grad_norm": 0.7213953733444214,
+      "learning_rate": 4.325686629291111e-05,
+      "loss": 0.3662,
+      "step": 9538
+    },
+    {
+      "epoch": 0.7014486359291124,
+      "grad_norm": 0.5742596387863159,
+      "learning_rate": 4.3237257164158816e-05,
+      "loss": 0.3595,
+      "step": 9539
+    },
+    {
+      "epoch": 0.7015221707478491,
+      "grad_norm": 0.6842775940895081,
+      "learning_rate": 4.3217651255049275e-05,
+      "loss": 0.3423,
+      "step": 9540
+    },
+    {
+      "epoch": 0.7015957055665858,
+      "grad_norm": 0.44120511412620544,
+      "learning_rate": 4.3198048566694397e-05,
+      "loss": 0.3678,
+      "step": 9541
+    },
+    {
+      "epoch": 0.7016692403853224,
+      "grad_norm": 0.7242489457130432,
+      "learning_rate": 4.317844910020615e-05,
+      "loss": 0.3621,
+      "step": 9542
+    },
+    {
+      "epoch": 0.7017427752040591,
+      "grad_norm": 0.481406569480896,
+      "learning_rate": 4.3158852856696285e-05,
+      "loss": 0.3617,
+      "step": 9543
+    },
+    {
+      "epoch": 0.7018163100227958,
+      "grad_norm": 0.47162774205207825,
+      "learning_rate": 4.313925983727622e-05,
+      "loss": 0.3773,
+      "step": 9544
+    },
+    {
+      "epoch": 0.7018898448415325,
+      "grad_norm": 0.4847866892814636,
+      "learning_rate": 4.311967004305736e-05,
+      "loss": 0.3539,
+      "step": 9545
+    },
+    {
+      "epoch": 0.7019633796602691,
+      "grad_norm": 0.5897793769836426,
+      "learning_rate": 4.3100083475150875e-05,
+      "loss": 0.3644,
+      "step": 9546
+    },
+    {
+      "epoch": 0.7020369144790058,
+      "grad_norm": 0.4971356987953186,
+      "learning_rate": 4.308050013466767e-05,
+      "loss": 0.3467,
+      "step": 9547
+    },
+    {
+      "epoch": 0.7021104492977425,
+      "grad_norm": 0.5718491673469543,
+      "learning_rate": 4.3060920022718556e-05,
+      "loss": 0.3738,
+      "step": 9548
+    },
+    {
+      "epoch": 0.7021839841164792,
+      "grad_norm": 0.46180734038352966,
+      "learning_rate": 4.304134314041419e-05,
+      "loss": 0.3527,
+      "step": 9549
+    },
+    {
+      "epoch": 0.7022575189352158,
+      "grad_norm": 0.4889228940010071,
+      "learning_rate": 4.302176948886493e-05,
+      "loss": 0.381,
+      "step": 9550
+    },
+    {
+      "epoch": 0.7023310537539525,
+      "grad_norm": 0.5249632596969604,
+      "learning_rate": 4.300219906918101e-05,
+      "loss": 0.3478,
+      "step": 9551
+    },
+    {
+      "epoch": 0.7024045885726892,
+      "grad_norm": 0.4337559640407562,
+      "learning_rate": 4.298263188247257e-05,
+      "loss": 0.3691,
+      "step": 9552
+    },
+    {
+      "epoch": 0.7024781233914258,
+      "grad_norm": 0.7319900393486023,
+      "learning_rate": 4.296306792984941e-05,
+      "loss": 0.3517,
+      "step": 9553
+    },
+    {
+      "epoch": 0.7025516582101625,
+      "grad_norm": 0.47767457365989685,
+      "learning_rate": 4.29435072124212e-05,
+      "loss": 0.3376,
+      "step": 9554
+    },
+    {
+      "epoch": 0.7026251930288991,
+      "grad_norm": 0.7217221260070801,
+      "learning_rate": 4.2923949731297474e-05,
+      "loss": 0.3686,
+      "step": 9555
+    },
+    {
+      "epoch": 0.7026987278476359,
+      "grad_norm": 0.47913384437561035,
+      "learning_rate": 4.290439548758759e-05,
+      "loss": 0.3519,
+      "step": 9556
+    },
+    {
+      "epoch": 0.7027722626663725,
+      "grad_norm": 0.6478757858276367,
+      "learning_rate": 4.288484448240061e-05,
+      "loss": 0.3649,
+      "step": 9557
+    },
+    {
+      "epoch": 0.7028457974851092,
+      "grad_norm": 0.7090504765510559,
+      "learning_rate": 4.286529671684553e-05,
+      "loss": 0.3595,
+      "step": 9558
+    },
+    {
+      "epoch": 0.7029193323038458,
+      "grad_norm": 0.48507583141326904,
+      "learning_rate": 4.284575219203115e-05,
+      "loss": 0.3597,
+      "step": 9559
+    },
+    {
+      "epoch": 0.7029928671225826,
+      "grad_norm": 0.777455747127533,
+      "learning_rate": 4.282621090906599e-05,
+      "loss": 0.3543,
+      "step": 9560
+    },
+    {
+      "epoch": 0.7030664019413192,
+      "grad_norm": 0.4590515196323395,
+      "learning_rate": 4.280667286905848e-05,
+      "loss": 0.3892,
+      "step": 9561
+    },
+    {
+      "epoch": 0.7031399367600559,
+      "grad_norm": 0.6535444259643555,
+      "learning_rate": 4.278713807311685e-05,
+      "loss": 0.359,
+      "step": 9562
+    },
+    {
+      "epoch": 0.7032134715787925,
+      "grad_norm": 0.42179903388023376,
+      "learning_rate": 4.2767606522349155e-05,
+      "loss": 0.3342,
+      "step": 9563
+    },
+    {
+      "epoch": 0.7032870063975293,
+      "grad_norm": 0.5287411212921143,
+      "learning_rate": 4.274807821786323e-05,
+      "loss": 0.3811,
+      "step": 9564
+    },
+    {
+      "epoch": 0.7033605412162659,
+      "grad_norm": 0.5788407325744629,
+      "learning_rate": 4.2728553160766685e-05,
+      "loss": 0.353,
+      "step": 9565
+    },
+    {
+      "epoch": 0.7034340760350025,
+      "grad_norm": 0.5032828450202942,
+      "learning_rate": 4.2709031352167086e-05,
+      "loss": 0.3655,
+      "step": 9566
+    },
+    {
+      "epoch": 0.7035076108537393,
+      "grad_norm": 0.5586323142051697,
+      "learning_rate": 4.268951279317165e-05,
+      "loss": 0.3636,
+      "step": 9567
+    },
+    {
+      "epoch": 0.7035811456724759,
+      "grad_norm": 0.914899468421936,
+      "learning_rate": 4.2669997484887536e-05,
+      "loss": 0.3562,
+      "step": 9568
+    },
+    {
+      "epoch": 0.7036546804912126,
+      "grad_norm": 0.418297678232193,
+      "learning_rate": 4.2650485428421696e-05,
+      "loss": 0.358,
+      "step": 9569
+    },
+    {
+      "epoch": 0.7037282153099492,
+      "grad_norm": 1.079282283782959,
+      "learning_rate": 4.2630976624880817e-05,
+      "loss": 0.3725,
+      "step": 9570
+    },
+    {
+      "epoch": 0.703801750128686,
+      "grad_norm": 0.5021114349365234,
+      "learning_rate": 4.2611471075371476e-05,
+      "loss": 0.3636,
+      "step": 9571
+    },
+    {
+      "epoch": 0.7038752849474226,
+      "grad_norm": 0.8740999698638916,
+      "learning_rate": 4.2591968781000116e-05,
+      "loss": 0.3545,
+      "step": 9572
+    },
+    {
+      "epoch": 0.7039488197661593,
+      "grad_norm": 0.5784381628036499,
+      "learning_rate": 4.2572469742872826e-05,
+      "loss": 0.3621,
+      "step": 9573
+    },
+    {
+      "epoch": 0.7040223545848959,
+      "grad_norm": 0.640758752822876,
+      "learning_rate": 4.2552973962095655e-05,
+      "loss": 0.3731,
+      "step": 9574
+    },
+    {
+      "epoch": 0.7040958894036327,
+      "grad_norm": 0.7233843803405762,
+      "learning_rate": 4.253348143977446e-05,
+      "loss": 0.3875,
+      "step": 9575
+    },
+    {
+      "epoch": 0.7041694242223693,
+      "grad_norm": 0.5189794898033142,
+      "learning_rate": 4.251399217701485e-05,
+      "loss": 0.4029,
+      "step": 9576
+    },
+    {
+      "epoch": 0.704242959041106,
+      "grad_norm": 0.5331882238388062,
+      "learning_rate": 4.249450617492223e-05,
+      "loss": 0.3836,
+      "step": 9577
+    },
+    {
+      "epoch": 0.7043164938598426,
+      "grad_norm": 0.6859387159347534,
+      "learning_rate": 4.247502343460191e-05,
+      "loss": 0.3405,
+      "step": 9578
+    },
+    {
+      "epoch": 0.7043900286785794,
+      "grad_norm": 0.6045767664909363,
+      "learning_rate": 4.2455543957159005e-05,
+      "loss": 0.3479,
+      "step": 9579
+    },
+    {
+      "epoch": 0.704463563497316,
+      "grad_norm": 0.6557610630989075,
+      "learning_rate": 4.243606774369835e-05,
+      "loss": 0.3317,
+      "step": 9580
+    },
+    {
+      "epoch": 0.7045370983160526,
+      "grad_norm": 0.5141816735267639,
+      "learning_rate": 4.241659479532467e-05,
+      "loss": 0.3513,
+      "step": 9581
+    },
+    {
+      "epoch": 0.7046106331347893,
+      "grad_norm": 0.7744680047035217,
+      "learning_rate": 4.2397125113142544e-05,
+      "loss": 0.3358,
+      "step": 9582
+    },
+    {
+      "epoch": 0.704684167953526,
+      "grad_norm": 0.5384450554847717,
+      "learning_rate": 4.237765869825624e-05,
+      "loss": 0.3613,
+      "step": 9583
+    },
+    {
+      "epoch": 0.7047577027722627,
+      "grad_norm": 0.718885600566864,
+      "learning_rate": 4.2358195551769944e-05,
+      "loss": 0.3655,
+      "step": 9584
+    },
+    {
+      "epoch": 0.7048312375909993,
+      "grad_norm": 0.5958392024040222,
+      "learning_rate": 4.233873567478762e-05,
+      "loss": 0.3517,
+      "step": 9585
+    },
+    {
+      "epoch": 0.704904772409736,
+      "grad_norm": 0.8572297692298889,
+      "learning_rate": 4.23192790684131e-05,
+      "loss": 0.3495,
+      "step": 9586
+    },
+    {
+      "epoch": 0.7049783072284727,
+      "grad_norm": 0.7354368567466736,
+      "learning_rate": 4.2299825733749954e-05,
+      "loss": 0.3473,
+      "step": 9587
+    },
+    {
+      "epoch": 0.7050518420472094,
+      "grad_norm": 0.749377965927124,
+      "learning_rate": 4.228037567190154e-05,
+      "loss": 0.3499,
+      "step": 9588
+    },
+    {
+      "epoch": 0.705125376865946,
+      "grad_norm": 0.8964629769325256,
+      "learning_rate": 4.2260928883971163e-05,
+      "loss": 0.3499,
+      "step": 9589
+    },
+    {
+      "epoch": 0.7051989116846827,
+      "grad_norm": 0.5141761302947998,
+      "learning_rate": 4.224148537106181e-05,
+      "loss": 0.3452,
+      "step": 9590
+    },
+    {
+      "epoch": 0.7052724465034194,
+      "grad_norm": 0.7987736463546753,
+      "learning_rate": 4.222204513427636e-05,
+      "loss": 0.3544,
+      "step": 9591
+    },
+    {
+      "epoch": 0.7053459813221561,
+      "grad_norm": 0.42904528975486755,
+      "learning_rate": 4.2202608174717517e-05,
+      "loss": 0.3465,
+      "step": 9592
+    },
+    {
+      "epoch": 0.7054195161408927,
+      "grad_norm": 0.7566087245941162,
+      "learning_rate": 4.21831744934877e-05,
+      "loss": 0.3416,
+      "step": 9593
+    },
+    {
+      "epoch": 0.7054930509596293,
+      "grad_norm": 0.610811173915863,
+      "learning_rate": 4.216374409168924e-05,
+      "loss": 0.3219,
+      "step": 9594
+    },
+    {
+      "epoch": 0.7055665857783661,
+      "grad_norm": 0.5245875716209412,
+      "learning_rate": 4.214431697042429e-05,
+      "loss": 0.3615,
+      "step": 9595
+    },
+    {
+      "epoch": 0.7056401205971027,
+      "grad_norm": 0.6885425448417664,
+      "learning_rate": 4.212489313079471e-05,
+      "loss": 0.3513,
+      "step": 9596
+    },
+    {
+      "epoch": 0.7057136554158394,
+      "grad_norm": 0.635259211063385,
+      "learning_rate": 4.210547257390226e-05,
+      "loss": 0.3348,
+      "step": 9597
+    },
+    {
+      "epoch": 0.705787190234576,
+      "grad_norm": 0.5152688026428223,
+      "learning_rate": 4.2086055300848505e-05,
+      "loss": 0.3462,
+      "step": 9598
+    },
+    {
+      "epoch": 0.7058607250533128,
+      "grad_norm": 0.506067156791687,
+      "learning_rate": 4.206664131273489e-05,
+      "loss": 0.3579,
+      "step": 9599
+    },
+    {
+      "epoch": 0.7059342598720494,
+      "grad_norm": 0.40938112139701843,
+      "learning_rate": 4.2047230610662435e-05,
+      "loss": 0.3455,
+      "step": 9600
+    },
+    {
+      "epoch": 0.7060077946907861,
+      "grad_norm": 0.4592011570930481,
+      "learning_rate": 4.202782319573223e-05,
+      "loss": 0.3313,
+      "step": 9601
+    },
+    {
+      "epoch": 0.7060813295095227,
+      "grad_norm": 0.49410727620124817,
+      "learning_rate": 4.200841906904511e-05,
+      "loss": 0.3431,
+      "step": 9602
+    },
+    {
+      "epoch": 0.7061548643282595,
+      "grad_norm": 0.4195888936519623,
+      "learning_rate": 4.1989018231701624e-05,
+      "loss": 0.329,
+      "step": 9603
+    },
+    {
+      "epoch": 0.7062283991469961,
+      "grad_norm": 0.520599365234375,
+      "learning_rate": 4.1969620684802245e-05,
+      "loss": 0.3459,
+      "step": 9604
+    },
+    {
+      "epoch": 0.7063019339657328,
+      "grad_norm": 0.4448087215423584,
+      "learning_rate": 4.195022642944726e-05,
+      "loss": 0.3574,
+      "step": 9605
+    },
+    {
+      "epoch": 0.7063754687844694,
+      "grad_norm": 0.4515336751937866,
+      "learning_rate": 4.193083546673666e-05,
+      "loss": 0.3665,
+      "step": 9606
+    },
+    {
+      "epoch": 0.7064490036032062,
+      "grad_norm": 0.439840167760849,
+      "learning_rate": 4.191144779777035e-05,
+      "loss": 0.3453,
+      "step": 9607
+    },
+    {
+      "epoch": 0.7065225384219428,
+      "grad_norm": 0.5253122448921204,
+      "learning_rate": 4.189206342364803e-05,
+      "loss": 0.3642,
+      "step": 9608
+    },
+    {
+      "epoch": 0.7065960732406794,
+      "grad_norm": 0.477455735206604,
+      "learning_rate": 4.1872682345469225e-05,
+      "loss": 0.3529,
+      "step": 9609
+    },
+    {
+      "epoch": 0.7066696080594161,
+      "grad_norm": 0.5238417387008667,
+      "learning_rate": 4.1853304564333227e-05,
+      "loss": 0.3616,
+      "step": 9610
+    },
+    {
+      "epoch": 0.7067431428781528,
+      "grad_norm": 0.6698228716850281,
+      "learning_rate": 4.1833930081339125e-05,
+      "loss": 0.3461,
+      "step": 9611
+    },
+    {
+      "epoch": 0.7068166776968895,
+      "grad_norm": 0.39860910177230835,
+      "learning_rate": 4.181455889758592e-05,
+      "loss": 0.3358,
+      "step": 9612
+    },
+    {
+      "epoch": 0.7068902125156261,
+      "grad_norm": 0.7711341381072998,
+      "learning_rate": 4.179519101417231e-05,
+      "loss": 0.4211,
+      "step": 9613
+    },
+    {
+      "epoch": 0.7069637473343628,
+      "grad_norm": 0.5313601493835449,
+      "learning_rate": 4.177582643219689e-05,
+      "loss": 0.3489,
+      "step": 9614
+    },
+    {
+      "epoch": 0.7070372821530995,
+      "grad_norm": 0.7458357214927673,
+      "learning_rate": 4.1756465152758086e-05,
+      "loss": 0.3681,
+      "step": 9615
+    },
+    {
+      "epoch": 0.7071108169718362,
+      "grad_norm": 0.5246988534927368,
+      "learning_rate": 4.1737107176953995e-05,
+      "loss": 0.367,
+      "step": 9616
+    },
+    {
+      "epoch": 0.7071843517905728,
+      "grad_norm": 0.775169849395752,
+      "learning_rate": 4.171775250588268e-05,
+      "loss": 0.3664,
+      "step": 9617
+    },
+    {
+      "epoch": 0.7072578866093095,
+      "grad_norm": 0.8492763042449951,
+      "learning_rate": 4.1698401140641984e-05,
+      "loss": 0.3695,
+      "step": 9618
+    },
+    {
+      "epoch": 0.7073314214280462,
+      "grad_norm": 0.5910110473632812,
+      "learning_rate": 4.1679053082329465e-05,
+      "loss": 0.3854,
+      "step": 9619
+    },
+    {
+      "epoch": 0.7074049562467829,
+      "grad_norm": 0.6654065847396851,
+      "learning_rate": 4.165970833204261e-05,
+      "loss": 0.3609,
+      "step": 9620
+    },
+    {
+      "epoch": 0.7074784910655195,
+      "grad_norm": 0.5381011962890625,
+      "learning_rate": 4.164036689087867e-05,
+      "loss": 0.3655,
+      "step": 9621
+    },
+    {
+      "epoch": 0.7075520258842561,
+      "grad_norm": 0.631980299949646,
+      "learning_rate": 4.1621028759934775e-05,
+      "loss": 0.3629,
+      "step": 9622
+    },
+    {
+      "epoch": 0.7076255607029929,
+      "grad_norm": 0.7260986566543579,
+      "learning_rate": 4.160169394030767e-05,
+      "loss": 0.3558,
+      "step": 9623
+    },
+    {
+      "epoch": 0.7076990955217295,
+      "grad_norm": 0.9518898725509644,
+      "learning_rate": 4.158236243309411e-05,
+      "loss": 0.357,
+      "step": 9624
+    },
+    {
+      "epoch": 0.7077726303404662,
+      "grad_norm": 0.43390101194381714,
+      "learning_rate": 4.156303423939064e-05,
+      "loss": 0.3469,
+      "step": 9625
+    },
+    {
+      "epoch": 0.7078461651592028,
+      "grad_norm": 0.8384736180305481,
+      "learning_rate": 4.154370936029351e-05,
+      "loss": 0.3326,
+      "step": 9626
+    },
+    {
+      "epoch": 0.7079196999779396,
+      "grad_norm": 0.4505997896194458,
+      "learning_rate": 4.1524387796898866e-05,
+      "loss": 0.3516,
+      "step": 9627
+    },
+    {
+      "epoch": 0.7079932347966762,
+      "grad_norm": 0.852266252040863,
+      "learning_rate": 4.150506955030271e-05,
+      "loss": 0.3758,
+      "step": 9628
+    },
+    {
+      "epoch": 0.7080667696154129,
+      "grad_norm": 0.6536331176757812,
+      "learning_rate": 4.1485754621600694e-05,
+      "loss": 0.3525,
+      "step": 9629
+    },
+    {
+      "epoch": 0.7081403044341495,
+      "grad_norm": 1.780083179473877,
+      "learning_rate": 4.1466443011888423e-05,
+      "loss": 0.3451,
+      "step": 9630
+    },
+    {
+      "epoch": 0.7082138392528863,
+      "grad_norm": 1.0648945569992065,
+      "learning_rate": 4.1447134722261285e-05,
+      "loss": 0.3761,
+      "step": 9631
+    },
+    {
+      "epoch": 0.7082873740716229,
+      "grad_norm": 1.3177554607391357,
+      "learning_rate": 4.1427829753814494e-05,
+      "loss": 0.3605,
+      "step": 9632
+    },
+    {
+      "epoch": 0.7083609088903596,
+      "grad_norm": 1.5018222332000732,
+      "learning_rate": 4.140852810764297e-05,
+      "loss": 0.3721,
+      "step": 9633
+    },
+    {
+      "epoch": 0.7084344437090963,
+      "grad_norm": 0.9634137749671936,
+      "learning_rate": 4.138922978484162e-05,
+      "loss": 0.3608,
+      "step": 9634
+    },
+    {
+      "epoch": 0.708507978527833,
+      "grad_norm": 0.7273737788200378,
+      "learning_rate": 4.1369934786504994e-05,
+      "loss": 0.3503,
+      "step": 9635
+    },
+    {
+      "epoch": 0.7085815133465696,
+      "grad_norm": 0.6562803387641907,
+      "learning_rate": 4.135064311372752e-05,
+      "loss": 0.3485,
+      "step": 9636
+    },
+    {
+      "epoch": 0.7086550481653062,
+      "grad_norm": 0.45692574977874756,
+      "learning_rate": 4.1331354767603445e-05,
+      "loss": 0.3556,
+      "step": 9637
+    },
+    {
+      "epoch": 0.708728582984043,
+      "grad_norm": 0.7395005226135254,
+      "learning_rate": 4.131206974922689e-05,
+      "loss": 0.3682,
+      "step": 9638
+    },
+    {
+      "epoch": 0.7088021178027796,
+      "grad_norm": 0.5045104026794434,
+      "learning_rate": 4.129278805969163e-05,
+      "loss": 0.3518,
+      "step": 9639
+    },
+    {
+      "epoch": 0.7088756526215163,
+      "grad_norm": 0.625607430934906,
+      "learning_rate": 4.127350970009141e-05,
+      "loss": 0.3342,
+      "step": 9640
+    },
+    {
+      "epoch": 0.7089491874402529,
+      "grad_norm": 0.5492692589759827,
+      "learning_rate": 4.1254234671519706e-05,
+      "loss": 0.3459,
+      "step": 9641
+    },
+    {
+      "epoch": 0.7090227222589897,
+      "grad_norm": 0.6384768486022949,
+      "learning_rate": 4.1234962975069794e-05,
+      "loss": 0.3377,
+      "step": 9642
+    },
+    {
+      "epoch": 0.7090962570777263,
+      "grad_norm": 0.6364525556564331,
+      "learning_rate": 4.121569461183479e-05,
+      "loss": 0.3313,
+      "step": 9643
+    },
+    {
+      "epoch": 0.709169791896463,
+      "grad_norm": 0.4404531419277191,
+      "learning_rate": 4.1196429582907614e-05,
+      "loss": 0.3497,
+      "step": 9644
+    },
+    {
+      "epoch": 0.7092433267151996,
+      "grad_norm": 0.5622639656066895,
+      "learning_rate": 4.117716788938106e-05,
+      "loss": 0.3573,
+      "step": 9645
+    },
+    {
+      "epoch": 0.7093168615339364,
+      "grad_norm": 0.489738792181015,
+      "learning_rate": 4.1157909532347616e-05,
+      "loss": 0.3433,
+      "step": 9646
+    },
+    {
+      "epoch": 0.709390396352673,
+      "grad_norm": 0.4964959919452667,
+      "learning_rate": 4.113865451289961e-05,
+      "loss": 0.3527,
+      "step": 9647
+    },
+    {
+      "epoch": 0.7094639311714097,
+      "grad_norm": 0.460058331489563,
+      "learning_rate": 4.111940283212927e-05,
+      "loss": 0.3492,
+      "step": 9648
+    },
+    {
+      "epoch": 0.7095374659901463,
+      "grad_norm": 0.5396617650985718,
+      "learning_rate": 4.1100154491128504e-05,
+      "loss": 0.3323,
+      "step": 9649
+    },
+    {
+      "epoch": 0.709611000808883,
+      "grad_norm": 0.4050856828689575,
+      "learning_rate": 4.108090949098914e-05,
+      "loss": 0.3534,
+      "step": 9650
+    },
+    {
+      "epoch": 0.7096845356276197,
+      "grad_norm": 0.4346780776977539,
+      "learning_rate": 4.10616678328028e-05,
+      "loss": 0.3549,
+      "step": 9651
+    },
+    {
+      "epoch": 0.7097580704463563,
+      "grad_norm": 0.41916924715042114,
+      "learning_rate": 4.104242951766083e-05,
+      "loss": 0.3443,
+      "step": 9652
+    },
+    {
+      "epoch": 0.709831605265093,
+      "grad_norm": 0.49599725008010864,
+      "learning_rate": 4.102319454665448e-05,
+      "loss": 0.3623,
+      "step": 9653
+    },
+    {
+      "epoch": 0.7099051400838297,
+      "grad_norm": 0.4317609965801239,
+      "learning_rate": 4.1003962920874805e-05,
+      "loss": 0.3628,
+      "step": 9654
+    },
+    {
+      "epoch": 0.7099786749025664,
+      "grad_norm": 0.5539970993995667,
+      "learning_rate": 4.098473464141258e-05,
+      "loss": 0.3662,
+      "step": 9655
+    },
+    {
+      "epoch": 0.710052209721303,
+      "grad_norm": 0.4686635136604309,
+      "learning_rate": 4.0965509709358486e-05,
+      "loss": 0.3455,
+      "step": 9656
+    },
+    {
+      "epoch": 0.7101257445400397,
+      "grad_norm": 0.5383292436599731,
+      "learning_rate": 4.0946288125803026e-05,
+      "loss": 0.3632,
+      "step": 9657
+    },
+    {
+      "epoch": 0.7101992793587764,
+      "grad_norm": 0.5563567280769348,
+      "learning_rate": 4.0927069891836415e-05,
+      "loss": 0.3712,
+      "step": 9658
+    },
+    {
+      "epoch": 0.7102728141775131,
+      "grad_norm": 0.43423953652381897,
+      "learning_rate": 4.090785500854871e-05,
+      "loss": 0.3624,
+      "step": 9659
+    },
+    {
+      "epoch": 0.7103463489962497,
+      "grad_norm": 0.5684024691581726,
+      "learning_rate": 4.088864347702983e-05,
+      "loss": 0.3759,
+      "step": 9660
+    },
+    {
+      "epoch": 0.7104198838149864,
+      "grad_norm": 0.4238218367099762,
+      "learning_rate": 4.086943529836952e-05,
+      "loss": 0.3576,
+      "step": 9661
+    },
+    {
+      "epoch": 0.7104934186337231,
+      "grad_norm": 0.46572113037109375,
+      "learning_rate": 4.0850230473657205e-05,
+      "loss": 0.3549,
+      "step": 9662
+    },
+    {
+      "epoch": 0.7105669534524598,
+      "grad_norm": 0.475348562002182,
+      "learning_rate": 4.083102900398222e-05,
+      "loss": 0.3698,
+      "step": 9663
+    },
+    {
+      "epoch": 0.7106404882711964,
+      "grad_norm": 0.4176289737224579,
+      "learning_rate": 4.081183089043378e-05,
+      "loss": 0.3516,
+      "step": 9664
+    },
+    {
+      "epoch": 0.710714023089933,
+      "grad_norm": 0.43927887082099915,
+      "learning_rate": 4.07926361341007e-05,
+      "loss": 0.3376,
+      "step": 9665
+    },
+    {
+      "epoch": 0.7107875579086698,
+      "grad_norm": 0.42121532559394836,
+      "learning_rate": 4.0773444736071786e-05,
+      "loss": 0.3577,
+      "step": 9666
+    },
+    {
+      "epoch": 0.7108610927274064,
+      "grad_norm": 0.43913817405700684,
+      "learning_rate": 4.075425669743559e-05,
+      "loss": 0.3551,
+      "step": 9667
+    },
+    {
+      "epoch": 0.7109346275461431,
+      "grad_norm": 0.39495164155960083,
+      "learning_rate": 4.0735072019280515e-05,
+      "loss": 0.3526,
+      "step": 9668
+    },
+    {
+      "epoch": 0.7110081623648797,
+      "grad_norm": 0.5011105537414551,
+      "learning_rate": 4.0715890702694705e-05,
+      "loss": 0.3544,
+      "step": 9669
+    },
+    {
+      "epoch": 0.7110816971836165,
+      "grad_norm": 0.5160853266716003,
+      "learning_rate": 4.06967127487661e-05,
+      "loss": 0.3701,
+      "step": 9670
+    },
+    {
+      "epoch": 0.7111552320023531,
+      "grad_norm": 0.5330423712730408,
+      "learning_rate": 4.0677538158582574e-05,
+      "loss": 0.3584,
+      "step": 9671
+    },
+    {
+      "epoch": 0.7112287668210898,
+      "grad_norm": 0.5578938722610474,
+      "learning_rate": 4.065836693323164e-05,
+      "loss": 0.3322,
+      "step": 9672
+    },
+    {
+      "epoch": 0.7113023016398264,
+      "grad_norm": 0.47802621126174927,
+      "learning_rate": 4.0639199073800774e-05,
+      "loss": 0.3504,
+      "step": 9673
+    },
+    {
+      "epoch": 0.7113758364585632,
+      "grad_norm": 0.561552882194519,
+      "learning_rate": 4.062003458137721e-05,
+      "loss": 0.3473,
+      "step": 9674
+    },
+    {
+      "epoch": 0.7114493712772998,
+      "grad_norm": 0.557141125202179,
+      "learning_rate": 4.060087345704792e-05,
+      "loss": 0.3347,
+      "step": 9675
+    },
+    {
+      "epoch": 0.7115229060960365,
+      "grad_norm": 0.5948776006698608,
+      "learning_rate": 4.0581715701899767e-05,
+      "loss": 0.3461,
+      "step": 9676
+    },
+    {
+      "epoch": 0.7115964409147731,
+      "grad_norm": 0.6248007416725159,
+      "learning_rate": 4.0562561317019445e-05,
+      "loss": 0.3545,
+      "step": 9677
+    },
+    {
+      "epoch": 0.7116699757335098,
+      "grad_norm": 0.6721441149711609,
+      "learning_rate": 4.054341030349334e-05,
+      "loss": 0.3745,
+      "step": 9678
+    },
+    {
+      "epoch": 0.7117435105522465,
+      "grad_norm": 0.5516718626022339,
+      "learning_rate": 4.052426266240774e-05,
+      "loss": 0.3854,
+      "step": 9679
+    },
+    {
+      "epoch": 0.7118170453709831,
+      "grad_norm": 0.5252427458763123,
+      "learning_rate": 4.050511839484874e-05,
+      "loss": 0.3591,
+      "step": 9680
+    },
+    {
+      "epoch": 0.7118905801897198,
+      "grad_norm": 0.5001206398010254,
+      "learning_rate": 4.0485977501902273e-05,
+      "loss": 0.3896,
+      "step": 9681
+    },
+    {
+      "epoch": 0.7119641150084565,
+      "grad_norm": 0.4874427914619446,
+      "learning_rate": 4.0466839984653904e-05,
+      "loss": 0.3485,
+      "step": 9682
+    },
+    {
+      "epoch": 0.7120376498271932,
+      "grad_norm": 0.5272668600082397,
+      "learning_rate": 4.0447705844189196e-05,
+      "loss": 0.3353,
+      "step": 9683
+    },
+    {
+      "epoch": 0.7121111846459298,
+      "grad_norm": 0.4794328808784485,
+      "learning_rate": 4.042857508159351e-05,
+      "loss": 0.3416,
+      "step": 9684
+    },
+    {
+      "epoch": 0.7121847194646665,
+      "grad_norm": 0.5085664391517639,
+      "learning_rate": 4.0409447697951864e-05,
+      "loss": 0.3623,
+      "step": 9685
+    },
+    {
+      "epoch": 0.7122582542834032,
+      "grad_norm": 0.45140454173088074,
+      "learning_rate": 4.039032369434924e-05,
+      "loss": 0.3397,
+      "step": 9686
+    },
+    {
+      "epoch": 0.7123317891021399,
+      "grad_norm": 0.6273918151855469,
+      "learning_rate": 4.037120307187039e-05,
+      "loss": 0.3886,
+      "step": 9687
+    },
+    {
+      "epoch": 0.7124053239208765,
+      "grad_norm": 0.4716092050075531,
+      "learning_rate": 4.0352085831599805e-05,
+      "loss": 0.3578,
+      "step": 9688
+    },
+    {
+      "epoch": 0.7124788587396133,
+      "grad_norm": 0.5312460660934448,
+      "learning_rate": 4.033297197462186e-05,
+      "loss": 0.3714,
+      "step": 9689
+    },
+    {
+      "epoch": 0.7125523935583499,
+      "grad_norm": 0.5569162964820862,
+      "learning_rate": 4.0313861502020714e-05,
+      "loss": 0.366,
+      "step": 9690
+    },
+    {
+      "epoch": 0.7126259283770866,
+      "grad_norm": 0.6374116539955139,
+      "learning_rate": 4.0294754414880356e-05,
+      "loss": 0.3582,
+      "step": 9691
+    },
+    {
+      "epoch": 0.7126994631958232,
+      "grad_norm": 0.49642089009284973,
+      "learning_rate": 4.027565071428455e-05,
+      "loss": 0.3989,
+      "step": 9692
+    },
+    {
+      "epoch": 0.7127729980145598,
+      "grad_norm": 0.8030725717544556,
+      "learning_rate": 4.025655040131683e-05,
+      "loss": 0.3348,
+      "step": 9693
+    },
+    {
+      "epoch": 0.7128465328332966,
+      "grad_norm": 0.4559403657913208,
+      "learning_rate": 4.023745347706065e-05,
+      "loss": 0.3422,
+      "step": 9694
+    },
+    {
+      "epoch": 0.7129200676520332,
+      "grad_norm": 0.6630784273147583,
+      "learning_rate": 4.021835994259914e-05,
+      "loss": 0.3899,
+      "step": 9695
+    },
+    {
+      "epoch": 0.7129936024707699,
+      "grad_norm": 0.5996916890144348,
+      "learning_rate": 4.019926979901535e-05,
+      "loss": 0.3522,
+      "step": 9696
+    },
+    {
+      "epoch": 0.7130671372895065,
+      "grad_norm": 0.8669707179069519,
+      "learning_rate": 4.018018304739213e-05,
+      "loss": 0.3925,
+      "step": 9697
+    },
+    {
+      "epoch": 0.7131406721082433,
+      "grad_norm": 0.5782567858695984,
+      "learning_rate": 4.016109968881201e-05,
+      "loss": 0.3373,
+      "step": 9698
+    },
+    {
+      "epoch": 0.7132142069269799,
+      "grad_norm": 0.8716286420822144,
+      "learning_rate": 4.014201972435747e-05,
+      "loss": 0.3791,
+      "step": 9699
+    },
+    {
+      "epoch": 0.7132877417457166,
+      "grad_norm": 0.5007483959197998,
+      "learning_rate": 4.012294315511078e-05,
+      "loss": 0.363,
+      "step": 9700
+    },
+    {
+      "epoch": 0.7133612765644533,
+      "grad_norm": 0.9557022452354431,
+      "learning_rate": 4.010386998215391e-05,
+      "loss": 0.3796,
+      "step": 9701
+    },
+    {
+      "epoch": 0.71343481138319,
+      "grad_norm": 0.6413217782974243,
+      "learning_rate": 4.008480020656874e-05,
+      "loss": 0.3565,
+      "step": 9702
+    },
+    {
+      "epoch": 0.7135083462019266,
+      "grad_norm": 0.6584614515304565,
+      "learning_rate": 4.006573382943694e-05,
+      "loss": 0.3363,
+      "step": 9703
+    },
+    {
+      "epoch": 0.7135818810206633,
+      "grad_norm": 0.470233678817749,
+      "learning_rate": 4.004667085184003e-05,
+      "loss": 0.3699,
+      "step": 9704
+    },
+    {
+      "epoch": 0.7136554158394,
+      "grad_norm": 0.678026556968689,
+      "learning_rate": 4.0027611274859155e-05,
+      "loss": 0.3446,
+      "step": 9705
+    },
+    {
+      "epoch": 0.7137289506581366,
+      "grad_norm": 0.5138819813728333,
+      "learning_rate": 4.0008555099575464e-05,
+      "loss": 0.3608,
+      "step": 9706
+    },
+    {
+      "epoch": 0.7138024854768733,
+      "grad_norm": 0.7729337811470032,
+      "learning_rate": 3.998950232706987e-05,
+      "loss": 0.3569,
+      "step": 9707
+    },
+    {
+      "epoch": 0.7138760202956099,
+      "grad_norm": 0.4437915086746216,
+      "learning_rate": 3.9970452958423e-05,
+      "loss": 0.3583,
+      "step": 9708
+    },
+    {
+      "epoch": 0.7139495551143467,
+      "grad_norm": 0.6811904311180115,
+      "learning_rate": 3.995140699471539e-05,
+      "loss": 0.3601,
+      "step": 9709
+    },
+    {
+      "epoch": 0.7140230899330833,
+      "grad_norm": 0.6284093856811523,
+      "learning_rate": 3.9932364437027393e-05,
+      "loss": 0.325,
+      "step": 9710
+    },
+    {
+      "epoch": 0.71409662475182,
+      "grad_norm": 0.7327436208724976,
+      "learning_rate": 3.991332528643903e-05,
+      "loss": 0.3453,
+      "step": 9711
+    },
+    {
+      "epoch": 0.7141701595705566,
+      "grad_norm": 0.9397602677345276,
+      "learning_rate": 3.989428954403027e-05,
+      "loss": 0.3522,
+      "step": 9712
+    },
+    {
+      "epoch": 0.7142436943892934,
+      "grad_norm": 0.47059985995292664,
+      "learning_rate": 3.9875257210880836e-05,
+      "loss": 0.3554,
+      "step": 9713
+    },
+    {
+      "epoch": 0.71431722920803,
+      "grad_norm": 0.6684949398040771,
+      "learning_rate": 3.9856228288070304e-05,
+      "loss": 0.3864,
+      "step": 9714
+    },
+    {
+      "epoch": 0.7143907640267667,
+      "grad_norm": 0.6445299983024597,
+      "learning_rate": 3.983720277667794e-05,
+      "loss": 0.37,
+      "step": 9715
+    },
+    {
+      "epoch": 0.7144642988455033,
+      "grad_norm": 0.6133123636245728,
+      "learning_rate": 3.981818067778297e-05,
+      "loss": 0.3529,
+      "step": 9716
+    },
+    {
+      "epoch": 0.7145378336642401,
+      "grad_norm": 0.7326287627220154,
+      "learning_rate": 3.979916199246429e-05,
+      "loss": 0.3515,
+      "step": 9717
+    },
+    {
+      "epoch": 0.7146113684829767,
+      "grad_norm": 0.5111469030380249,
+      "learning_rate": 3.9780146721800645e-05,
+      "loss": 0.3371,
+      "step": 9718
+    },
+    {
+      "epoch": 0.7146849033017134,
+      "grad_norm": 0.8477935194969177,
+      "learning_rate": 3.976113486687064e-05,
+      "loss": 0.3509,
+      "step": 9719
+    },
+    {
+      "epoch": 0.71475843812045,
+      "grad_norm": 0.5543200969696045,
+      "learning_rate": 3.974212642875266e-05,
+      "loss": 0.3764,
+      "step": 9720
+    },
+    {
+      "epoch": 0.7148319729391867,
+      "grad_norm": 0.7423045635223389,
+      "learning_rate": 3.9723121408524834e-05,
+      "loss": 0.3369,
+      "step": 9721
+    },
+    {
+      "epoch": 0.7149055077579234,
+      "grad_norm": 0.6606454253196716,
+      "learning_rate": 3.9704119807265174e-05,
+      "loss": 0.3382,
+      "step": 9722
+    },
+    {
+      "epoch": 0.71497904257666,
+      "grad_norm": 0.602641224861145,
+      "learning_rate": 3.9685121626051506e-05,
+      "loss": 0.3485,
+      "step": 9723
+    },
+    {
+      "epoch": 0.7150525773953967,
+      "grad_norm": 0.5461050271987915,
+      "learning_rate": 3.966612686596135e-05,
+      "loss": 0.3395,
+      "step": 9724
+    },
+    {
+      "epoch": 0.7151261122141334,
+      "grad_norm": 0.5231101512908936,
+      "learning_rate": 3.964713552807215e-05,
+      "loss": 0.3428,
+      "step": 9725
+    },
+    {
+      "epoch": 0.7151996470328701,
+      "grad_norm": 0.5285569429397583,
+      "learning_rate": 3.962814761346113e-05,
+      "loss": 0.3565,
+      "step": 9726
+    },
+    {
+      "epoch": 0.7152731818516067,
+      "grad_norm": 0.969703197479248,
+      "learning_rate": 3.960916312320535e-05,
+      "loss": 0.3727,
+      "step": 9727
+    },
+    {
+      "epoch": 0.7153467166703434,
+      "grad_norm": 0.8189136981964111,
+      "learning_rate": 3.9590182058381505e-05,
+      "loss": 0.3361,
+      "step": 9728
+    },
+    {
+      "epoch": 0.7154202514890801,
+      "grad_norm": 0.5305330157279968,
+      "learning_rate": 3.957120442006629e-05,
+      "loss": 0.3572,
+      "step": 9729
+    },
+    {
+      "epoch": 0.7154937863078168,
+      "grad_norm": 0.9707690477371216,
+      "learning_rate": 3.955223020933616e-05,
+      "loss": 0.3472,
+      "step": 9730
+    },
+    {
+      "epoch": 0.7155673211265534,
+      "grad_norm": 0.5444461703300476,
+      "learning_rate": 3.95332594272673e-05,
+      "loss": 0.3391,
+      "step": 9731
+    },
+    {
+      "epoch": 0.7156408559452901,
+      "grad_norm": 0.8093369007110596,
+      "learning_rate": 3.9514292074935765e-05,
+      "loss": 0.3411,
+      "step": 9732
+    },
+    {
+      "epoch": 0.7157143907640268,
+      "grad_norm": 0.5589645504951477,
+      "learning_rate": 3.9495328153417465e-05,
+      "loss": 0.3463,
+      "step": 9733
+    },
+    {
+      "epoch": 0.7157879255827634,
+      "grad_norm": 0.6737709641456604,
+      "learning_rate": 3.9476367663787964e-05,
+      "loss": 0.3685,
+      "step": 9734
+    },
+    {
+      "epoch": 0.7158614604015001,
+      "grad_norm": 0.46605363488197327,
+      "learning_rate": 3.945741060712276e-05,
+      "loss": 0.3319,
+      "step": 9735
+    },
+    {
+      "epoch": 0.7159349952202367,
+      "grad_norm": 0.6717389822006226,
+      "learning_rate": 3.9438456984497116e-05,
+      "loss": 0.346,
+      "step": 9736
+    },
+    {
+      "epoch": 0.7160085300389735,
+      "grad_norm": 0.5115780830383301,
+      "learning_rate": 3.941950679698615e-05,
+      "loss": 0.3282,
+      "step": 9737
+    },
+    {
+      "epoch": 0.7160820648577101,
+      "grad_norm": 0.6883215308189392,
+      "learning_rate": 3.9400560045664645e-05,
+      "loss": 0.3386,
+      "step": 9738
+    },
+    {
+      "epoch": 0.7161555996764468,
+      "grad_norm": 0.49773108959198,
+      "learning_rate": 3.9381616731607374e-05,
+      "loss": 0.3515,
+      "step": 9739
+    },
+    {
+      "epoch": 0.7162291344951834,
+      "grad_norm": 0.8992039561271667,
+      "learning_rate": 3.936267685588877e-05,
+      "loss": 0.3628,
+      "step": 9740
+    },
+    {
+      "epoch": 0.7163026693139202,
+      "grad_norm": 0.40958893299102783,
+      "learning_rate": 3.93437404195831e-05,
+      "loss": 0.3595,
+      "step": 9741
+    },
+    {
+      "epoch": 0.7163762041326568,
+      "grad_norm": 0.739706814289093,
+      "learning_rate": 3.932480742376448e-05,
+      "loss": 0.3618,
+      "step": 9742
+    },
+    {
+      "epoch": 0.7164497389513935,
+      "grad_norm": 0.4118359386920929,
+      "learning_rate": 3.930587786950685e-05,
+      "loss": 0.3582,
+      "step": 9743
+    },
+    {
+      "epoch": 0.7165232737701301,
+      "grad_norm": 0.6746358275413513,
+      "learning_rate": 3.928695175788385e-05,
+      "loss": 0.365,
+      "step": 9744
+    },
+    {
+      "epoch": 0.7165968085888669,
+      "grad_norm": 0.5314933657646179,
+      "learning_rate": 3.926802908996902e-05,
+      "loss": 0.3421,
+      "step": 9745
+    },
+    {
+      "epoch": 0.7166703434076035,
+      "grad_norm": 0.4774525761604309,
+      "learning_rate": 3.92491098668357e-05,
+      "loss": 0.3355,
+      "step": 9746
+    },
+    {
+      "epoch": 0.7167438782263402,
+      "grad_norm": 0.510736346244812,
+      "learning_rate": 3.923019408955695e-05,
+      "loss": 0.3263,
+      "step": 9747
+    },
+    {
+      "epoch": 0.7168174130450768,
+      "grad_norm": 0.6187212467193604,
+      "learning_rate": 3.921128175920572e-05,
+      "loss": 0.341,
+      "step": 9748
+    },
+    {
+      "epoch": 0.7168909478638135,
+      "grad_norm": 0.8508075475692749,
+      "learning_rate": 3.9192372876854746e-05,
+      "loss": 0.3757,
+      "step": 9749
+    },
+    {
+      "epoch": 0.7169644826825502,
+      "grad_norm": 0.5115249752998352,
+      "learning_rate": 3.9173467443576596e-05,
+      "loss": 0.3499,
+      "step": 9750
+    },
+    {
+      "epoch": 0.7170380175012868,
+      "grad_norm": 0.5831030011177063,
+      "learning_rate": 3.9154565460443556e-05,
+      "loss": 0.3743,
+      "step": 9751
+    },
+    {
+      "epoch": 0.7171115523200235,
+      "grad_norm": 0.9068475365638733,
+      "learning_rate": 3.913566692852775e-05,
+      "loss": 0.3483,
+      "step": 9752
+    },
+    {
+      "epoch": 0.7171850871387602,
+      "grad_norm": 0.42369771003723145,
+      "learning_rate": 3.911677184890117e-05,
+      "loss": 0.3707,
+      "step": 9753
+    },
+    {
+      "epoch": 0.7172586219574969,
+      "grad_norm": 0.7763363718986511,
+      "learning_rate": 3.909788022263552e-05,
+      "loss": 0.3437,
+      "step": 9754
+    },
+    {
+      "epoch": 0.7173321567762335,
+      "grad_norm": 0.42290136218070984,
+      "learning_rate": 3.907899205080237e-05,
+      "loss": 0.3444,
+      "step": 9755
+    },
+    {
+      "epoch": 0.7174056915949703,
+      "grad_norm": 0.5533301830291748,
+      "learning_rate": 3.906010733447312e-05,
+      "loss": 0.3456,
+      "step": 9756
+    },
+    {
+      "epoch": 0.7174792264137069,
+      "grad_norm": 0.5654445886611938,
+      "learning_rate": 3.904122607471886e-05,
+      "loss": 0.3354,
+      "step": 9757
+    },
+    {
+      "epoch": 0.7175527612324436,
+      "grad_norm": 0.8558905720710754,
+      "learning_rate": 3.902234827261059e-05,
+      "loss": 0.3458,
+      "step": 9758
+    },
+    {
+      "epoch": 0.7176262960511802,
+      "grad_norm": 0.4961293041706085,
+      "learning_rate": 3.900347392921908e-05,
+      "loss": 0.3519,
+      "step": 9759
+    },
+    {
+      "epoch": 0.717699830869917,
+      "grad_norm": 0.7686862349510193,
+      "learning_rate": 3.8984603045614934e-05,
+      "loss": 0.3346,
+      "step": 9760
+    },
+    {
+      "epoch": 0.7177733656886536,
+      "grad_norm": 0.418269544839859,
+      "learning_rate": 3.896573562286846e-05,
+      "loss": 0.3401,
+      "step": 9761
+    },
+    {
+      "epoch": 0.7178469005073902,
+      "grad_norm": 1.1076583862304688,
+      "learning_rate": 3.894687166204991e-05,
+      "loss": 0.3263,
+      "step": 9762
+    },
+    {
+      "epoch": 0.7179204353261269,
+      "grad_norm": 0.41663792729377747,
+      "learning_rate": 3.892801116422923e-05,
+      "loss": 0.3014,
+      "step": 9763
+    },
+    {
+      "epoch": 0.7179939701448635,
+      "grad_norm": 1.0567744970321655,
+      "learning_rate": 3.890915413047618e-05,
+      "loss": 0.3409,
+      "step": 9764
+    },
+    {
+      "epoch": 0.7180675049636003,
+      "grad_norm": 0.47608932852745056,
+      "learning_rate": 3.889030056186036e-05,
+      "loss": 0.3208,
+      "step": 9765
+    },
+    {
+      "epoch": 0.7181410397823369,
+      "grad_norm": 0.8711642026901245,
+      "learning_rate": 3.887145045945124e-05,
+      "loss": 0.3208,
+      "step": 9766
+    },
+    {
+      "epoch": 0.7182145746010736,
+      "grad_norm": 0.5572077631950378,
+      "learning_rate": 3.885260382431791e-05,
+      "loss": 0.3367,
+      "step": 9767
+    },
+    {
+      "epoch": 0.7182881094198103,
+      "grad_norm": 0.9539721012115479,
+      "learning_rate": 3.883376065752944e-05,
+      "loss": 0.3622,
+      "step": 9768
+    },
+    {
+      "epoch": 0.718361644238547,
+      "grad_norm": 0.592133104801178,
+      "learning_rate": 3.8814920960154646e-05,
+      "loss": 0.3532,
+      "step": 9769
+    },
+    {
+      "epoch": 0.7184351790572836,
+      "grad_norm": 0.8332561254501343,
+      "learning_rate": 3.879608473326206e-05,
+      "loss": 0.362,
+      "step": 9770
+    },
+    {
+      "epoch": 0.7185087138760203,
+      "grad_norm": 0.9274919629096985,
+      "learning_rate": 3.877725197792015e-05,
+      "loss": 0.3407,
+      "step": 9771
+    },
+    {
+      "epoch": 0.718582248694757,
+      "grad_norm": 1.1379884481430054,
+      "learning_rate": 3.875842269519712e-05,
+      "loss": 0.3467,
+      "step": 9772
+    },
+    {
+      "epoch": 0.7186557835134937,
+      "grad_norm": 0.8034078478813171,
+      "learning_rate": 3.873959688616101e-05,
+      "loss": 0.321,
+      "step": 9773
+    },
+    {
+      "epoch": 0.7187293183322303,
+      "grad_norm": 0.938740611076355,
+      "learning_rate": 3.872077455187963e-05,
+      "loss": 0.347,
+      "step": 9774
+    },
+    {
+      "epoch": 0.718802853150967,
+      "grad_norm": 0.5463693141937256,
+      "learning_rate": 3.8701955693420534e-05,
+      "loss": 0.3203,
+      "step": 9775
+    },
+    {
+      "epoch": 0.7188763879697037,
+      "grad_norm": 0.43969962000846863,
+      "learning_rate": 3.868314031185125e-05,
+      "loss": 0.3443,
+      "step": 9776
+    },
+    {
+      "epoch": 0.7189499227884403,
+      "grad_norm": 0.7928131222724915,
+      "learning_rate": 3.866432840823891e-05,
+      "loss": 0.3639,
+      "step": 9777
+    },
+    {
+      "epoch": 0.719023457607177,
+      "grad_norm": 0.45381882786750793,
+      "learning_rate": 3.864551998365061e-05,
+      "loss": 0.321,
+      "step": 9778
+    },
+    {
+      "epoch": 0.7190969924259136,
+      "grad_norm": 0.8548365235328674,
+      "learning_rate": 3.862671503915319e-05,
+      "loss": 0.3492,
+      "step": 9779
+    },
+    {
+      "epoch": 0.7191705272446504,
+      "grad_norm": 0.5189006328582764,
+      "learning_rate": 3.860791357581324e-05,
+      "loss": 0.3378,
+      "step": 9780
+    },
+    {
+      "epoch": 0.719244062063387,
+      "grad_norm": 0.5522581934928894,
+      "learning_rate": 3.858911559469721e-05,
+      "loss": 0.3166,
+      "step": 9781
+    },
+    {
+      "epoch": 0.7193175968821237,
+      "grad_norm": 0.41656172275543213,
+      "learning_rate": 3.857032109687135e-05,
+      "loss": 0.3251,
+      "step": 9782
+    },
+    {
+      "epoch": 0.7193911317008603,
+      "grad_norm": 0.47851231694221497,
+      "learning_rate": 3.855153008340175e-05,
+      "loss": 0.3424,
+      "step": 9783
+    },
+    {
+      "epoch": 0.7194646665195971,
+      "grad_norm": 0.6768891215324402,
+      "learning_rate": 3.853274255535418e-05,
+      "loss": 0.36,
+      "step": 9784
+    },
+    {
+      "epoch": 0.7195382013383337,
+      "grad_norm": 0.8288326263427734,
+      "learning_rate": 3.851395851379431e-05,
+      "loss": 0.3371,
+      "step": 9785
+    },
+    {
+      "epoch": 0.7196117361570704,
+      "grad_norm": 0.5022973418235779,
+      "learning_rate": 3.849517795978769e-05,
+      "loss": 0.3069,
+      "step": 9786
+    },
+    {
+      "epoch": 0.719685270975807,
+      "grad_norm": 0.8832995891571045,
+      "learning_rate": 3.84764008943994e-05,
+      "loss": 0.3724,
+      "step": 9787
+    },
+    {
+      "epoch": 0.7197588057945438,
+      "grad_norm": 0.3820752799510956,
+      "learning_rate": 3.845762731869458e-05,
+      "loss": 0.3365,
+      "step": 9788
+    },
+    {
+      "epoch": 0.7198323406132804,
+      "grad_norm": 0.736340343952179,
+      "learning_rate": 3.843885723373812e-05,
+      "loss": 0.4025,
+      "step": 9789
+    },
+    {
+      "epoch": 0.719905875432017,
+      "grad_norm": 0.47364574670791626,
+      "learning_rate": 3.842009064059462e-05,
+      "loss": 0.3461,
+      "step": 9790
+    },
+    {
+      "epoch": 0.7199794102507537,
+      "grad_norm": 0.4872604012489319,
+      "learning_rate": 3.8401327540328566e-05,
+      "loss": 0.3672,
+      "step": 9791
+    },
+    {
+      "epoch": 0.7200529450694904,
+      "grad_norm": 0.4738563001155853,
+      "learning_rate": 3.8382567934004255e-05,
+      "loss": 0.3652,
+      "step": 9792
+    },
+    {
+      "epoch": 0.7201264798882271,
+      "grad_norm": 0.42797425389289856,
+      "learning_rate": 3.836381182268568e-05,
+      "loss": 0.3737,
+      "step": 9793
+    },
+    {
+      "epoch": 0.7202000147069637,
+      "grad_norm": 0.43101218342781067,
+      "learning_rate": 3.834505920743675e-05,
+      "loss": 0.3645,
+      "step": 9794
+    },
+    {
+      "epoch": 0.7202735495257004,
+      "grad_norm": 0.453080952167511,
+      "learning_rate": 3.832631008932113e-05,
+      "loss": 0.3454,
+      "step": 9795
+    },
+    {
+      "epoch": 0.7203470843444371,
+      "grad_norm": 0.408916175365448,
+      "learning_rate": 3.8307564469402316e-05,
+      "loss": 0.3284,
+      "step": 9796
+    },
+    {
+      "epoch": 0.7204206191631738,
+      "grad_norm": 0.4809834361076355,
+      "learning_rate": 3.828882234874356e-05,
+      "loss": 0.3598,
+      "step": 9797
+    },
+    {
+      "epoch": 0.7204941539819104,
+      "grad_norm": 0.45207539200782776,
+      "learning_rate": 3.827008372840788e-05,
+      "loss": 0.3203,
+      "step": 9798
+    },
+    {
+      "epoch": 0.7205676888006471,
+      "grad_norm": 0.4112846255302429,
+      "learning_rate": 3.825134860945824e-05,
+      "loss": 0.3336,
+      "step": 9799
+    },
+    {
+      "epoch": 0.7206412236193838,
+      "grad_norm": 0.5914629101753235,
+      "learning_rate": 3.8232616992957236e-05,
+      "loss": 0.3503,
+      "step": 9800
+    },
+    {
+      "epoch": 0.7207147584381205,
+      "grad_norm": 0.5606886148452759,
+      "learning_rate": 3.821388887996739e-05,
+      "loss": 0.3268,
+      "step": 9801
+    },
+    {
+      "epoch": 0.7207882932568571,
+      "grad_norm": 0.47566699981689453,
+      "learning_rate": 3.8195164271550995e-05,
+      "loss": 0.3283,
+      "step": 9802
+    },
+    {
+      "epoch": 0.7208618280755938,
+      "grad_norm": 0.4449048638343811,
+      "learning_rate": 3.817644316877008e-05,
+      "loss": 0.3395,
+      "step": 9803
+    },
+    {
+      "epoch": 0.7209353628943305,
+      "grad_norm": 0.4127355217933655,
+      "learning_rate": 3.815772557268654e-05,
+      "loss": 0.3425,
+      "step": 9804
+    },
+    {
+      "epoch": 0.7210088977130671,
+      "grad_norm": 0.45135316252708435,
+      "learning_rate": 3.813901148436212e-05,
+      "loss": 0.3227,
+      "step": 9805
+    },
+    {
+      "epoch": 0.7210824325318038,
+      "grad_norm": 0.4727770984172821,
+      "learning_rate": 3.812030090485822e-05,
+      "loss": 0.3345,
+      "step": 9806
+    },
+    {
+      "epoch": 0.7211559673505404,
+      "grad_norm": 0.604666531085968,
+      "learning_rate": 3.810159383523615e-05,
+      "loss": 0.3785,
+      "step": 9807
+    },
+    {
+      "epoch": 0.7212295021692772,
+      "grad_norm": 0.4510834515094757,
+      "learning_rate": 3.8082890276557e-05,
+      "loss": 0.3399,
+      "step": 9808
+    },
+    {
+      "epoch": 0.7213030369880138,
+      "grad_norm": 0.5576369166374207,
+      "learning_rate": 3.806419022988173e-05,
+      "loss": 0.32,
+      "step": 9809
+    },
+    {
+      "epoch": 0.7213765718067505,
+      "grad_norm": 0.39969536662101746,
+      "learning_rate": 3.804549369627088e-05,
+      "loss": 0.3632,
+      "step": 9810
+    },
+    {
+      "epoch": 0.7214501066254871,
+      "grad_norm": 0.4421914219856262,
+      "learning_rate": 3.8026800676785024e-05,
+      "loss": 0.3503,
+      "step": 9811
+    },
+    {
+      "epoch": 0.7215236414442239,
+      "grad_norm": 0.50775545835495,
+      "learning_rate": 3.800811117248446e-05,
+      "loss": 0.3213,
+      "step": 9812
+    },
+    {
+      "epoch": 0.7215971762629605,
+      "grad_norm": 0.47110888361930847,
+      "learning_rate": 3.798942518442923e-05,
+      "loss": 0.3598,
+      "step": 9813
+    },
+    {
+      "epoch": 0.7216707110816972,
+      "grad_norm": 0.5291804075241089,
+      "learning_rate": 3.797074271367923e-05,
+      "loss": 0.3405,
+      "step": 9814
+    },
+    {
+      "epoch": 0.7217442459004338,
+      "grad_norm": 0.5074454545974731,
+      "learning_rate": 3.795206376129421e-05,
+      "loss": 0.3282,
+      "step": 9815
+    },
+    {
+      "epoch": 0.7218177807191706,
+      "grad_norm": 0.5158198475837708,
+      "learning_rate": 3.793338832833358e-05,
+      "loss": 0.3714,
+      "step": 9816
+    },
+    {
+      "epoch": 0.7218913155379072,
+      "grad_norm": 0.4771977663040161,
+      "learning_rate": 3.7914716415856655e-05,
+      "loss": 0.3524,
+      "step": 9817
+    },
+    {
+      "epoch": 0.7219648503566438,
+      "grad_norm": 0.5938387513160706,
+      "learning_rate": 3.789604802492255e-05,
+      "loss": 0.3526,
+      "step": 9818
+    },
+    {
+      "epoch": 0.7220383851753805,
+      "grad_norm": 0.4127366840839386,
+      "learning_rate": 3.7877383156590166e-05,
+      "loss": 0.3413,
+      "step": 9819
+    },
+    {
+      "epoch": 0.7221119199941172,
+      "grad_norm": 0.6239690780639648,
+      "learning_rate": 3.785872181191814e-05,
+      "loss": 0.3443,
+      "step": 9820
+    },
+    {
+      "epoch": 0.7221854548128539,
+      "grad_norm": 0.4818781614303589,
+      "learning_rate": 3.7840063991965024e-05,
+      "loss": 0.3439,
+      "step": 9821
+    },
+    {
+      "epoch": 0.7222589896315905,
+      "grad_norm": 0.5813525319099426,
+      "learning_rate": 3.782140969778908e-05,
+      "loss": 0.3564,
+      "step": 9822
+    },
+    {
+      "epoch": 0.7223325244503273,
+      "grad_norm": 0.6553165316581726,
+      "learning_rate": 3.780275893044837e-05,
+      "loss": 0.3608,
+      "step": 9823
+    },
+    {
+      "epoch": 0.7224060592690639,
+      "grad_norm": 0.4820668697357178,
+      "learning_rate": 3.77841116910008e-05,
+      "loss": 0.3413,
+      "step": 9824
+    },
+    {
+      "epoch": 0.7224795940878006,
+      "grad_norm": 0.6710219979286194,
+      "learning_rate": 3.776546798050412e-05,
+      "loss": 0.3719,
+      "step": 9825
+    },
+    {
+      "epoch": 0.7225531289065372,
+      "grad_norm": 0.5222293138504028,
+      "learning_rate": 3.774682780001574e-05,
+      "loss": 0.3207,
+      "step": 9826
+    },
+    {
+      "epoch": 0.722626663725274,
+      "grad_norm": 0.5248188376426697,
+      "learning_rate": 3.772819115059298e-05,
+      "loss": 0.3299,
+      "step": 9827
+    },
+    {
+      "epoch": 0.7227001985440106,
+      "grad_norm": 0.4902292490005493,
+      "learning_rate": 3.770955803329297e-05,
+      "loss": 0.3361,
+      "step": 9828
+    },
+    {
+      "epoch": 0.7227737333627473,
+      "grad_norm": 0.55344158411026,
+      "learning_rate": 3.769092844917254e-05,
+      "loss": 0.3622,
+      "step": 9829
+    },
+    {
+      "epoch": 0.7228472681814839,
+      "grad_norm": 0.7509507536888123,
+      "learning_rate": 3.76723023992884e-05,
+      "loss": 0.3422,
+      "step": 9830
+    },
+    {
+      "epoch": 0.7229208030002207,
+      "grad_norm": 0.4209148585796356,
+      "learning_rate": 3.765367988469704e-05,
+      "loss": 0.3428,
+      "step": 9831
+    },
+    {
+      "epoch": 0.7229943378189573,
+      "grad_norm": 0.6276606917381287,
+      "learning_rate": 3.763506090645482e-05,
+      "loss": 0.3428,
+      "step": 9832
+    },
+    {
+      "epoch": 0.7230678726376939,
+      "grad_norm": 0.5559438467025757,
+      "learning_rate": 3.761644546561771e-05,
+      "loss": 0.3519,
+      "step": 9833
+    },
+    {
+      "epoch": 0.7231414074564306,
+      "grad_norm": 0.8304657340049744,
+      "learning_rate": 3.759783356324164e-05,
+      "loss": 0.364,
+      "step": 9834
+    },
+    {
+      "epoch": 0.7232149422751672,
+      "grad_norm": 0.5153989195823669,
+      "learning_rate": 3.757922520038234e-05,
+      "loss": 0.3512,
+      "step": 9835
+    },
+    {
+      "epoch": 0.723288477093904,
+      "grad_norm": 0.7977685928344727,
+      "learning_rate": 3.756062037809522e-05,
+      "loss": 0.3589,
+      "step": 9836
+    },
+    {
+      "epoch": 0.7233620119126406,
+      "grad_norm": 0.694600522518158,
+      "learning_rate": 3.754201909743562e-05,
+      "loss": 0.3256,
+      "step": 9837
+    },
+    {
+      "epoch": 0.7234355467313773,
+      "grad_norm": 0.7794986963272095,
+      "learning_rate": 3.752342135945863e-05,
+      "loss": 0.3081,
+      "step": 9838
+    },
+    {
+      "epoch": 0.723509081550114,
+      "grad_norm": 0.6306648254394531,
+      "learning_rate": 3.750482716521909e-05,
+      "loss": 0.3337,
+      "step": 9839
+    },
+    {
+      "epoch": 0.7235826163688507,
+      "grad_norm": 0.7836182117462158,
+      "learning_rate": 3.748623651577171e-05,
+      "loss": 0.325,
+      "step": 9840
+    },
+    {
+      "epoch": 0.7236561511875873,
+      "grad_norm": 0.6278282999992371,
+      "learning_rate": 3.746764941217097e-05,
+      "loss": 0.3377,
+      "step": 9841
+    },
+    {
+      "epoch": 0.723729686006324,
+      "grad_norm": 0.7145109176635742,
+      "learning_rate": 3.7449065855471175e-05,
+      "loss": 0.3433,
+      "step": 9842
+    },
+    {
+      "epoch": 0.7238032208250607,
+      "grad_norm": 0.6652292013168335,
+      "learning_rate": 3.743048584672635e-05,
+      "loss": 0.3295,
+      "step": 9843
+    },
+    {
+      "epoch": 0.7238767556437974,
+      "grad_norm": 0.6445231437683105,
+      "learning_rate": 3.741190938699043e-05,
+      "loss": 0.327,
+      "step": 9844
+    },
+    {
+      "epoch": 0.723950290462534,
+      "grad_norm": 0.8116956353187561,
+      "learning_rate": 3.7393336477317064e-05,
+      "loss": 0.3483,
+      "step": 9845
+    },
+    {
+      "epoch": 0.7240238252812706,
+      "grad_norm": 0.558398425579071,
+      "learning_rate": 3.737476711875968e-05,
+      "loss": 0.3231,
+      "step": 9846
+    },
+    {
+      "epoch": 0.7240973601000074,
+      "grad_norm": 0.7271535992622375,
+      "learning_rate": 3.7356201312371605e-05,
+      "loss": 0.3386,
+      "step": 9847
+    },
+    {
+      "epoch": 0.724170894918744,
+      "grad_norm": 0.5139985084533691,
+      "learning_rate": 3.733763905920594e-05,
+      "loss": 0.3421,
+      "step": 9848
+    },
+    {
+      "epoch": 0.7242444297374807,
+      "grad_norm": 0.7436094880104065,
+      "learning_rate": 3.7319080360315475e-05,
+      "loss": 0.351,
+      "step": 9849
+    },
+    {
+      "epoch": 0.7243179645562173,
+      "grad_norm": 0.5024223923683167,
+      "learning_rate": 3.730052521675293e-05,
+      "loss": 0.3637,
+      "step": 9850
+    },
+    {
+      "epoch": 0.7243914993749541,
+      "grad_norm": 0.517486035823822,
+      "learning_rate": 3.728197362957079e-05,
+      "loss": 0.3241,
+      "step": 9851
+    },
+    {
+      "epoch": 0.7244650341936907,
+      "grad_norm": 0.41244181990623474,
+      "learning_rate": 3.726342559982127e-05,
+      "loss": 0.3329,
+      "step": 9852
+    },
+    {
+      "epoch": 0.7245385690124274,
+      "grad_norm": 0.6391118168830872,
+      "learning_rate": 3.7244881128556455e-05,
+      "loss": 0.3033,
+      "step": 9853
+    },
+    {
+      "epoch": 0.724612103831164,
+      "grad_norm": 0.4782552421092987,
+      "learning_rate": 3.7226340216828206e-05,
+      "loss": 0.3309,
+      "step": 9854
+    },
+    {
+      "epoch": 0.7246856386499008,
+      "grad_norm": 0.7414394617080688,
+      "learning_rate": 3.720780286568822e-05,
+      "loss": 0.3222,
+      "step": 9855
+    },
+    {
+      "epoch": 0.7247591734686374,
+      "grad_norm": 0.5041128993034363,
+      "learning_rate": 3.7189269076187925e-05,
+      "loss": 0.343,
+      "step": 9856
+    },
+    {
+      "epoch": 0.7248327082873741,
+      "grad_norm": 0.6677822470664978,
+      "learning_rate": 3.717073884937854e-05,
+      "loss": 0.353,
+      "step": 9857
+    },
+    {
+      "epoch": 0.7249062431061107,
+      "grad_norm": 0.4907589852809906,
+      "learning_rate": 3.71522121863112e-05,
+      "loss": 0.3281,
+      "step": 9858
+    },
+    {
+      "epoch": 0.7249797779248475,
+      "grad_norm": 0.54115229845047,
+      "learning_rate": 3.7133689088036664e-05,
+      "loss": 0.3246,
+      "step": 9859
+    },
+    {
+      "epoch": 0.7250533127435841,
+      "grad_norm": 0.605371356010437,
+      "learning_rate": 3.711516955560563e-05,
+      "loss": 0.3425,
+      "step": 9860
+    },
+    {
+      "epoch": 0.7251268475623207,
+      "grad_norm": 0.4973986744880676,
+      "learning_rate": 3.7096653590068574e-05,
+      "loss": 0.3496,
+      "step": 9861
+    },
+    {
+      "epoch": 0.7252003823810574,
+      "grad_norm": 0.7738516926765442,
+      "learning_rate": 3.707814119247569e-05,
+      "loss": 0.3667,
+      "step": 9862
+    },
+    {
+      "epoch": 0.7252739171997941,
+      "grad_norm": 0.519071102142334,
+      "learning_rate": 3.705963236387703e-05,
+      "loss": 0.327,
+      "step": 9863
+    },
+    {
+      "epoch": 0.7253474520185308,
+      "grad_norm": 0.8536173105239868,
+      "learning_rate": 3.704112710532245e-05,
+      "loss": 0.3328,
+      "step": 9864
+    },
+    {
+      "epoch": 0.7254209868372674,
+      "grad_norm": 0.5727661848068237,
+      "learning_rate": 3.702262541786161e-05,
+      "loss": 0.3405,
+      "step": 9865
+    },
+    {
+      "epoch": 0.7254945216560041,
+      "grad_norm": 0.8028700947761536,
+      "learning_rate": 3.70041273025439e-05,
+      "loss": 0.3186,
+      "step": 9866
+    },
+    {
+      "epoch": 0.7255680564747408,
+      "grad_norm": 0.6486015915870667,
+      "learning_rate": 3.6985632760418565e-05,
+      "loss": 0.3389,
+      "step": 9867
+    },
+    {
+      "epoch": 0.7256415912934775,
+      "grad_norm": 0.7877935171127319,
+      "learning_rate": 3.696714179253471e-05,
+      "loss": 0.3292,
+      "step": 9868
+    },
+    {
+      "epoch": 0.7257151261122141,
+      "grad_norm": 0.5433515906333923,
+      "learning_rate": 3.694865439994102e-05,
+      "loss": 0.3304,
+      "step": 9869
+    },
+    {
+      "epoch": 0.7257886609309508,
+      "grad_norm": 0.9332024455070496,
+      "learning_rate": 3.6930170583686195e-05,
+      "loss": 0.3379,
+      "step": 9870
+    },
+    {
+      "epoch": 0.7258621957496875,
+      "grad_norm": 0.6044676303863525,
+      "learning_rate": 3.69116903448187e-05,
+      "loss": 0.3357,
+      "step": 9871
+    },
+    {
+      "epoch": 0.7259357305684242,
+      "grad_norm": 0.7076053023338318,
+      "learning_rate": 3.689321368438666e-05,
+      "loss": 0.3182,
+      "step": 9872
+    },
+    {
+      "epoch": 0.7260092653871608,
+      "grad_norm": 0.6316573023796082,
+      "learning_rate": 3.687474060343814e-05,
+      "loss": 0.3418,
+      "step": 9873
+    },
+    {
+      "epoch": 0.7260828002058974,
+      "grad_norm": 0.5046650171279907,
+      "learning_rate": 3.6856271103021e-05,
+      "loss": 0.3275,
+      "step": 9874
+    },
+    {
+      "epoch": 0.7261563350246342,
+      "grad_norm": 0.4997893273830414,
+      "learning_rate": 3.683780518418275e-05,
+      "loss": 0.3267,
+      "step": 9875
+    },
+    {
+      "epoch": 0.7262298698433708,
+      "grad_norm": 0.5840170383453369,
+      "learning_rate": 3.681934284797086e-05,
+      "loss": 0.341,
+      "step": 9876
+    },
+    {
+      "epoch": 0.7263034046621075,
+      "grad_norm": 0.36245986819267273,
+      "learning_rate": 3.680088409543252e-05,
+      "loss": 0.3629,
+      "step": 9877
+    },
+    {
+      "epoch": 0.7263769394808441,
+      "grad_norm": 0.78248131275177,
+      "learning_rate": 3.6782428927614754e-05,
+      "loss": 0.3396,
+      "step": 9878
+    },
+    {
+      "epoch": 0.7264504742995809,
+      "grad_norm": 0.498477041721344,
+      "learning_rate": 3.676397734556435e-05,
+      "loss": 0.3399,
+      "step": 9879
+    },
+    {
+      "epoch": 0.7265240091183175,
+      "grad_norm": 0.7129235863685608,
+      "learning_rate": 3.6745529350327835e-05,
+      "loss": 0.3281,
+      "step": 9880
+    },
+    {
+      "epoch": 0.7265975439370542,
+      "grad_norm": 0.6656875014305115,
+      "learning_rate": 3.6727084942951696e-05,
+      "loss": 0.3221,
+      "step": 9881
+    },
+    {
+      "epoch": 0.7266710787557908,
+      "grad_norm": 0.7821856141090393,
+      "learning_rate": 3.670864412448204e-05,
+      "loss": 0.3367,
+      "step": 9882
+    },
+    {
+      "epoch": 0.7267446135745276,
+      "grad_norm": 0.6801834106445312,
+      "learning_rate": 3.669020689596488e-05,
+      "loss": 0.3252,
+      "step": 9883
+    },
+    {
+      "epoch": 0.7268181483932642,
+      "grad_norm": 0.6096077561378479,
+      "learning_rate": 3.667177325844603e-05,
+      "loss": 0.3164,
+      "step": 9884
+    },
+    {
+      "epoch": 0.7268916832120009,
+      "grad_norm": 0.6421117186546326,
+      "learning_rate": 3.6653343212971006e-05,
+      "loss": 0.3312,
+      "step": 9885
+    },
+    {
+      "epoch": 0.7269652180307375,
+      "grad_norm": 0.4650944769382477,
+      "learning_rate": 3.6634916760585206e-05,
+      "loss": 0.328,
+      "step": 9886
+    },
+    {
+      "epoch": 0.7270387528494743,
+      "grad_norm": 0.5150191783905029,
+      "learning_rate": 3.6616493902333795e-05,
+      "loss": 0.328,
+      "step": 9887
+    },
+    {
+      "epoch": 0.7271122876682109,
+      "grad_norm": 0.6283262372016907,
+      "learning_rate": 3.659807463926178e-05,
+      "loss": 0.3287,
+      "step": 9888
+    },
+    {
+      "epoch": 0.7271858224869475,
+      "grad_norm": 0.4481499195098877,
+      "learning_rate": 3.6579658972413844e-05,
+      "loss": 0.3277,
+      "step": 9889
+    },
+    {
+      "epoch": 0.7272593573056843,
+      "grad_norm": 0.6189059615135193,
+      "learning_rate": 3.656124690283459e-05,
+      "loss": 0.3242,
+      "step": 9890
+    },
+    {
+      "epoch": 0.7273328921244209,
+      "grad_norm": 0.42047062516212463,
+      "learning_rate": 3.654283843156842e-05,
+      "loss": 0.3148,
+      "step": 9891
+    },
+    {
+      "epoch": 0.7274064269431576,
+      "grad_norm": 0.8232445120811462,
+      "learning_rate": 3.652443355965936e-05,
+      "loss": 0.3291,
+      "step": 9892
+    },
+    {
+      "epoch": 0.7274799617618942,
+      "grad_norm": 0.5400626063346863,
+      "learning_rate": 3.650603228815141e-05,
+      "loss": 0.348,
+      "step": 9893
+    },
+    {
+      "epoch": 0.727553496580631,
+      "grad_norm": 0.7077703475952148,
+      "learning_rate": 3.648763461808835e-05,
+      "loss": 0.3357,
+      "step": 9894
+    },
+    {
+      "epoch": 0.7276270313993676,
+      "grad_norm": 0.47724708914756775,
+      "learning_rate": 3.6469240550513664e-05,
+      "loss": 0.3358,
+      "step": 9895
+    },
+    {
+      "epoch": 0.7277005662181043,
+      "grad_norm": 1.0990368127822876,
+      "learning_rate": 3.6450850086470677e-05,
+      "loss": 0.3356,
+      "step": 9896
+    },
+    {
+      "epoch": 0.7277741010368409,
+      "grad_norm": 0.5341111421585083,
+      "learning_rate": 3.6432463227002586e-05,
+      "loss": 0.3381,
+      "step": 9897
+    },
+    {
+      "epoch": 0.7278476358555777,
+      "grad_norm": 0.539677619934082,
+      "learning_rate": 3.641407997315224e-05,
+      "loss": 0.3366,
+      "step": 9898
+    },
+    {
+      "epoch": 0.7279211706743143,
+      "grad_norm": 0.46368083357810974,
+      "learning_rate": 3.639570032596237e-05,
+      "loss": 0.335,
+      "step": 9899
+    },
+    {
+      "epoch": 0.727994705493051,
+      "grad_norm": 0.5474029183387756,
+      "learning_rate": 3.637732428647551e-05,
+      "loss": 0.3172,
+      "step": 9900
+    },
+    {
+      "epoch": 0.7280682403117876,
+      "grad_norm": 0.4732530415058136,
+      "learning_rate": 3.635895185573399e-05,
+      "loss": 0.3447,
+      "step": 9901
+    },
+    {
+      "epoch": 0.7281417751305242,
+      "grad_norm": 0.5638248324394226,
+      "learning_rate": 3.634058303477985e-05,
+      "loss": 0.346,
+      "step": 9902
+    },
+    {
+      "epoch": 0.728215309949261,
+      "grad_norm": 0.38803187012672424,
+      "learning_rate": 3.632221782465506e-05,
+      "loss": 0.3283,
+      "step": 9903
+    },
+    {
+      "epoch": 0.7282888447679976,
+      "grad_norm": 0.8055105805397034,
+      "learning_rate": 3.6303856226401264e-05,
+      "loss": 0.2991,
+      "step": 9904
+    },
+    {
+      "epoch": 0.7283623795867343,
+      "grad_norm": 0.4557923674583435,
+      "learning_rate": 3.6285498241059944e-05,
+      "loss": 0.32,
+      "step": 9905
+    },
+    {
+      "epoch": 0.728435914405471,
+      "grad_norm": 1.0599162578582764,
+      "learning_rate": 3.6267143869672405e-05,
+      "loss": 0.3162,
+      "step": 9906
+    },
+    {
+      "epoch": 0.7285094492242077,
+      "grad_norm": 0.4667108952999115,
+      "learning_rate": 3.624879311327975e-05,
+      "loss": 0.3399,
+      "step": 9907
+    },
+    {
+      "epoch": 0.7285829840429443,
+      "grad_norm": 0.6675218343734741,
+      "learning_rate": 3.62304459729228e-05,
+      "loss": 0.3265,
+      "step": 9908
+    },
+    {
+      "epoch": 0.728656518861681,
+      "grad_norm": 0.6008273363113403,
+      "learning_rate": 3.621210244964226e-05,
+      "loss": 0.3128,
+      "step": 9909
+    },
+    {
+      "epoch": 0.7287300536804177,
+      "grad_norm": 0.5127565264701843,
+      "learning_rate": 3.6193762544478574e-05,
+      "loss": 0.3307,
+      "step": 9910
+    },
+    {
+      "epoch": 0.7288035884991544,
+      "grad_norm": 0.7895171642303467,
+      "learning_rate": 3.6175426258472053e-05,
+      "loss": 0.3172,
+      "step": 9911
+    },
+    {
+      "epoch": 0.728877123317891,
+      "grad_norm": 0.5364162921905518,
+      "learning_rate": 3.6157093592662686e-05,
+      "loss": 0.3406,
+      "step": 9912
+    },
+    {
+      "epoch": 0.7289506581366277,
+      "grad_norm": 0.696061909198761,
+      "learning_rate": 3.613876454809033e-05,
+      "loss": 0.3299,
+      "step": 9913
+    },
+    {
+      "epoch": 0.7290241929553644,
+      "grad_norm": 0.6490724682807922,
+      "learning_rate": 3.612043912579473e-05,
+      "loss": 0.3062,
+      "step": 9914
+    },
+    {
+      "epoch": 0.7290977277741011,
+      "grad_norm": 0.4828961193561554,
+      "learning_rate": 3.6102117326815155e-05,
+      "loss": 0.3202,
+      "step": 9915
+    },
+    {
+      "epoch": 0.7291712625928377,
+      "grad_norm": 0.5312831997871399,
+      "learning_rate": 3.608379915219092e-05,
+      "loss": 0.319,
+      "step": 9916
+    },
+    {
+      "epoch": 0.7292447974115743,
+      "grad_norm": 0.6442121863365173,
+      "learning_rate": 3.60654846029611e-05,
+      "loss": 0.3079,
+      "step": 9917
+    },
+    {
+      "epoch": 0.7293183322303111,
+      "grad_norm": 0.4524915814399719,
+      "learning_rate": 3.6047173680164416e-05,
+      "loss": 0.3034,
+      "step": 9918
+    },
+    {
+      "epoch": 0.7293918670490477,
+      "grad_norm": 0.8913872838020325,
+      "learning_rate": 3.6028866384839543e-05,
+      "loss": 0.318,
+      "step": 9919
+    },
+    {
+      "epoch": 0.7294654018677844,
+      "grad_norm": 0.6247218251228333,
+      "learning_rate": 3.601056271802491e-05,
+      "loss": 0.319,
+      "step": 9920
+    },
+    {
+      "epoch": 0.729538936686521,
+      "grad_norm": 0.5163448452949524,
+      "learning_rate": 3.599226268075866e-05,
+      "loss": 0.366,
+      "step": 9921
+    },
+    {
+      "epoch": 0.7296124715052578,
+      "grad_norm": 0.5645058155059814,
+      "learning_rate": 3.5973966274078806e-05,
+      "loss": 0.2959,
+      "step": 9922
+    },
+    {
+      "epoch": 0.7296860063239944,
+      "grad_norm": 0.429625928401947,
+      "learning_rate": 3.595567349902317e-05,
+      "loss": 0.2926,
+      "step": 9923
+    },
+    {
+      "epoch": 0.7297595411427311,
+      "grad_norm": 0.543308436870575,
+      "learning_rate": 3.5937384356629354e-05,
+      "loss": 0.3392,
+      "step": 9924
+    },
+    {
+      "epoch": 0.7298330759614677,
+      "grad_norm": 0.563117265701294,
+      "learning_rate": 3.5919098847934666e-05,
+      "loss": 0.3156,
+      "step": 9925
+    },
+    {
+      "epoch": 0.7299066107802045,
+      "grad_norm": 0.4096028506755829,
+      "learning_rate": 3.5900816973976356e-05,
+      "loss": 0.3198,
+      "step": 9926
+    },
+    {
+      "epoch": 0.7299801455989411,
+      "grad_norm": 0.47077274322509766,
+      "learning_rate": 3.588253873579135e-05,
+      "loss": 0.3252,
+      "step": 9927
+    },
+    {
+      "epoch": 0.7300536804176778,
+      "grad_norm": 0.5068917274475098,
+      "learning_rate": 3.586426413441638e-05,
+      "loss": 0.3309,
+      "step": 9928
+    },
+    {
+      "epoch": 0.7301272152364144,
+      "grad_norm": 0.5459272861480713,
+      "learning_rate": 3.584599317088804e-05,
+      "loss": 0.3281,
+      "step": 9929
+    },
+    {
+      "epoch": 0.7302007500551511,
+      "grad_norm": 0.5035730004310608,
+      "learning_rate": 3.58277258462427e-05,
+      "loss": 0.313,
+      "step": 9930
+    },
+    {
+      "epoch": 0.7302742848738878,
+      "grad_norm": 0.8488215804100037,
+      "learning_rate": 3.5809462161516436e-05,
+      "loss": 0.3147,
+      "step": 9931
+    },
+    {
+      "epoch": 0.7303478196926244,
+      "grad_norm": 0.42042770981788635,
+      "learning_rate": 3.5791202117745226e-05,
+      "loss": 0.3163,
+      "step": 9932
+    },
+    {
+      "epoch": 0.7304213545113611,
+      "grad_norm": 0.6922115683555603,
+      "learning_rate": 3.577294571596483e-05,
+      "loss": 0.3228,
+      "step": 9933
+    },
+    {
+      "epoch": 0.7304948893300978,
+      "grad_norm": 0.560339093208313,
+      "learning_rate": 3.57546929572107e-05,
+      "loss": 0.3139,
+      "step": 9934
+    },
+    {
+      "epoch": 0.7305684241488345,
+      "grad_norm": 0.6504924297332764,
+      "learning_rate": 3.5736443842518184e-05,
+      "loss": 0.312,
+      "step": 9935
+    },
+    {
+      "epoch": 0.7306419589675711,
+      "grad_norm": 0.7144525647163391,
+      "learning_rate": 3.57181983729224e-05,
+      "loss": 0.3039,
+      "step": 9936
+    },
+    {
+      "epoch": 0.7307154937863078,
+      "grad_norm": 0.7261081337928772,
+      "learning_rate": 3.569995654945826e-05,
+      "loss": 0.2898,
+      "step": 9937
+    },
+    {
+      "epoch": 0.7307890286050445,
+      "grad_norm": 1.0432000160217285,
+      "learning_rate": 3.5681718373160457e-05,
+      "loss": 0.305,
+      "step": 9938
+    },
+    {
+      "epoch": 0.7308625634237812,
+      "grad_norm": 0.9771158695220947,
+      "learning_rate": 3.566348384506343e-05,
+      "loss": 0.3179,
+      "step": 9939
+    },
+    {
+      "epoch": 0.7309360982425178,
+      "grad_norm": 0.9437362551689148,
+      "learning_rate": 3.564525296620152e-05,
+      "loss": 0.3169,
+      "step": 9940
+    },
+    {
+      "epoch": 0.7310096330612545,
+      "grad_norm": 0.9133866429328918,
+      "learning_rate": 3.5627025737608764e-05,
+      "loss": 0.3007,
+      "step": 9941
+    },
+    {
+      "epoch": 0.7310831678799912,
+      "grad_norm": 0.8609307408332825,
+      "learning_rate": 3.560880216031903e-05,
+      "loss": 0.2847,
+      "step": 9942
+    },
+    {
+      "epoch": 0.7311567026987279,
+      "grad_norm": 0.6631449460983276,
+      "learning_rate": 3.559058223536603e-05,
+      "loss": 0.3091,
+      "step": 9943
+    },
+    {
+      "epoch": 0.7312302375174645,
+      "grad_norm": 1.1436527967453003,
+      "learning_rate": 3.557236596378315e-05,
+      "loss": 0.3271,
+      "step": 9944
+    },
+    {
+      "epoch": 0.7313037723362011,
+      "grad_norm": 0.44793426990509033,
+      "learning_rate": 3.555415334660366e-05,
+      "loss": 0.2793,
+      "step": 9945
+    },
+    {
+      "epoch": 0.7313773071549379,
+      "grad_norm": 1.085526943206787,
+      "learning_rate": 3.55359443848606e-05,
+      "loss": 0.3297,
+      "step": 9946
+    },
+    {
+      "epoch": 0.7314508419736745,
+      "grad_norm": 0.5073277354240417,
+      "learning_rate": 3.5517739079586844e-05,
+      "loss": 0.3199,
+      "step": 9947
+    },
+    {
+      "epoch": 0.7315243767924112,
+      "grad_norm": 0.9715854525566101,
+      "learning_rate": 3.5499537431814946e-05,
+      "loss": 0.3163,
+      "step": 9948
+    },
+    {
+      "epoch": 0.7315979116111478,
+      "grad_norm": 0.6046716570854187,
+      "learning_rate": 3.548133944257739e-05,
+      "loss": 0.3428,
+      "step": 9949
+    },
+    {
+      "epoch": 0.7316714464298846,
+      "grad_norm": 0.7521737813949585,
+      "learning_rate": 3.546314511290635e-05,
+      "loss": 0.3236,
+      "step": 9950
+    },
+    {
+      "epoch": 0.7317449812486212,
+      "grad_norm": 0.5281789898872375,
+      "learning_rate": 3.544495444383379e-05,
+      "loss": 0.3118,
+      "step": 9951
+    },
+    {
+      "epoch": 0.7318185160673579,
+      "grad_norm": 0.8013538718223572,
+      "learning_rate": 3.542676743639154e-05,
+      "loss": 0.3172,
+      "step": 9952
+    },
+    {
+      "epoch": 0.7318920508860945,
+      "grad_norm": 0.6479586362838745,
+      "learning_rate": 3.540858409161123e-05,
+      "loss": 0.3009,
+      "step": 9953
+    },
+    {
+      "epoch": 0.7319655857048313,
+      "grad_norm": 0.722156822681427,
+      "learning_rate": 3.539040441052416e-05,
+      "loss": 0.3166,
+      "step": 9954
+    },
+    {
+      "epoch": 0.7320391205235679,
+      "grad_norm": 0.8239865303039551,
+      "learning_rate": 3.5372228394161534e-05,
+      "loss": 0.3241,
+      "step": 9955
+    },
+    {
+      "epoch": 0.7321126553423046,
+      "grad_norm": 0.49318820238113403,
+      "learning_rate": 3.535405604355435e-05,
+      "loss": 0.3058,
+      "step": 9956
+    },
+    {
+      "epoch": 0.7321861901610413,
+      "grad_norm": 0.6644738912582397,
+      "learning_rate": 3.53358873597333e-05,
+      "loss": 0.3363,
+      "step": 9957
+    },
+    {
+      "epoch": 0.7322597249797779,
+      "grad_norm": 0.41076475381851196,
+      "learning_rate": 3.531772234372897e-05,
+      "loss": 0.2993,
+      "step": 9958
+    },
+    {
+      "epoch": 0.7323332597985146,
+      "grad_norm": 0.5016968846321106,
+      "learning_rate": 3.529956099657167e-05,
+      "loss": 0.3115,
+      "step": 9959
+    },
+    {
+      "epoch": 0.7324067946172512,
+      "grad_norm": 0.5745558142662048,
+      "learning_rate": 3.5281403319291596e-05,
+      "loss": 0.3261,
+      "step": 9960
+    },
+    {
+      "epoch": 0.732480329435988,
+      "grad_norm": 0.5427923798561096,
+      "learning_rate": 3.5263249312918635e-05,
+      "loss": 0.2925,
+      "step": 9961
+    },
+    {
+      "epoch": 0.7325538642547246,
+      "grad_norm": 0.6688679456710815,
+      "learning_rate": 3.524509897848245e-05,
+      "loss": 0.3353,
+      "step": 9962
+    },
+    {
+      "epoch": 0.7326273990734613,
+      "grad_norm": 0.5233243107795715,
+      "learning_rate": 3.522695231701262e-05,
+      "loss": 0.2941,
+      "step": 9963
+    },
+    {
+      "epoch": 0.7327009338921979,
+      "grad_norm": 0.6522636413574219,
+      "learning_rate": 3.5208809329538384e-05,
+      "loss": 0.3075,
+      "step": 9964
+    },
+    {
+      "epoch": 0.7327744687109347,
+      "grad_norm": 0.8374513387680054,
+      "learning_rate": 3.5190670017088855e-05,
+      "loss": 0.296,
+      "step": 9965
+    },
+    {
+      "epoch": 0.7328480035296713,
+      "grad_norm": 0.4160602390766144,
+      "learning_rate": 3.5172534380692957e-05,
+      "loss": 0.3141,
+      "step": 9966
+    },
+    {
+      "epoch": 0.732921538348408,
+      "grad_norm": 0.551238477230072,
+      "learning_rate": 3.515440242137928e-05,
+      "loss": 0.3184,
+      "step": 9967
+    },
+    {
+      "epoch": 0.7329950731671446,
+      "grad_norm": 0.566423773765564,
+      "learning_rate": 3.513627414017635e-05,
+      "loss": 0.3147,
+      "step": 9968
+    },
+    {
+      "epoch": 0.7330686079858814,
+      "grad_norm": 0.5912564992904663,
+      "learning_rate": 3.51181495381124e-05,
+      "loss": 0.2945,
+      "step": 9969
+    },
+    {
+      "epoch": 0.733142142804618,
+      "grad_norm": 0.629987895488739,
+      "learning_rate": 3.510002861621551e-05,
+      "loss": 0.342,
+      "step": 9970
+    },
+    {
+      "epoch": 0.7332156776233547,
+      "grad_norm": 0.5731266736984253,
+      "learning_rate": 3.5081911375513447e-05,
+      "loss": 0.3055,
+      "step": 9971
+    },
+    {
+      "epoch": 0.7332892124420913,
+      "grad_norm": 0.43575626611709595,
+      "learning_rate": 3.506379781703389e-05,
+      "loss": 0.321,
+      "step": 9972
+    },
+    {
+      "epoch": 0.733362747260828,
+      "grad_norm": 0.6095768213272095,
+      "learning_rate": 3.504568794180432e-05,
+      "loss": 0.3263,
+      "step": 9973
+    },
+    {
+      "epoch": 0.7334362820795647,
+      "grad_norm": 0.5660428404808044,
+      "learning_rate": 3.50275817508518e-05,
+      "loss": 0.304,
+      "step": 9974
+    },
+    {
+      "epoch": 0.7335098168983013,
+      "grad_norm": 0.5450838208198547,
+      "learning_rate": 3.500947924520343e-05,
+      "loss": 0.3073,
+      "step": 9975
+    },
+    {
+      "epoch": 0.733583351717038,
+      "grad_norm": 0.652557373046875,
+      "learning_rate": 3.4991380425886e-05,
+      "loss": 0.3284,
+      "step": 9976
+    },
+    {
+      "epoch": 0.7336568865357747,
+      "grad_norm": 0.4503464102745056,
+      "learning_rate": 3.497328529392605e-05,
+      "loss": 0.3045,
+      "step": 9977
+    },
+    {
+      "epoch": 0.7337304213545114,
+      "grad_norm": 0.5421918034553528,
+      "learning_rate": 3.495519385034999e-05,
+      "loss": 0.2969,
+      "step": 9978
+    },
+    {
+      "epoch": 0.733803956173248,
+      "grad_norm": 0.6240904331207275,
+      "learning_rate": 3.493710609618401e-05,
+      "loss": 0.3188,
+      "step": 9979
+    },
+    {
+      "epoch": 0.7338774909919847,
+      "grad_norm": 0.48236969113349915,
+      "learning_rate": 3.4919022032453995e-05,
+      "loss": 0.3002,
+      "step": 9980
+    },
+    {
+      "epoch": 0.7339510258107214,
+      "grad_norm": 0.5655259490013123,
+      "learning_rate": 3.490094166018575e-05,
+      "loss": 0.3204,
+      "step": 9981
+    },
+    {
+      "epoch": 0.7340245606294581,
+      "grad_norm": 0.5276697278022766,
+      "learning_rate": 3.488286498040478e-05,
+      "loss": 0.2998,
+      "step": 9982
+    },
+    {
+      "epoch": 0.7340980954481947,
+      "grad_norm": 0.435646653175354,
+      "learning_rate": 3.486479199413647e-05,
+      "loss": 0.3124,
+      "step": 9983
+    },
+    {
+      "epoch": 0.7341716302669314,
+      "grad_norm": 0.7330833077430725,
+      "learning_rate": 3.48467227024059e-05,
+      "loss": 0.3404,
+      "step": 9984
+    },
+    {
+      "epoch": 0.7342451650856681,
+      "grad_norm": 0.5996506810188293,
+      "learning_rate": 3.482865710623794e-05,
+      "loss": 0.311,
+      "step": 9985
+    },
+    {
+      "epoch": 0.7343186999044047,
+      "grad_norm": 0.6841238141059875,
+      "learning_rate": 3.481059520665736e-05,
+      "loss": 0.3082,
+      "step": 9986
+    },
+    {
+      "epoch": 0.7343922347231414,
+      "grad_norm": 0.6138297915458679,
+      "learning_rate": 3.4792537004688575e-05,
+      "loss": 0.3098,
+      "step": 9987
+    },
+    {
+      "epoch": 0.734465769541878,
+      "grad_norm": 0.4932144284248352,
+      "learning_rate": 3.477448250135592e-05,
+      "loss": 0.2996,
+      "step": 9988
+    },
+    {
+      "epoch": 0.7345393043606148,
+      "grad_norm": 0.7059671878814697,
+      "learning_rate": 3.475643169768348e-05,
+      "loss": 0.3358,
+      "step": 9989
+    },
+    {
+      "epoch": 0.7346128391793514,
+      "grad_norm": 0.7591435313224792,
+      "learning_rate": 3.473838459469506e-05,
+      "loss": 0.3189,
+      "step": 9990
+    },
+    {
+      "epoch": 0.7346863739980881,
+      "grad_norm": 0.45743873715400696,
+      "learning_rate": 3.472034119341434e-05,
+      "loss": 0.3088,
+      "step": 9991
+    },
+    {
+      "epoch": 0.7347599088168247,
+      "grad_norm": 0.5724465250968933,
+      "learning_rate": 3.470230149486475e-05,
+      "loss": 0.3155,
+      "step": 9992
+    },
+    {
+      "epoch": 0.7348334436355615,
+      "grad_norm": 0.48154541850090027,
+      "learning_rate": 3.468426550006957e-05,
+      "loss": 0.3134,
+      "step": 9993
+    },
+    {
+      "epoch": 0.7349069784542981,
+      "grad_norm": 0.46424156427383423,
+      "learning_rate": 3.4666233210051744e-05,
+      "loss": 0.2986,
+      "step": 9994
+    },
+    {
+      "epoch": 0.7349805132730348,
+      "grad_norm": 0.666040301322937,
+      "learning_rate": 3.4648204625834124e-05,
+      "loss": 0.3048,
+      "step": 9995
+    },
+    {
+      "epoch": 0.7350540480917714,
+      "grad_norm": 0.4170384407043457,
+      "learning_rate": 3.463017974843937e-05,
+      "loss": 0.3147,
+      "step": 9996
+    },
+    {
+      "epoch": 0.7351275829105082,
+      "grad_norm": 0.6071352958679199,
+      "learning_rate": 3.461215857888975e-05,
+      "loss": 0.302,
+      "step": 9997
+    },
+    {
+      "epoch": 0.7352011177292448,
+      "grad_norm": 0.4251909852027893,
+      "learning_rate": 3.45941411182075e-05,
+      "loss": 0.3171,
+      "step": 9998
+    },
+    {
+      "epoch": 0.7352746525479815,
+      "grad_norm": 0.577324628829956,
+      "learning_rate": 3.4576127367414625e-05,
+      "loss": 0.3155,
+      "step": 9999
+    },
+    {
+      "epoch": 0.7353481873667181,
+      "grad_norm": 0.4614168107509613,
+      "learning_rate": 3.4558117327532804e-05,
+      "loss": 0.3078,
+      "step": 10000
+    },
+    {
+      "epoch": 0.7354217221854548,
+      "grad_norm": 0.5941044092178345,
+      "learning_rate": 3.454011099958364e-05,
+      "loss": 0.3034,
+      "step": 10001
+    },
+    {
+      "epoch": 0.7354952570041915,
+      "grad_norm": 0.4467896819114685,
+      "learning_rate": 3.45221083845885e-05,
+      "loss": 0.3158,
+      "step": 10002
+    },
+    {
+      "epoch": 0.7355687918229281,
+      "grad_norm": 0.6321160197257996,
+      "learning_rate": 3.450410948356844e-05,
+      "loss": 0.3338,
+      "step": 10003
+    },
+    {
+      "epoch": 0.7356423266416648,
+      "grad_norm": 0.46940967440605164,
+      "learning_rate": 3.4486114297544406e-05,
+      "loss": 0.3307,
+      "step": 10004
+    },
+    {
+      "epoch": 0.7357158614604015,
+      "grad_norm": 0.6125577092170715,
+      "learning_rate": 3.446812282753711e-05,
+      "loss": 0.3187,
+      "step": 10005
+    },
+    {
+      "epoch": 0.7357893962791382,
+      "grad_norm": 0.5485306978225708,
+      "learning_rate": 3.445013507456708e-05,
+      "loss": 0.3449,
+      "step": 10006
+    },
+    {
+      "epoch": 0.7358629310978748,
+      "grad_norm": 0.46508827805519104,
+      "learning_rate": 3.443215103965452e-05,
+      "loss": 0.3272,
+      "step": 10007
+    },
+    {
+      "epoch": 0.7359364659166115,
+      "grad_norm": 0.746712327003479,
+      "learning_rate": 3.44141707238196e-05,
+      "loss": 0.3138,
+      "step": 10008
+    },
+    {
+      "epoch": 0.7360100007353482,
+      "grad_norm": 0.532949686050415,
+      "learning_rate": 3.439619412808214e-05,
+      "loss": 0.2967,
+      "step": 10009
+    },
+    {
+      "epoch": 0.7360835355540849,
+      "grad_norm": 0.6242986917495728,
+      "learning_rate": 3.437822125346173e-05,
+      "loss": 0.3215,
+      "step": 10010
+    },
+    {
+      "epoch": 0.7361570703728215,
+      "grad_norm": 0.5394580364227295,
+      "learning_rate": 3.436025210097787e-05,
+      "loss": 0.3228,
+      "step": 10011
+    },
+    {
+      "epoch": 0.7362306051915583,
+      "grad_norm": 0.8371850252151489,
+      "learning_rate": 3.4342286671649815e-05,
+      "loss": 0.3275,
+      "step": 10012
+    },
+    {
+      "epoch": 0.7363041400102949,
+      "grad_norm": 0.7356728315353394,
+      "learning_rate": 3.432432496649652e-05,
+      "loss": 0.306,
+      "step": 10013
+    },
+    {
+      "epoch": 0.7363776748290315,
+      "grad_norm": 0.5488924384117126,
+      "learning_rate": 3.430636698653683e-05,
+      "loss": 0.3372,
+      "step": 10014
+    },
+    {
+      "epoch": 0.7364512096477682,
+      "grad_norm": 0.6774764657020569,
+      "learning_rate": 3.428841273278932e-05,
+      "loss": 0.3107,
+      "step": 10015
+    },
+    {
+      "epoch": 0.7365247444665048,
+      "grad_norm": 0.5456016659736633,
+      "learning_rate": 3.427046220627244e-05,
+      "loss": 0.3149,
+      "step": 10016
+    },
+    {
+      "epoch": 0.7365982792852416,
+      "grad_norm": 0.5583640933036804,
+      "learning_rate": 3.4252515408004257e-05,
+      "loss": 0.3498,
+      "step": 10017
+    },
+    {
+      "epoch": 0.7366718141039782,
+      "grad_norm": 0.6597028970718384,
+      "learning_rate": 3.42345723390028e-05,
+      "loss": 0.3101,
+      "step": 10018
+    },
+    {
+      "epoch": 0.7367453489227149,
+      "grad_norm": 0.4885706305503845,
+      "learning_rate": 3.421663300028587e-05,
+      "loss": 0.3106,
+      "step": 10019
+    },
+    {
+      "epoch": 0.7368188837414515,
+      "grad_norm": 0.9751814603805542,
+      "learning_rate": 3.419869739287087e-05,
+      "loss": 0.308,
+      "step": 10020
+    },
+    {
+      "epoch": 0.7368924185601883,
+      "grad_norm": 0.4035402536392212,
+      "learning_rate": 3.4180765517775195e-05,
+      "loss": 0.3166,
+      "step": 10021
+    },
+    {
+      "epoch": 0.7369659533789249,
+      "grad_norm": 0.984874963760376,
+      "learning_rate": 3.4162837376015996e-05,
+      "loss": 0.3217,
+      "step": 10022
+    },
+    {
+      "epoch": 0.7370394881976616,
+      "grad_norm": 0.48116010427474976,
+      "learning_rate": 3.414491296861011e-05,
+      "loss": 0.2953,
+      "step": 10023
+    },
+    {
+      "epoch": 0.7371130230163982,
+      "grad_norm": 1.1947702169418335,
+      "learning_rate": 3.412699229657425e-05,
+      "loss": 0.3151,
+      "step": 10024
+    },
+    {
+      "epoch": 0.737186557835135,
+      "grad_norm": 0.585121750831604,
+      "learning_rate": 3.410907536092496e-05,
+      "loss": 0.3265,
+      "step": 10025
+    },
+    {
+      "epoch": 0.7372600926538716,
+      "grad_norm": 1.3808531761169434,
+      "learning_rate": 3.40911621626784e-05,
+      "loss": 0.3307,
+      "step": 10026
+    },
+    {
+      "epoch": 0.7373336274726083,
+      "grad_norm": 1.1031168699264526,
+      "learning_rate": 3.40732527028507e-05,
+      "loss": 0.3226,
+      "step": 10027
+    },
+    {
+      "epoch": 0.737407162291345,
+      "grad_norm": 1.1554040908813477,
+      "learning_rate": 3.405534698245766e-05,
+      "loss": 0.301,
+      "step": 10028
+    },
+    {
+      "epoch": 0.7374806971100816,
+      "grad_norm": 0.8877071738243103,
+      "learning_rate": 3.403744500251498e-05,
+      "loss": 0.3144,
+      "step": 10029
+    },
+    {
+      "epoch": 0.7375542319288183,
+      "grad_norm": 0.8070380091667175,
+      "learning_rate": 3.401954676403799e-05,
+      "loss": 0.3191,
+      "step": 10030
+    },
+    {
+      "epoch": 0.7376277667475549,
+      "grad_norm": 0.842049777507782,
+      "learning_rate": 3.400165226804198e-05,
+      "loss": 0.3378,
+      "step": 10031
+    },
+    {
+      "epoch": 0.7377013015662917,
+      "grad_norm": 0.5374936461448669,
+      "learning_rate": 3.39837615155419e-05,
+      "loss": 0.3146,
+      "step": 10032
+    },
+    {
+      "epoch": 0.7377748363850283,
+      "grad_norm": 1.021384835243225,
+      "learning_rate": 3.396587450755251e-05,
+      "loss": 0.3283,
+      "step": 10033
+    },
+    {
+      "epoch": 0.737848371203765,
+      "grad_norm": 0.620278537273407,
+      "learning_rate": 3.39479912450884e-05,
+      "loss": 0.3018,
+      "step": 10034
+    },
+    {
+      "epoch": 0.7379219060225016,
+      "grad_norm": 0.8071222305297852,
+      "learning_rate": 3.393011172916397e-05,
+      "loss": 0.3198,
+      "step": 10035
+    },
+    {
+      "epoch": 0.7379954408412384,
+      "grad_norm": 0.5738754272460938,
+      "learning_rate": 3.391223596079329e-05,
+      "loss": 0.3241,
+      "step": 10036
+    },
+    {
+      "epoch": 0.738068975659975,
+      "grad_norm": 0.5853444337844849,
+      "learning_rate": 3.389436394099034e-05,
+      "loss": 0.3437,
+      "step": 10037
+    },
+    {
+      "epoch": 0.7381425104787117,
+      "grad_norm": 0.4435904026031494,
+      "learning_rate": 3.3876495670768826e-05,
+      "loss": 0.3258,
+      "step": 10038
+    },
+    {
+      "epoch": 0.7382160452974483,
+      "grad_norm": 0.4264938235282898,
+      "learning_rate": 3.3858631151142295e-05,
+      "loss": 0.3503,
+      "step": 10039
+    },
+    {
+      "epoch": 0.7382895801161851,
+      "grad_norm": 0.45124751329421997,
+      "learning_rate": 3.3840770383123974e-05,
+      "loss": 0.3399,
+      "step": 10040
+    },
+    {
+      "epoch": 0.7383631149349217,
+      "grad_norm": 0.41816478967666626,
+      "learning_rate": 3.3822913367726974e-05,
+      "loss": 0.3088,
+      "step": 10041
+    },
+    {
+      "epoch": 0.7384366497536583,
+      "grad_norm": 0.48556655645370483,
+      "learning_rate": 3.3805060105964204e-05,
+      "loss": 0.3374,
+      "step": 10042
+    },
+    {
+      "epoch": 0.738510184572395,
+      "grad_norm": 0.4177401065826416,
+      "learning_rate": 3.3787210598848284e-05,
+      "loss": 0.331,
+      "step": 10043
+    },
+    {
+      "epoch": 0.7385837193911317,
+      "grad_norm": 0.42479208111763,
+      "learning_rate": 3.3769364847391626e-05,
+      "loss": 0.333,
+      "step": 10044
+    },
+    {
+      "epoch": 0.7386572542098684,
+      "grad_norm": 0.4634307622909546,
+      "learning_rate": 3.375152285260651e-05,
+      "loss": 0.3167,
+      "step": 10045
+    },
+    {
+      "epoch": 0.738730789028605,
+      "grad_norm": 0.44740739464759827,
+      "learning_rate": 3.3733684615504916e-05,
+      "loss": 0.319,
+      "step": 10046
+    },
+    {
+      "epoch": 0.7388043238473417,
+      "grad_norm": 0.5201365351676941,
+      "learning_rate": 3.371585013709865e-05,
+      "loss": 0.373,
+      "step": 10047
+    },
+    {
+      "epoch": 0.7388778586660784,
+      "grad_norm": 0.4712975025177002,
+      "learning_rate": 3.3698019418399365e-05,
+      "loss": 0.3045,
+      "step": 10048
+    },
+    {
+      "epoch": 0.7389513934848151,
+      "grad_norm": 0.4864865839481354,
+      "learning_rate": 3.368019246041834e-05,
+      "loss": 0.3278,
+      "step": 10049
+    },
+    {
+      "epoch": 0.7390249283035517,
+      "grad_norm": 0.48493221402168274,
+      "learning_rate": 3.3662369264166794e-05,
+      "loss": 0.3267,
+      "step": 10050
+    },
+    {
+      "epoch": 0.7390984631222884,
+      "grad_norm": 0.41076770424842834,
+      "learning_rate": 3.364454983065568e-05,
+      "loss": 0.2999,
+      "step": 10051
+    },
+    {
+      "epoch": 0.7391719979410251,
+      "grad_norm": 0.4457550644874573,
+      "learning_rate": 3.362673416089576e-05,
+      "loss": 0.3231,
+      "step": 10052
+    },
+    {
+      "epoch": 0.7392455327597618,
+      "grad_norm": 0.555755615234375,
+      "learning_rate": 3.360892225589748e-05,
+      "loss": 0.319,
+      "step": 10053
+    },
+    {
+      "epoch": 0.7393190675784984,
+      "grad_norm": 0.39492011070251465,
+      "learning_rate": 3.359111411667123e-05,
+      "loss": 0.3398,
+      "step": 10054
+    },
+    {
+      "epoch": 0.7393926023972351,
+      "grad_norm": 0.5070074200630188,
+      "learning_rate": 3.357330974422708e-05,
+      "loss": 0.3265,
+      "step": 10055
+    },
+    {
+      "epoch": 0.7394661372159718,
+      "grad_norm": 0.42292603850364685,
+      "learning_rate": 3.355550913957487e-05,
+      "loss": 0.3336,
+      "step": 10056
+    },
+    {
+      "epoch": 0.7395396720347084,
+      "grad_norm": 0.4987037777900696,
+      "learning_rate": 3.35377123037243e-05,
+      "loss": 0.3016,
+      "step": 10057
+    },
+    {
+      "epoch": 0.7396132068534451,
+      "grad_norm": 0.43377384543418884,
+      "learning_rate": 3.351991923768486e-05,
+      "loss": 0.3406,
+      "step": 10058
+    },
+    {
+      "epoch": 0.7396867416721817,
+      "grad_norm": 0.533263087272644,
+      "learning_rate": 3.3502129942465735e-05,
+      "loss": 0.315,
+      "step": 10059
+    },
+    {
+      "epoch": 0.7397602764909185,
+      "grad_norm": 0.5662903189659119,
+      "learning_rate": 3.348434441907596e-05,
+      "loss": 0.3208,
+      "step": 10060
+    },
+    {
+      "epoch": 0.7398338113096551,
+      "grad_norm": 0.7723186612129211,
+      "learning_rate": 3.346656266852442e-05,
+      "loss": 0.3273,
+      "step": 10061
+    },
+    {
+      "epoch": 0.7399073461283918,
+      "grad_norm": 0.40989232063293457,
+      "learning_rate": 3.3448784691819615e-05,
+      "loss": 0.2936,
+      "step": 10062
+    },
+    {
+      "epoch": 0.7399808809471284,
+      "grad_norm": 0.7740196585655212,
+      "learning_rate": 3.343101048996998e-05,
+      "loss": 0.3245,
+      "step": 10063
+    },
+    {
+      "epoch": 0.7400544157658652,
+      "grad_norm": 0.4444553554058075,
+      "learning_rate": 3.3413240063983686e-05,
+      "loss": 0.3359,
+      "step": 10064
+    },
+    {
+      "epoch": 0.7401279505846018,
+      "grad_norm": 0.6335516571998596,
+      "learning_rate": 3.3395473414868716e-05,
+      "loss": 0.3281,
+      "step": 10065
+    },
+    {
+      "epoch": 0.7402014854033385,
+      "grad_norm": 0.4378620684146881,
+      "learning_rate": 3.337771054363279e-05,
+      "loss": 0.3055,
+      "step": 10066
+    },
+    {
+      "epoch": 0.7402750202220751,
+      "grad_norm": 0.9696934819221497,
+      "learning_rate": 3.335995145128339e-05,
+      "loss": 0.3291,
+      "step": 10067
+    },
+    {
+      "epoch": 0.7403485550408119,
+      "grad_norm": 0.48451513051986694,
+      "learning_rate": 3.334219613882791e-05,
+      "loss": 0.3002,
+      "step": 10068
+    },
+    {
+      "epoch": 0.7404220898595485,
+      "grad_norm": 1.2159574031829834,
+      "learning_rate": 3.332444460727339e-05,
+      "loss": 0.327,
+      "step": 10069
+    },
+    {
+      "epoch": 0.7404956246782851,
+      "grad_norm": 0.4683122932910919,
+      "learning_rate": 3.330669685762672e-05,
+      "loss": 0.3174,
+      "step": 10070
+    },
+    {
+      "epoch": 0.7405691594970218,
+      "grad_norm": 0.742274284362793,
+      "learning_rate": 3.328895289089462e-05,
+      "loss": 0.2982,
+      "step": 10071
+    },
+    {
+      "epoch": 0.7406426943157585,
+      "grad_norm": 0.6192287802696228,
+      "learning_rate": 3.327121270808349e-05,
+      "loss": 0.3317,
+      "step": 10072
+    },
+    {
+      "epoch": 0.7407162291344952,
+      "grad_norm": 0.4915502965450287,
+      "learning_rate": 3.3253476310199597e-05,
+      "loss": 0.3189,
+      "step": 10073
+    },
+    {
+      "epoch": 0.7407897639532318,
+      "grad_norm": 0.7386595010757446,
+      "learning_rate": 3.323574369824897e-05,
+      "loss": 0.3289,
+      "step": 10074
+    },
+    {
+      "epoch": 0.7408632987719685,
+      "grad_norm": 0.41749119758605957,
+      "learning_rate": 3.321801487323745e-05,
+      "loss": 0.3109,
+      "step": 10075
+    },
+    {
+      "epoch": 0.7409368335907052,
+      "grad_norm": 0.7794691920280457,
+      "learning_rate": 3.3200289836170574e-05,
+      "loss": 0.2992,
+      "step": 10076
+    },
+    {
+      "epoch": 0.7410103684094419,
+      "grad_norm": 0.4041176438331604,
+      "learning_rate": 3.318256858805375e-05,
+      "loss": 0.3221,
+      "step": 10077
+    },
+    {
+      "epoch": 0.7410839032281785,
+      "grad_norm": 0.7584613561630249,
+      "learning_rate": 3.3164851129892215e-05,
+      "loss": 0.327,
+      "step": 10078
+    },
+    {
+      "epoch": 0.7411574380469153,
+      "grad_norm": 0.5594965219497681,
+      "learning_rate": 3.31471374626908e-05,
+      "loss": 0.3197,
+      "step": 10079
+    },
+    {
+      "epoch": 0.7412309728656519,
+      "grad_norm": 0.6011794805526733,
+      "learning_rate": 3.3129427587454295e-05,
+      "loss": 0.3059,
+      "step": 10080
+    },
+    {
+      "epoch": 0.7413045076843886,
+      "grad_norm": 0.794447124004364,
+      "learning_rate": 3.311172150518726e-05,
+      "loss": 0.3057,
+      "step": 10081
+    },
+    {
+      "epoch": 0.7413780425031252,
+      "grad_norm": 0.638353168964386,
+      "learning_rate": 3.3094019216893944e-05,
+      "loss": 0.3202,
+      "step": 10082
+    },
+    {
+      "epoch": 0.741451577321862,
+      "grad_norm": 0.987686276435852,
+      "learning_rate": 3.3076320723578466e-05,
+      "loss": 0.3747,
+      "step": 10083
+    },
+    {
+      "epoch": 0.7415251121405986,
+      "grad_norm": 0.5530634522438049,
+      "learning_rate": 3.305862602624473e-05,
+      "loss": 0.3466,
+      "step": 10084
+    },
+    {
+      "epoch": 0.7415986469593352,
+      "grad_norm": 1.1262869834899902,
+      "learning_rate": 3.304093512589634e-05,
+      "loss": 0.3848,
+      "step": 10085
+    },
+    {
+      "epoch": 0.7416721817780719,
+      "grad_norm": 0.47100600600242615,
+      "learning_rate": 3.302324802353678e-05,
+      "loss": 0.3236,
+      "step": 10086
+    },
+    {
+      "epoch": 0.7417457165968085,
+      "grad_norm": 0.9150334596633911,
+      "learning_rate": 3.300556472016928e-05,
+      "loss": 0.3105,
+      "step": 10087
+    },
+    {
+      "epoch": 0.7418192514155453,
+      "grad_norm": 0.43377527594566345,
+      "learning_rate": 3.298788521679688e-05,
+      "loss": 0.3392,
+      "step": 10088
+    },
+    {
+      "epoch": 0.7418927862342819,
+      "grad_norm": 0.5491644144058228,
+      "learning_rate": 3.297020951442231e-05,
+      "loss": 0.3487,
+      "step": 10089
+    },
+    {
+      "epoch": 0.7419663210530186,
+      "grad_norm": 0.49921298027038574,
+      "learning_rate": 3.295253761404824e-05,
+      "loss": 0.3254,
+      "step": 10090
+    },
+    {
+      "epoch": 0.7420398558717552,
+      "grad_norm": 0.4414997398853302,
+      "learning_rate": 3.293486951667698e-05,
+      "loss": 0.3244,
+      "step": 10091
+    },
+    {
+      "epoch": 0.742113390690492,
+      "grad_norm": 0.6026118397712708,
+      "learning_rate": 3.291720522331068e-05,
+      "loss": 0.3431,
+      "step": 10092
+    },
+    {
+      "epoch": 0.7421869255092286,
+      "grad_norm": 0.41518235206604004,
+      "learning_rate": 3.289954473495129e-05,
+      "loss": 0.3245,
+      "step": 10093
+    },
+    {
+      "epoch": 0.7422604603279653,
+      "grad_norm": 0.8247800469398499,
+      "learning_rate": 3.2881888052600565e-05,
+      "loss": 0.3286,
+      "step": 10094
+    },
+    {
+      "epoch": 0.742333995146702,
+      "grad_norm": 0.5349140763282776,
+      "learning_rate": 3.286423517725995e-05,
+      "loss": 0.3232,
+      "step": 10095
+    },
+    {
+      "epoch": 0.7424075299654387,
+      "grad_norm": 0.4402865767478943,
+      "learning_rate": 3.2846586109930775e-05,
+      "loss": 0.2997,
+      "step": 10096
+    },
+    {
+      "epoch": 0.7424810647841753,
+      "grad_norm": 0.49878621101379395,
+      "learning_rate": 3.28289408516141e-05,
+      "loss": 0.3291,
+      "step": 10097
+    },
+    {
+      "epoch": 0.7425545996029119,
+      "grad_norm": 0.7739441990852356,
+      "learning_rate": 3.281129940331083e-05,
+      "loss": 0.3226,
+      "step": 10098
+    },
+    {
+      "epoch": 0.7426281344216487,
+      "grad_norm": 0.4460623264312744,
+      "learning_rate": 3.279366176602151e-05,
+      "loss": 0.3379,
+      "step": 10099
+    },
+    {
+      "epoch": 0.7427016692403853,
+      "grad_norm": 0.4435238242149353,
+      "learning_rate": 3.277602794074663e-05,
+      "loss": 0.2918,
+      "step": 10100
+    },
+    {
+      "epoch": 0.742775204059122,
+      "grad_norm": 0.4376389682292938,
+      "learning_rate": 3.275839792848646e-05,
+      "loss": 0.3309,
+      "step": 10101
+    },
+    {
+      "epoch": 0.7428487388778586,
+      "grad_norm": 0.46337807178497314,
+      "learning_rate": 3.274077173024086e-05,
+      "loss": 0.3293,
+      "step": 10102
+    },
+    {
+      "epoch": 0.7429222736965954,
+      "grad_norm": 0.4121059775352478,
+      "learning_rate": 3.2723149347009665e-05,
+      "loss": 0.3344,
+      "step": 10103
+    },
+    {
+      "epoch": 0.742995808515332,
+      "grad_norm": 0.39614781737327576,
+      "learning_rate": 3.270553077979247e-05,
+      "loss": 0.3334,
+      "step": 10104
+    },
+    {
+      "epoch": 0.7430693433340687,
+      "grad_norm": 0.5104303359985352,
+      "learning_rate": 3.268791602958856e-05,
+      "loss": 0.3303,
+      "step": 10105
+    },
+    {
+      "epoch": 0.7431428781528053,
+      "grad_norm": 0.459448903799057,
+      "learning_rate": 3.267030509739709e-05,
+      "loss": 0.3433,
+      "step": 10106
+    },
+    {
+      "epoch": 0.7432164129715421,
+      "grad_norm": 0.41692522168159485,
+      "learning_rate": 3.265269798421699e-05,
+      "loss": 0.322,
+      "step": 10107
+    },
+    {
+      "epoch": 0.7432899477902787,
+      "grad_norm": 0.6991991400718689,
+      "learning_rate": 3.2635094691046916e-05,
+      "loss": 0.3118,
+      "step": 10108
+    },
+    {
+      "epoch": 0.7433634826090154,
+      "grad_norm": 0.3763180077075958,
+      "learning_rate": 3.2617495218885365e-05,
+      "loss": 0.2886,
+      "step": 10109
+    },
+    {
+      "epoch": 0.743437017427752,
+      "grad_norm": 0.5760794878005981,
+      "learning_rate": 3.2599899568730586e-05,
+      "loss": 0.294,
+      "step": 10110
+    },
+    {
+      "epoch": 0.7435105522464888,
+      "grad_norm": 0.5332213044166565,
+      "learning_rate": 3.258230774158067e-05,
+      "loss": 0.3189,
+      "step": 10111
+    },
+    {
+      "epoch": 0.7435840870652254,
+      "grad_norm": 0.6370073556900024,
+      "learning_rate": 3.256471973843338e-05,
+      "loss": 0.2953,
+      "step": 10112
+    },
+    {
+      "epoch": 0.743657621883962,
+      "grad_norm": 0.5073726177215576,
+      "learning_rate": 3.254713556028639e-05,
+      "loss": 0.3168,
+      "step": 10113
+    },
+    {
+      "epoch": 0.7437311567026987,
+      "grad_norm": 0.47853752970695496,
+      "learning_rate": 3.2529555208137054e-05,
+      "loss": 0.3002,
+      "step": 10114
+    },
+    {
+      "epoch": 0.7438046915214354,
+      "grad_norm": 0.5615386366844177,
+      "learning_rate": 3.2511978682982516e-05,
+      "loss": 0.3199,
+      "step": 10115
+    },
+    {
+      "epoch": 0.7438782263401721,
+      "grad_norm": 0.4452940821647644,
+      "learning_rate": 3.249440598581976e-05,
+      "loss": 0.3219,
+      "step": 10116
+    },
+    {
+      "epoch": 0.7439517611589087,
+      "grad_norm": 0.48771750926971436,
+      "learning_rate": 3.247683711764559e-05,
+      "loss": 0.3219,
+      "step": 10117
+    },
+    {
+      "epoch": 0.7440252959776454,
+      "grad_norm": 0.49034279584884644,
+      "learning_rate": 3.2459272079456427e-05,
+      "loss": 0.3152,
+      "step": 10118
+    },
+    {
+      "epoch": 0.7440988307963821,
+      "grad_norm": 0.678661584854126,
+      "learning_rate": 3.244171087224864e-05,
+      "loss": 0.296,
+      "step": 10119
+    },
+    {
+      "epoch": 0.7441723656151188,
+      "grad_norm": 0.4985355734825134,
+      "learning_rate": 3.2424153497018305e-05,
+      "loss": 0.3031,
+      "step": 10120
+    },
+    {
+      "epoch": 0.7442459004338554,
+      "grad_norm": 0.8295179009437561,
+      "learning_rate": 3.240659995476134e-05,
+      "loss": 0.3269,
+      "step": 10121
+    },
+    {
+      "epoch": 0.7443194352525921,
+      "grad_norm": 0.49996310472488403,
+      "learning_rate": 3.238905024647332e-05,
+      "loss": 0.3069,
+      "step": 10122
+    },
+    {
+      "epoch": 0.7443929700713288,
+      "grad_norm": 0.433904230594635,
+      "learning_rate": 3.237150437314972e-05,
+      "loss": 0.3175,
+      "step": 10123
+    },
+    {
+      "epoch": 0.7444665048900655,
+      "grad_norm": 0.9562931656837463,
+      "learning_rate": 3.235396233578581e-05,
+      "loss": 0.3448,
+      "step": 10124
+    },
+    {
+      "epoch": 0.7445400397088021,
+      "grad_norm": 0.5063918232917786,
+      "learning_rate": 3.233642413537653e-05,
+      "loss": 0.3055,
+      "step": 10125
+    },
+    {
+      "epoch": 0.7446135745275387,
+      "grad_norm": 0.8739487528800964,
+      "learning_rate": 3.231888977291666e-05,
+      "loss": 0.3322,
+      "step": 10126
+    },
+    {
+      "epoch": 0.7446871093462755,
+      "grad_norm": 0.49363887310028076,
+      "learning_rate": 3.230135924940081e-05,
+      "loss": 0.3063,
+      "step": 10127
+    },
+    {
+      "epoch": 0.7447606441650121,
+      "grad_norm": 0.5610170364379883,
+      "learning_rate": 3.228383256582329e-05,
+      "loss": 0.3135,
+      "step": 10128
+    },
+    {
+      "epoch": 0.7448341789837488,
+      "grad_norm": 0.5196980237960815,
+      "learning_rate": 3.226630972317824e-05,
+      "loss": 0.3626,
+      "step": 10129
+    },
+    {
+      "epoch": 0.7449077138024854,
+      "grad_norm": 0.4404999911785126,
+      "learning_rate": 3.224879072245962e-05,
+      "loss": 0.3011,
+      "step": 10130
+    },
+    {
+      "epoch": 0.7449812486212222,
+      "grad_norm": 0.5421916842460632,
+      "learning_rate": 3.223127556466106e-05,
+      "loss": 0.3164,
+      "step": 10131
+    },
+    {
+      "epoch": 0.7450547834399588,
+      "grad_norm": 0.5185849070549011,
+      "learning_rate": 3.2213764250776065e-05,
+      "loss": 0.3025,
+      "step": 10132
+    },
+    {
+      "epoch": 0.7451283182586955,
+      "grad_norm": 0.4397890865802765,
+      "learning_rate": 3.219625678179791e-05,
+      "loss": 0.2981,
+      "step": 10133
+    },
+    {
+      "epoch": 0.7452018530774321,
+      "grad_norm": 0.42630696296691895,
+      "learning_rate": 3.2178753158719645e-05,
+      "loss": 0.3164,
+      "step": 10134
+    },
+    {
+      "epoch": 0.7452753878961689,
+      "grad_norm": 0.5276486277580261,
+      "learning_rate": 3.216125338253405e-05,
+      "loss": 0.3234,
+      "step": 10135
+    },
+    {
+      "epoch": 0.7453489227149055,
+      "grad_norm": 0.521729052066803,
+      "learning_rate": 3.214375745423379e-05,
+      "loss": 0.3076,
+      "step": 10136
+    },
+    {
+      "epoch": 0.7454224575336422,
+      "grad_norm": 0.4487251937389374,
+      "learning_rate": 3.2126265374811214e-05,
+      "loss": 0.3665,
+      "step": 10137
+    },
+    {
+      "epoch": 0.7454959923523788,
+      "grad_norm": 0.5915623903274536,
+      "learning_rate": 3.210877714525847e-05,
+      "loss": 0.3514,
+      "step": 10138
+    },
+    {
+      "epoch": 0.7455695271711156,
+      "grad_norm": 0.5053547620773315,
+      "learning_rate": 3.209129276656753e-05,
+      "loss": 0.3254,
+      "step": 10139
+    },
+    {
+      "epoch": 0.7456430619898522,
+      "grad_norm": 0.5567373633384705,
+      "learning_rate": 3.207381223973016e-05,
+      "loss": 0.3187,
+      "step": 10140
+    },
+    {
+      "epoch": 0.7457165968085888,
+      "grad_norm": 0.4785602390766144,
+      "learning_rate": 3.2056335565737814e-05,
+      "loss": 0.3125,
+      "step": 10141
+    },
+    {
+      "epoch": 0.7457901316273255,
+      "grad_norm": 0.7830932140350342,
+      "learning_rate": 3.203886274558183e-05,
+      "loss": 0.3157,
+      "step": 10142
+    },
+    {
+      "epoch": 0.7458636664460622,
+      "grad_norm": 0.46483802795410156,
+      "learning_rate": 3.202139378025326e-05,
+      "loss": 0.3308,
+      "step": 10143
+    },
+    {
+      "epoch": 0.7459372012647989,
+      "grad_norm": 0.835345447063446,
+      "learning_rate": 3.200392867074302e-05,
+      "loss": 0.3089,
+      "step": 10144
+    },
+    {
+      "epoch": 0.7460107360835355,
+      "grad_norm": 0.4497796595096588,
+      "learning_rate": 3.1986467418041665e-05,
+      "loss": 0.2993,
+      "step": 10145
+    },
+    {
+      "epoch": 0.7460842709022723,
+      "grad_norm": 0.7621421813964844,
+      "learning_rate": 3.196901002313966e-05,
+      "loss": 0.3164,
+      "step": 10146
+    },
+    {
+      "epoch": 0.7461578057210089,
+      "grad_norm": 0.7424991130828857,
+      "learning_rate": 3.195155648702722e-05,
+      "loss": 0.3049,
+      "step": 10147
+    },
+    {
+      "epoch": 0.7462313405397456,
+      "grad_norm": 0.7245610356330872,
+      "learning_rate": 3.193410681069433e-05,
+      "loss": 0.3133,
+      "step": 10148
+    },
+    {
+      "epoch": 0.7463048753584822,
+      "grad_norm": 0.7114962935447693,
+      "learning_rate": 3.1916660995130684e-05,
+      "loss": 0.3194,
+      "step": 10149
+    },
+    {
+      "epoch": 0.746378410177219,
+      "grad_norm": 0.717814028263092,
+      "learning_rate": 3.1899219041325924e-05,
+      "loss": 0.3285,
+      "step": 10150
+    },
+    {
+      "epoch": 0.7464519449959556,
+      "grad_norm": 0.5798618197441101,
+      "learning_rate": 3.188178095026929e-05,
+      "loss": 0.3222,
+      "step": 10151
+    },
+    {
+      "epoch": 0.7465254798146923,
+      "grad_norm": 0.4943339228630066,
+      "learning_rate": 3.186434672294993e-05,
+      "loss": 0.3173,
+      "step": 10152
+    },
+    {
+      "epoch": 0.7465990146334289,
+      "grad_norm": 0.8797395825386047,
+      "learning_rate": 3.184691636035677e-05,
+      "loss": 0.3635,
+      "step": 10153
+    },
+    {
+      "epoch": 0.7466725494521655,
+      "grad_norm": 0.563988447189331,
+      "learning_rate": 3.1829489863478404e-05,
+      "loss": 0.3183,
+      "step": 10154
+    },
+    {
+      "epoch": 0.7467460842709023,
+      "grad_norm": 0.7561256885528564,
+      "learning_rate": 3.181206723330331e-05,
+      "loss": 0.312,
+      "step": 10155
+    },
+    {
+      "epoch": 0.7468196190896389,
+      "grad_norm": 0.45153531432151794,
+      "learning_rate": 3.179464847081972e-05,
+      "loss": 0.324,
+      "step": 10156
+    },
+    {
+      "epoch": 0.7468931539083756,
+      "grad_norm": 0.8295782208442688,
+      "learning_rate": 3.17772335770157e-05,
+      "loss": 0.3055,
+      "step": 10157
+    },
+    {
+      "epoch": 0.7469666887271122,
+      "grad_norm": 0.4782620370388031,
+      "learning_rate": 3.175982255287894e-05,
+      "loss": 0.3338,
+      "step": 10158
+    },
+    {
+      "epoch": 0.747040223545849,
+      "grad_norm": 0.7815575003623962,
+      "learning_rate": 3.1742415399397065e-05,
+      "loss": 0.3155,
+      "step": 10159
+    },
+    {
+      "epoch": 0.7471137583645856,
+      "grad_norm": 0.5246520042419434,
+      "learning_rate": 3.17250121175575e-05,
+      "loss": 0.3014,
+      "step": 10160
+    },
+    {
+      "epoch": 0.7471872931833223,
+      "grad_norm": 0.7085018754005432,
+      "learning_rate": 3.1707612708347213e-05,
+      "loss": 0.3303,
+      "step": 10161
+    },
+    {
+      "epoch": 0.747260828002059,
+      "grad_norm": 0.6450973749160767,
+      "learning_rate": 3.169021717275322e-05,
+      "loss": 0.3497,
+      "step": 10162
+    },
+    {
+      "epoch": 0.7473343628207957,
+      "grad_norm": 0.651629626750946,
+      "learning_rate": 3.167282551176223e-05,
+      "loss": 0.3406,
+      "step": 10163
+    },
+    {
+      "epoch": 0.7474078976395323,
+      "grad_norm": 0.4638863503932953,
+      "learning_rate": 3.165543772636066e-05,
+      "loss": 0.3226,
+      "step": 10164
+    },
+    {
+      "epoch": 0.747481432458269,
+      "grad_norm": 0.46578460931777954,
+      "learning_rate": 3.163805381753478e-05,
+      "loss": 0.31,
+      "step": 10165
+    },
+    {
+      "epoch": 0.7475549672770057,
+      "grad_norm": 0.7172430157661438,
+      "learning_rate": 3.1620673786270625e-05,
+      "loss": 0.3123,
+      "step": 10166
+    },
+    {
+      "epoch": 0.7476285020957424,
+      "grad_norm": 0.39498621225357056,
+      "learning_rate": 3.160329763355406e-05,
+      "loss": 0.3216,
+      "step": 10167
+    },
+    {
+      "epoch": 0.747702036914479,
+      "grad_norm": 0.8077290058135986,
+      "learning_rate": 3.158592536037061e-05,
+      "loss": 0.3284,
+      "step": 10168
+    },
+    {
+      "epoch": 0.7477755717332156,
+      "grad_norm": 0.41470640897750854,
+      "learning_rate": 3.1568556967705666e-05,
+      "loss": 0.3058,
+      "step": 10169
+    },
+    {
+      "epoch": 0.7478491065519524,
+      "grad_norm": 0.5661654472351074,
+      "learning_rate": 3.155119245654442e-05,
+      "loss": 0.3388,
+      "step": 10170
+    },
+    {
+      "epoch": 0.747922641370689,
+      "grad_norm": 0.3801719844341278,
+      "learning_rate": 3.153383182787177e-05,
+      "loss": 0.3053,
+      "step": 10171
+    },
+    {
+      "epoch": 0.7479961761894257,
+      "grad_norm": 0.48732927441596985,
+      "learning_rate": 3.1516475082672415e-05,
+      "loss": 0.2951,
+      "step": 10172
+    },
+    {
+      "epoch": 0.7480697110081623,
+      "grad_norm": 0.49019983410835266,
+      "learning_rate": 3.149912222193089e-05,
+      "loss": 0.355,
+      "step": 10173
+    },
+    {
+      "epoch": 0.7481432458268991,
+      "grad_norm": 0.4377724528312683,
+      "learning_rate": 3.148177324663141e-05,
+      "loss": 0.323,
+      "step": 10174
+    },
+    {
+      "epoch": 0.7482167806456357,
+      "grad_norm": 0.4562240242958069,
+      "learning_rate": 3.146442815775806e-05,
+      "loss": 0.3268,
+      "step": 10175
+    },
+    {
+      "epoch": 0.7482903154643724,
+      "grad_norm": 0.4380819499492645,
+      "learning_rate": 3.1447086956294716e-05,
+      "loss": 0.361,
+      "step": 10176
+    },
+    {
+      "epoch": 0.748363850283109,
+      "grad_norm": 0.572411835193634,
+      "learning_rate": 3.142974964322492e-05,
+      "loss": 0.3455,
+      "step": 10177
+    },
+    {
+      "epoch": 0.7484373851018458,
+      "grad_norm": 0.40827223658561707,
+      "learning_rate": 3.141241621953207e-05,
+      "loss": 0.3088,
+      "step": 10178
+    },
+    {
+      "epoch": 0.7485109199205824,
+      "grad_norm": 0.48491087555885315,
+      "learning_rate": 3.1395086686199357e-05,
+      "loss": 0.3322,
+      "step": 10179
+    },
+    {
+      "epoch": 0.7485844547393191,
+      "grad_norm": 0.4588286280632019,
+      "learning_rate": 3.1377761044209776e-05,
+      "loss": 0.3464,
+      "step": 10180
+    },
+    {
+      "epoch": 0.7486579895580557,
+      "grad_norm": 0.4956115186214447,
+      "learning_rate": 3.136043929454596e-05,
+      "loss": 0.3205,
+      "step": 10181
+    },
+    {
+      "epoch": 0.7487315243767924,
+      "grad_norm": 0.4424450993537903,
+      "learning_rate": 3.134312143819048e-05,
+      "loss": 0.3151,
+      "step": 10182
+    },
+    {
+      "epoch": 0.7488050591955291,
+      "grad_norm": 0.7747414112091064,
+      "learning_rate": 3.1325807476125666e-05,
+      "loss": 0.3214,
+      "step": 10183
+    },
+    {
+      "epoch": 0.7488785940142657,
+      "grad_norm": 0.4710461497306824,
+      "learning_rate": 3.130849740933346e-05,
+      "loss": 0.3233,
+      "step": 10184
+    },
+    {
+      "epoch": 0.7489521288330024,
+      "grad_norm": 0.7111046314239502,
+      "learning_rate": 3.129119123879578e-05,
+      "loss": 0.3086,
+      "step": 10185
+    },
+    {
+      "epoch": 0.7490256636517391,
+      "grad_norm": 0.5070613026618958,
+      "learning_rate": 3.127388896549427e-05,
+      "loss": 0.3245,
+      "step": 10186
+    },
+    {
+      "epoch": 0.7490991984704758,
+      "grad_norm": 0.5250900983810425,
+      "learning_rate": 3.125659059041028e-05,
+      "loss": 0.3201,
+      "step": 10187
+    },
+    {
+      "epoch": 0.7491727332892124,
+      "grad_norm": 0.5881357192993164,
+      "learning_rate": 3.123929611452502e-05,
+      "loss": 0.3176,
+      "step": 10188
+    },
+    {
+      "epoch": 0.7492462681079491,
+      "grad_norm": 0.3758205473423004,
+      "learning_rate": 3.1222005538819444e-05,
+      "loss": 0.3037,
+      "step": 10189
+    },
+    {
+      "epoch": 0.7493198029266858,
+      "grad_norm": 0.561244010925293,
+      "learning_rate": 3.120471886427434e-05,
+      "loss": 0.3049,
+      "step": 10190
+    },
+    {
+      "epoch": 0.7493933377454225,
+      "grad_norm": 0.5183084011077881,
+      "learning_rate": 3.1187436091870146e-05,
+      "loss": 0.306,
+      "step": 10191
+    },
+    {
+      "epoch": 0.7494668725641591,
+      "grad_norm": 0.6505194306373596,
+      "learning_rate": 3.117015722258719e-05,
+      "loss": 0.3209,
+      "step": 10192
+    },
+    {
+      "epoch": 0.7495404073828958,
+      "grad_norm": 0.6810802817344666,
+      "learning_rate": 3.11528822574056e-05,
+      "loss": 0.3205,
+      "step": 10193
+    },
+    {
+      "epoch": 0.7496139422016325,
+      "grad_norm": 0.7695670127868652,
+      "learning_rate": 3.113561119730516e-05,
+      "loss": 0.342,
+      "step": 10194
+    },
+    {
+      "epoch": 0.7496874770203692,
+      "grad_norm": 0.5762351751327515,
+      "learning_rate": 3.1118344043265545e-05,
+      "loss": 0.3225,
+      "step": 10195
+    },
+    {
+      "epoch": 0.7497610118391058,
+      "grad_norm": 0.6479451060295105,
+      "learning_rate": 3.1101080796266154e-05,
+      "loss": 0.3137,
+      "step": 10196
+    },
+    {
+      "epoch": 0.7498345466578424,
+      "grad_norm": 0.5483207106590271,
+      "learning_rate": 3.1083821457286145e-05,
+      "loss": 0.3029,
+      "step": 10197
+    },
+    {
+      "epoch": 0.7499080814765792,
+      "grad_norm": 0.7277947664260864,
+      "learning_rate": 3.10665660273045e-05,
+      "loss": 0.3472,
+      "step": 10198
+    },
+    {
+      "epoch": 0.7499816162953158,
+      "grad_norm": 0.5600482225418091,
+      "learning_rate": 3.104931450730002e-05,
+      "loss": 0.3339,
+      "step": 10199
+    },
+    {
+      "epoch": 0.7500551511140525,
+      "grad_norm": 0.5252562761306763,
+      "learning_rate": 3.103206689825116e-05,
+      "loss": 0.3013,
+      "step": 10200
+    },
+    {
+      "epoch": 0.7501286859327891,
+      "grad_norm": 0.6101149916648865,
+      "learning_rate": 3.101482320113624e-05,
+      "loss": 0.3523,
+      "step": 10201
+    },
+    {
+      "epoch": 0.7502022207515259,
+      "grad_norm": 0.3909684121608734,
+      "learning_rate": 3.0997583416933364e-05,
+      "loss": 0.3312,
+      "step": 10202
+    },
+    {
+      "epoch": 0.7502757555702625,
+      "grad_norm": 0.6834962964057922,
+      "learning_rate": 3.09803475466204e-05,
+      "loss": 0.3091,
+      "step": 10203
+    },
+    {
+      "epoch": 0.7503492903889992,
+      "grad_norm": 0.4209718406200409,
+      "learning_rate": 3.096311559117493e-05,
+      "loss": 0.2983,
+      "step": 10204
+    },
+    {
+      "epoch": 0.7504228252077358,
+      "grad_norm": 0.5767913460731506,
+      "learning_rate": 3.094588755157441e-05,
+      "loss": 0.3197,
+      "step": 10205
+    },
+    {
+      "epoch": 0.7504963600264726,
+      "grad_norm": 0.48430168628692627,
+      "learning_rate": 3.092866342879608e-05,
+      "loss": 0.3219,
+      "step": 10206
+    },
+    {
+      "epoch": 0.7505698948452092,
+      "grad_norm": 0.47225192189216614,
+      "learning_rate": 3.0911443223816796e-05,
+      "loss": 0.3182,
+      "step": 10207
+    },
+    {
+      "epoch": 0.7506434296639459,
+      "grad_norm": 0.4849843978881836,
+      "learning_rate": 3.089422693761336e-05,
+      "loss": 0.3475,
+      "step": 10208
+    },
+    {
+      "epoch": 0.7507169644826825,
+      "grad_norm": 0.5246581435203552,
+      "learning_rate": 3.087701457116232e-05,
+      "loss": 0.3205,
+      "step": 10209
+    },
+    {
+      "epoch": 0.7507904993014192,
+      "grad_norm": 0.6543416976928711,
+      "learning_rate": 3.085980612543995e-05,
+      "loss": 0.3328,
+      "step": 10210
+    },
+    {
+      "epoch": 0.7508640341201559,
+      "grad_norm": 0.44081172347068787,
+      "learning_rate": 3.084260160142234e-05,
+      "loss": 0.2939,
+      "step": 10211
+    },
+    {
+      "epoch": 0.7509375689388925,
+      "grad_norm": 0.47760823369026184,
+      "learning_rate": 3.0825401000085375e-05,
+      "loss": 0.318,
+      "step": 10212
+    },
+    {
+      "epoch": 0.7510111037576293,
+      "grad_norm": 0.46308445930480957,
+      "learning_rate": 3.0808204322404645e-05,
+      "loss": 0.3332,
+      "step": 10213
+    },
+    {
+      "epoch": 0.7510846385763659,
+      "grad_norm": 0.4469332993030548,
+      "learning_rate": 3.0791011569355585e-05,
+      "loss": 0.3097,
+      "step": 10214
+    },
+    {
+      "epoch": 0.7511581733951026,
+      "grad_norm": 0.4377128779888153,
+      "learning_rate": 3.077382274191338e-05,
+      "loss": 0.3348,
+      "step": 10215
+    },
+    {
+      "epoch": 0.7512317082138392,
+      "grad_norm": 0.48642978072166443,
+      "learning_rate": 3.075663784105305e-05,
+      "loss": 0.3449,
+      "step": 10216
+    },
+    {
+      "epoch": 0.751305243032576,
+      "grad_norm": 0.3878629803657532,
+      "learning_rate": 3.073945686774926e-05,
+      "loss": 0.3169,
+      "step": 10217
+    },
+    {
+      "epoch": 0.7513787778513126,
+      "grad_norm": 0.355671763420105,
+      "learning_rate": 3.0722279822976607e-05,
+      "loss": 0.3109,
+      "step": 10218
+    },
+    {
+      "epoch": 0.7514523126700493,
+      "grad_norm": 0.436323344707489,
+      "learning_rate": 3.070510670770935e-05,
+      "loss": 0.3213,
+      "step": 10219
+    },
+    {
+      "epoch": 0.7515258474887859,
+      "grad_norm": 0.3945727050304413,
+      "learning_rate": 3.068793752292154e-05,
+      "loss": 0.3165,
+      "step": 10220
+    },
+    {
+      "epoch": 0.7515993823075227,
+      "grad_norm": 0.4334068298339844,
+      "learning_rate": 3.0670772269587054e-05,
+      "loss": 0.2832,
+      "step": 10221
+    },
+    {
+      "epoch": 0.7516729171262593,
+      "grad_norm": 0.4243036210536957,
+      "learning_rate": 3.0653610948679565e-05,
+      "loss": 0.3225,
+      "step": 10222
+    },
+    {
+      "epoch": 0.751746451944996,
+      "grad_norm": 0.3855290114879608,
+      "learning_rate": 3.063645356117243e-05,
+      "loss": 0.3779,
+      "step": 10223
+    },
+    {
+      "epoch": 0.7518199867637326,
+      "grad_norm": 0.4744561016559601,
+      "learning_rate": 3.061930010803884e-05,
+      "loss": 0.3473,
+      "step": 10224
+    },
+    {
+      "epoch": 0.7518935215824692,
+      "grad_norm": 0.46133413910865784,
+      "learning_rate": 3.0602150590251765e-05,
+      "loss": 0.3134,
+      "step": 10225
+    },
+    {
+      "epoch": 0.751967056401206,
+      "grad_norm": 0.5225798487663269,
+      "learning_rate": 3.058500500878398e-05,
+      "loss": 0.3276,
+      "step": 10226
+    },
+    {
+      "epoch": 0.7520405912199426,
+      "grad_norm": 0.4610948860645294,
+      "learning_rate": 3.056786336460794e-05,
+      "loss": 0.2894,
+      "step": 10227
+    },
+    {
+      "epoch": 0.7521141260386793,
+      "grad_norm": 0.6266395449638367,
+      "learning_rate": 3.0550725658695965e-05,
+      "loss": 0.3091,
+      "step": 10228
+    },
+    {
+      "epoch": 0.752187660857416,
+      "grad_norm": 0.45861881971359253,
+      "learning_rate": 3.053359189202015e-05,
+      "loss": 0.3217,
+      "step": 10229
+    },
+    {
+      "epoch": 0.7522611956761527,
+      "grad_norm": 0.5321691632270813,
+      "learning_rate": 3.051646206555232e-05,
+      "loss": 0.3184,
+      "step": 10230
+    },
+    {
+      "epoch": 0.7523347304948893,
+      "grad_norm": 0.4613167643547058,
+      "learning_rate": 3.0499336180264047e-05,
+      "loss": 0.3107,
+      "step": 10231
+    },
+    {
+      "epoch": 0.752408265313626,
+      "grad_norm": 0.4138275682926178,
+      "learning_rate": 3.0482214237126815e-05,
+      "loss": 0.3111,
+      "step": 10232
+    },
+    {
+      "epoch": 0.7524818001323627,
+      "grad_norm": 0.5284433364868164,
+      "learning_rate": 3.046509623711172e-05,
+      "loss": 0.3286,
+      "step": 10233
+    },
+    {
+      "epoch": 0.7525553349510994,
+      "grad_norm": 0.46926525235176086,
+      "learning_rate": 3.044798218118976e-05,
+      "loss": 0.3103,
+      "step": 10234
+    },
+    {
+      "epoch": 0.752628869769836,
+      "grad_norm": 0.7528896927833557,
+      "learning_rate": 3.0430872070331674e-05,
+      "loss": 0.3422,
+      "step": 10235
+    },
+    {
+      "epoch": 0.7527024045885727,
+      "grad_norm": 0.43969228863716125,
+      "learning_rate": 3.041376590550791e-05,
+      "loss": 0.3325,
+      "step": 10236
+    },
+    {
+      "epoch": 0.7527759394073094,
+      "grad_norm": 0.5720844268798828,
+      "learning_rate": 3.0396663687688786e-05,
+      "loss": 0.3442,
+      "step": 10237
+    },
+    {
+      "epoch": 0.752849474226046,
+      "grad_norm": 0.4392065703868866,
+      "learning_rate": 3.0379565417844345e-05,
+      "loss": 0.2971,
+      "step": 10238
+    },
+    {
+      "epoch": 0.7529230090447827,
+      "grad_norm": 0.5750499367713928,
+      "learning_rate": 3.0362471096944466e-05,
+      "loss": 0.3263,
+      "step": 10239
+    },
+    {
+      "epoch": 0.7529965438635193,
+      "grad_norm": 0.438201367855072,
+      "learning_rate": 3.0345380725958682e-05,
+      "loss": 0.3082,
+      "step": 10240
+    },
+    {
+      "epoch": 0.7530700786822561,
+      "grad_norm": 0.46300262212753296,
+      "learning_rate": 3.0328294305856454e-05,
+      "loss": 0.3285,
+      "step": 10241
+    },
+    {
+      "epoch": 0.7531436135009927,
+      "grad_norm": 0.5409001111984253,
+      "learning_rate": 3.031121183760689e-05,
+      "loss": 0.3413,
+      "step": 10242
+    },
+    {
+      "epoch": 0.7532171483197294,
+      "grad_norm": 0.4069308340549469,
+      "learning_rate": 3.0294133322178907e-05,
+      "loss": 0.3212,
+      "step": 10243
+    },
+    {
+      "epoch": 0.753290683138466,
+      "grad_norm": 0.5080545544624329,
+      "learning_rate": 3.0277058760541255e-05,
+      "loss": 0.3364,
+      "step": 10244
+    },
+    {
+      "epoch": 0.7533642179572028,
+      "grad_norm": 0.4962013065814972,
+      "learning_rate": 3.0259988153662434e-05,
+      "loss": 0.3187,
+      "step": 10245
+    },
+    {
+      "epoch": 0.7534377527759394,
+      "grad_norm": 0.5116958022117615,
+      "learning_rate": 3.0242921502510647e-05,
+      "loss": 0.3201,
+      "step": 10246
+    },
+    {
+      "epoch": 0.7535112875946761,
+      "grad_norm": 0.578018844127655,
+      "learning_rate": 3.0225858808053975e-05,
+      "loss": 0.3007,
+      "step": 10247
+    },
+    {
+      "epoch": 0.7535848224134127,
+      "grad_norm": 0.6080332398414612,
+      "learning_rate": 3.0208800071260236e-05,
+      "loss": 0.3183,
+      "step": 10248
+    },
+    {
+      "epoch": 0.7536583572321495,
+      "grad_norm": 0.4733300805091858,
+      "learning_rate": 3.019174529309703e-05,
+      "loss": 0.3358,
+      "step": 10249
+    },
+    {
+      "epoch": 0.7537318920508861,
+      "grad_norm": 0.5149821043014526,
+      "learning_rate": 3.0174694474531673e-05,
+      "loss": 0.3136,
+      "step": 10250
+    },
+    {
+      "epoch": 0.7538054268696228,
+      "grad_norm": 0.6174405217170715,
+      "learning_rate": 3.0157647616531338e-05,
+      "loss": 0.3075,
+      "step": 10251
+    },
+    {
+      "epoch": 0.7538789616883594,
+      "grad_norm": 0.42921724915504456,
+      "learning_rate": 3.014060472006297e-05,
+      "loss": 0.3209,
+      "step": 10252
+    },
+    {
+      "epoch": 0.7539524965070961,
+      "grad_norm": 0.6167609095573425,
+      "learning_rate": 3.012356578609323e-05,
+      "loss": 0.301,
+      "step": 10253
+    },
+    {
+      "epoch": 0.7540260313258328,
+      "grad_norm": 0.48328518867492676,
+      "learning_rate": 3.0106530815588553e-05,
+      "loss": 0.3216,
+      "step": 10254
+    },
+    {
+      "epoch": 0.7540995661445694,
+      "grad_norm": 0.561008870601654,
+      "learning_rate": 3.008949980951523e-05,
+      "loss": 0.3004,
+      "step": 10255
+    },
+    {
+      "epoch": 0.7541731009633061,
+      "grad_norm": 0.5337584614753723,
+      "learning_rate": 3.0072472768839233e-05,
+      "loss": 0.282,
+      "step": 10256
+    },
+    {
+      "epoch": 0.7542466357820428,
+      "grad_norm": 0.681376039981842,
+      "learning_rate": 3.005544969452637e-05,
+      "loss": 0.3482,
+      "step": 10257
+    },
+    {
+      "epoch": 0.7543201706007795,
+      "grad_norm": 0.438615083694458,
+      "learning_rate": 3.0038430587542255e-05,
+      "loss": 0.334,
+      "step": 10258
+    },
+    {
+      "epoch": 0.7543937054195161,
+      "grad_norm": 0.5496267080307007,
+      "learning_rate": 3.0021415448852162e-05,
+      "loss": 0.2887,
+      "step": 10259
+    },
+    {
+      "epoch": 0.7544672402382528,
+      "grad_norm": 0.5027125477790833,
+      "learning_rate": 3.000440427942123e-05,
+      "loss": 0.3113,
+      "step": 10260
+    },
+    {
+      "epoch": 0.7545407750569895,
+      "grad_norm": 0.5161908268928528,
+      "learning_rate": 2.9987397080214364e-05,
+      "loss": 0.3215,
+      "step": 10261
+    },
+    {
+      "epoch": 0.7546143098757262,
+      "grad_norm": 0.6639412641525269,
+      "learning_rate": 2.9970393852196244e-05,
+      "loss": 0.3518,
+      "step": 10262
+    },
+    {
+      "epoch": 0.7546878446944628,
+      "grad_norm": 0.4998006224632263,
+      "learning_rate": 2.9953394596331263e-05,
+      "loss": 0.3169,
+      "step": 10263
+    },
+    {
+      "epoch": 0.7547613795131995,
+      "grad_norm": 0.5161516666412354,
+      "learning_rate": 2.9936399313583663e-05,
+      "loss": 0.3142,
+      "step": 10264
+    },
+    {
+      "epoch": 0.7548349143319362,
+      "grad_norm": 0.6369035243988037,
+      "learning_rate": 2.9919408004917503e-05,
+      "loss": 0.3087,
+      "step": 10265
+    },
+    {
+      "epoch": 0.7549084491506728,
+      "grad_norm": 0.4111957252025604,
+      "learning_rate": 2.9902420671296404e-05,
+      "loss": 0.3224,
+      "step": 10266
+    },
+    {
+      "epoch": 0.7549819839694095,
+      "grad_norm": 0.5864478349685669,
+      "learning_rate": 2.9885437313683996e-05,
+      "loss": 0.3017,
+      "step": 10267
+    },
+    {
+      "epoch": 0.7550555187881461,
+      "grad_norm": 0.4727792739868164,
+      "learning_rate": 2.9868457933043603e-05,
+      "loss": 0.3031,
+      "step": 10268
+    },
+    {
+      "epoch": 0.7551290536068829,
+      "grad_norm": 0.4844178855419159,
+      "learning_rate": 2.985148253033826e-05,
+      "loss": 0.3064,
+      "step": 10269
+    },
+    {
+      "epoch": 0.7552025884256195,
+      "grad_norm": 0.398190975189209,
+      "learning_rate": 2.983451110653086e-05,
+      "loss": 0.3023,
+      "step": 10270
+    },
+    {
+      "epoch": 0.7552761232443562,
+      "grad_norm": 0.46402984857559204,
+      "learning_rate": 2.9817543662584037e-05,
+      "loss": 0.3029,
+      "step": 10271
+    },
+    {
+      "epoch": 0.7553496580630928,
+      "grad_norm": 0.4029982388019562,
+      "learning_rate": 2.9800580199460226e-05,
+      "loss": 0.3223,
+      "step": 10272
+    },
+    {
+      "epoch": 0.7554231928818296,
+      "grad_norm": 0.4600641429424286,
+      "learning_rate": 2.978362071812156e-05,
+      "loss": 0.31,
+      "step": 10273
+    },
+    {
+      "epoch": 0.7554967277005662,
+      "grad_norm": 0.43906304240226746,
+      "learning_rate": 2.9766665219530044e-05,
+      "loss": 0.3534,
+      "step": 10274
+    },
+    {
+      "epoch": 0.7555702625193029,
+      "grad_norm": 0.4307094216346741,
+      "learning_rate": 2.974971370464742e-05,
+      "loss": 0.3085,
+      "step": 10275
+    },
+    {
+      "epoch": 0.7556437973380395,
+      "grad_norm": 0.5371949672698975,
+      "learning_rate": 2.9732766174435178e-05,
+      "loss": 0.309,
+      "step": 10276
+    },
+    {
+      "epoch": 0.7557173321567763,
+      "grad_norm": 0.46532803773880005,
+      "learning_rate": 2.971582262985455e-05,
+      "loss": 0.3134,
+      "step": 10277
+    },
+    {
+      "epoch": 0.7557908669755129,
+      "grad_norm": 0.6690793037414551,
+      "learning_rate": 2.9698883071866678e-05,
+      "loss": 0.3235,
+      "step": 10278
+    },
+    {
+      "epoch": 0.7558644017942496,
+      "grad_norm": 0.4065893292427063,
+      "learning_rate": 2.9681947501432317e-05,
+      "loss": 0.3242,
+      "step": 10279
+    },
+    {
+      "epoch": 0.7559379366129862,
+      "grad_norm": 0.7391971945762634,
+      "learning_rate": 2.9665015919512106e-05,
+      "loss": 0.3082,
+      "step": 10280
+    },
+    {
+      "epoch": 0.7560114714317229,
+      "grad_norm": 0.45264968276023865,
+      "learning_rate": 2.9648088327066447e-05,
+      "loss": 0.3537,
+      "step": 10281
+    },
+    {
+      "epoch": 0.7560850062504596,
+      "grad_norm": 0.5847695469856262,
+      "learning_rate": 2.9631164725055426e-05,
+      "loss": 0.2855,
+      "step": 10282
+    },
+    {
+      "epoch": 0.7561585410691962,
+      "grad_norm": 0.5549497008323669,
+      "learning_rate": 2.961424511443902e-05,
+      "loss": 0.3144,
+      "step": 10283
+    },
+    {
+      "epoch": 0.756232075887933,
+      "grad_norm": 0.4156021773815155,
+      "learning_rate": 2.9597329496176896e-05,
+      "loss": 0.3082,
+      "step": 10284
+    },
+    {
+      "epoch": 0.7563056107066696,
+      "grad_norm": 0.754436194896698,
+      "learning_rate": 2.9580417871228584e-05,
+      "loss": 0.2992,
+      "step": 10285
+    },
+    {
+      "epoch": 0.7563791455254063,
+      "grad_norm": 0.4394018054008484,
+      "learning_rate": 2.956351024055325e-05,
+      "loss": 0.3055,
+      "step": 10286
+    },
+    {
+      "epoch": 0.7564526803441429,
+      "grad_norm": 0.8407710194587708,
+      "learning_rate": 2.9546606605109962e-05,
+      "loss": 0.306,
+      "step": 10287
+    },
+    {
+      "epoch": 0.7565262151628797,
+      "grad_norm": 0.6325837969779968,
+      "learning_rate": 2.952970696585755e-05,
+      "loss": 0.3195,
+      "step": 10288
+    },
+    {
+      "epoch": 0.7565997499816163,
+      "grad_norm": 0.6916733980178833,
+      "learning_rate": 2.9512811323754464e-05,
+      "loss": 0.3126,
+      "step": 10289
+    },
+    {
+      "epoch": 0.756673284800353,
+      "grad_norm": 0.7335968613624573,
+      "learning_rate": 2.949591967975912e-05,
+      "loss": 0.3051,
+      "step": 10290
+    },
+    {
+      "epoch": 0.7567468196190896,
+      "grad_norm": 0.861624538898468,
+      "learning_rate": 2.9479032034829633e-05,
+      "loss": 0.3109,
+      "step": 10291
+    },
+    {
+      "epoch": 0.7568203544378264,
+      "grad_norm": 0.9733031392097473,
+      "learning_rate": 2.946214838992386e-05,
+      "loss": 0.3005,
+      "step": 10292
+    },
+    {
+      "epoch": 0.756893889256563,
+      "grad_norm": 0.47712501883506775,
+      "learning_rate": 2.9445268745999454e-05,
+      "loss": 0.3087,
+      "step": 10293
+    },
+    {
+      "epoch": 0.7569674240752996,
+      "grad_norm": 0.9331419467926025,
+      "learning_rate": 2.9428393104013872e-05,
+      "loss": 0.2889,
+      "step": 10294
+    },
+    {
+      "epoch": 0.7570409588940363,
+      "grad_norm": 0.5229402184486389,
+      "learning_rate": 2.9411521464924352e-05,
+      "loss": 0.3188,
+      "step": 10295
+    },
+    {
+      "epoch": 0.757114493712773,
+      "grad_norm": 0.5294033885002136,
+      "learning_rate": 2.939465382968778e-05,
+      "loss": 0.3017,
+      "step": 10296
+    },
+    {
+      "epoch": 0.7571880285315097,
+      "grad_norm": 0.7397599816322327,
+      "learning_rate": 2.9377790199260968e-05,
+      "loss": 0.2961,
+      "step": 10297
+    },
+    {
+      "epoch": 0.7572615633502463,
+      "grad_norm": 0.3878781497478485,
+      "learning_rate": 2.9360930574600454e-05,
+      "loss": 0.3107,
+      "step": 10298
+    },
+    {
+      "epoch": 0.757335098168983,
+      "grad_norm": 0.5805470943450928,
+      "learning_rate": 2.934407495666247e-05,
+      "loss": 0.3135,
+      "step": 10299
+    },
+    {
+      "epoch": 0.7574086329877197,
+      "grad_norm": 0.5700618028640747,
+      "learning_rate": 2.9327223346403155e-05,
+      "loss": 0.338,
+      "step": 10300
+    },
+    {
+      "epoch": 0.7574821678064564,
+      "grad_norm": 0.39749008417129517,
+      "learning_rate": 2.9310375744778318e-05,
+      "loss": 0.3279,
+      "step": 10301
+    },
+    {
+      "epoch": 0.757555702625193,
+      "grad_norm": 0.7393819689750671,
+      "learning_rate": 2.929353215274354e-05,
+      "loss": 0.3233,
+      "step": 10302
+    },
+    {
+      "epoch": 0.7576292374439297,
+      "grad_norm": 0.41598573327064514,
+      "learning_rate": 2.9276692571254238e-05,
+      "loss": 0.3305,
+      "step": 10303
+    },
+    {
+      "epoch": 0.7577027722626664,
+      "grad_norm": 0.44467324018478394,
+      "learning_rate": 2.92598570012656e-05,
+      "loss": 0.3255,
+      "step": 10304
+    },
+    {
+      "epoch": 0.7577763070814031,
+      "grad_norm": 0.6290749907493591,
+      "learning_rate": 2.9243025443732497e-05,
+      "loss": 0.3126,
+      "step": 10305
+    },
+    {
+      "epoch": 0.7578498419001397,
+      "grad_norm": 0.6455585360527039,
+      "learning_rate": 2.9226197899609665e-05,
+      "loss": 0.2998,
+      "step": 10306
+    },
+    {
+      "epoch": 0.7579233767188764,
+      "grad_norm": 0.5909944176673889,
+      "learning_rate": 2.920937436985157e-05,
+      "loss": 0.3544,
+      "step": 10307
+    },
+    {
+      "epoch": 0.7579969115376131,
+      "grad_norm": 0.48851659893989563,
+      "learning_rate": 2.9192554855412512e-05,
+      "loss": 0.3139,
+      "step": 10308
+    },
+    {
+      "epoch": 0.7580704463563497,
+      "grad_norm": 0.9239702224731445,
+      "learning_rate": 2.9175739357246434e-05,
+      "loss": 0.3392,
+      "step": 10309
+    },
+    {
+      "epoch": 0.7581439811750864,
+      "grad_norm": 0.487999826669693,
+      "learning_rate": 2.915892787630716e-05,
+      "loss": 0.3291,
+      "step": 10310
+    },
+    {
+      "epoch": 0.758217515993823,
+      "grad_norm": 0.7125412821769714,
+      "learning_rate": 2.914212041354829e-05,
+      "loss": 0.2925,
+      "step": 10311
+    },
+    {
+      "epoch": 0.7582910508125598,
+      "grad_norm": 0.46075868606567383,
+      "learning_rate": 2.9125316969923134e-05,
+      "loss": 0.3036,
+      "step": 10312
+    },
+    {
+      "epoch": 0.7583645856312964,
+      "grad_norm": 0.9468078017234802,
+      "learning_rate": 2.9108517546384763e-05,
+      "loss": 0.3178,
+      "step": 10313
+    },
+    {
+      "epoch": 0.7584381204500331,
+      "grad_norm": 0.4444507360458374,
+      "learning_rate": 2.9091722143886125e-05,
+      "loss": 0.3072,
+      "step": 10314
+    },
+    {
+      "epoch": 0.7585116552687697,
+      "grad_norm": 0.6423125863075256,
+      "learning_rate": 2.9074930763379804e-05,
+      "loss": 0.321,
+      "step": 10315
+    },
+    {
+      "epoch": 0.7585851900875065,
+      "grad_norm": 0.6964449882507324,
+      "learning_rate": 2.9058143405818262e-05,
+      "loss": 0.3065,
+      "step": 10316
+    },
+    {
+      "epoch": 0.7586587249062431,
+      "grad_norm": 0.6389700174331665,
+      "learning_rate": 2.9041360072153702e-05,
+      "loss": 0.3146,
+      "step": 10317
+    },
+    {
+      "epoch": 0.7587322597249798,
+      "grad_norm": 0.6098455190658569,
+      "learning_rate": 2.9024580763338105e-05,
+      "loss": 0.3408,
+      "step": 10318
+    },
+    {
+      "epoch": 0.7588057945437164,
+      "grad_norm": 0.4360286593437195,
+      "learning_rate": 2.900780548032316e-05,
+      "loss": 0.3002,
+      "step": 10319
+    },
+    {
+      "epoch": 0.7588793293624532,
+      "grad_norm": 0.6129658222198486,
+      "learning_rate": 2.8991034224060408e-05,
+      "loss": 0.311,
+      "step": 10320
+    },
+    {
+      "epoch": 0.7589528641811898,
+      "grad_norm": 0.4690982401371002,
+      "learning_rate": 2.8974266995501165e-05,
+      "loss": 0.3378,
+      "step": 10321
+    },
+    {
+      "epoch": 0.7590263989999264,
+      "grad_norm": 0.5573740005493164,
+      "learning_rate": 2.895750379559643e-05,
+      "loss": 0.3045,
+      "step": 10322
+    },
+    {
+      "epoch": 0.7590999338186631,
+      "grad_norm": 0.5683220028877258,
+      "learning_rate": 2.894074462529708e-05,
+      "loss": 0.329,
+      "step": 10323
+    },
+    {
+      "epoch": 0.7591734686373998,
+      "grad_norm": 0.40095415711402893,
+      "learning_rate": 2.8923989485553695e-05,
+      "loss": 0.3077,
+      "step": 10324
+    },
+    {
+      "epoch": 0.7592470034561365,
+      "grad_norm": 0.5015828013420105,
+      "learning_rate": 2.8907238377316594e-05,
+      "loss": 0.2999,
+      "step": 10325
+    },
+    {
+      "epoch": 0.7593205382748731,
+      "grad_norm": 0.3790073096752167,
+      "learning_rate": 2.889049130153596e-05,
+      "loss": 0.3056,
+      "step": 10326
+    },
+    {
+      "epoch": 0.7593940730936098,
+      "grad_norm": 0.46402138471603394,
+      "learning_rate": 2.8873748259161737e-05,
+      "loss": 0.315,
+      "step": 10327
+    },
+    {
+      "epoch": 0.7594676079123465,
+      "grad_norm": 0.44380104541778564,
+      "learning_rate": 2.885700925114354e-05,
+      "loss": 0.3519,
+      "step": 10328
+    },
+    {
+      "epoch": 0.7595411427310832,
+      "grad_norm": 0.3582659661769867,
+      "learning_rate": 2.8840274278430868e-05,
+      "loss": 0.3256,
+      "step": 10329
+    },
+    {
+      "epoch": 0.7596146775498198,
+      "grad_norm": 0.43544018268585205,
+      "learning_rate": 2.8823543341972937e-05,
+      "loss": 0.3174,
+      "step": 10330
+    },
+    {
+      "epoch": 0.7596882123685565,
+      "grad_norm": 0.43452343344688416,
+      "learning_rate": 2.8806816442718765e-05,
+      "loss": 0.3127,
+      "step": 10331
+    },
+    {
+      "epoch": 0.7597617471872932,
+      "grad_norm": 0.37278813123703003,
+      "learning_rate": 2.8790093581617072e-05,
+      "loss": 0.3186,
+      "step": 10332
+    },
+    {
+      "epoch": 0.7598352820060299,
+      "grad_norm": 0.43050849437713623,
+      "learning_rate": 2.8773374759616423e-05,
+      "loss": 0.3105,
+      "step": 10333
+    },
+    {
+      "epoch": 0.7599088168247665,
+      "grad_norm": 0.39098986983299255,
+      "learning_rate": 2.8756659977665157e-05,
+      "loss": 0.3066,
+      "step": 10334
+    },
+    {
+      "epoch": 0.7599823516435033,
+      "grad_norm": 0.4489831328392029,
+      "learning_rate": 2.8739949236711316e-05,
+      "loss": 0.3204,
+      "step": 10335
+    },
+    {
+      "epoch": 0.7600558864622399,
+      "grad_norm": 0.40666770935058594,
+      "learning_rate": 2.8723242537702733e-05,
+      "loss": 0.3006,
+      "step": 10336
+    },
+    {
+      "epoch": 0.7601294212809765,
+      "grad_norm": 0.4783036708831787,
+      "learning_rate": 2.870653988158709e-05,
+      "loss": 0.3308,
+      "step": 10337
+    },
+    {
+      "epoch": 0.7602029560997132,
+      "grad_norm": 0.42637526988983154,
+      "learning_rate": 2.8689841269311703e-05,
+      "loss": 0.316,
+      "step": 10338
+    },
+    {
+      "epoch": 0.7602764909184498,
+      "grad_norm": 0.6010774374008179,
+      "learning_rate": 2.8673146701823783e-05,
+      "loss": 0.3356,
+      "step": 10339
+    },
+    {
+      "epoch": 0.7603500257371866,
+      "grad_norm": 0.481101930141449,
+      "learning_rate": 2.865645618007028e-05,
+      "loss": 0.3265,
+      "step": 10340
+    },
+    {
+      "epoch": 0.7604235605559232,
+      "grad_norm": 0.5436930060386658,
+      "learning_rate": 2.8639769704997853e-05,
+      "loss": 0.3243,
+      "step": 10341
+    },
+    {
+      "epoch": 0.7604970953746599,
+      "grad_norm": 0.47295060753822327,
+      "learning_rate": 2.8623087277552996e-05,
+      "loss": 0.3157,
+      "step": 10342
+    },
+    {
+      "epoch": 0.7605706301933965,
+      "grad_norm": 0.502816379070282,
+      "learning_rate": 2.8606408898681946e-05,
+      "loss": 0.3248,
+      "step": 10343
+    },
+    {
+      "epoch": 0.7606441650121333,
+      "grad_norm": 0.4070824682712555,
+      "learning_rate": 2.8589734569330774e-05,
+      "loss": 0.3212,
+      "step": 10344
+    },
+    {
+      "epoch": 0.7607176998308699,
+      "grad_norm": 0.5192257761955261,
+      "learning_rate": 2.8573064290445172e-05,
+      "loss": 0.3342,
+      "step": 10345
+    },
+    {
+      "epoch": 0.7607912346496066,
+      "grad_norm": 0.4398528039455414,
+      "learning_rate": 2.8556398062970758e-05,
+      "loss": 0.3719,
+      "step": 10346
+    },
+    {
+      "epoch": 0.7608647694683432,
+      "grad_norm": 0.46241822838783264,
+      "learning_rate": 2.853973588785289e-05,
+      "loss": 0.321,
+      "step": 10347
+    },
+    {
+      "epoch": 0.76093830428708,
+      "grad_norm": 0.40207916498184204,
+      "learning_rate": 2.8523077766036565e-05,
+      "loss": 0.3126,
+      "step": 10348
+    },
+    {
+      "epoch": 0.7610118391058166,
+      "grad_norm": 0.5691240429878235,
+      "learning_rate": 2.85064236984667e-05,
+      "loss": 0.3051,
+      "step": 10349
+    },
+    {
+      "epoch": 0.7610853739245532,
+      "grad_norm": 0.6645870208740234,
+      "learning_rate": 2.8489773686087972e-05,
+      "loss": 0.3442,
+      "step": 10350
+    },
+    {
+      "epoch": 0.76115890874329,
+      "grad_norm": 0.48774272203445435,
+      "learning_rate": 2.8473127729844706e-05,
+      "loss": 0.3031,
+      "step": 10351
+    },
+    {
+      "epoch": 0.7612324435620266,
+      "grad_norm": 0.4587843120098114,
+      "learning_rate": 2.8456485830681122e-05,
+      "loss": 0.3347,
+      "step": 10352
+    },
+    {
+      "epoch": 0.7613059783807633,
+      "grad_norm": 0.5149400234222412,
+      "learning_rate": 2.843984798954117e-05,
+      "loss": 0.3095,
+      "step": 10353
+    },
+    {
+      "epoch": 0.7613795131994999,
+      "grad_norm": 0.48604616522789,
+      "learning_rate": 2.8423214207368588e-05,
+      "loss": 0.3582,
+      "step": 10354
+    },
+    {
+      "epoch": 0.7614530480182367,
+      "grad_norm": 0.609449565410614,
+      "learning_rate": 2.84065844851068e-05,
+      "loss": 0.3278,
+      "step": 10355
+    },
+    {
+      "epoch": 0.7615265828369733,
+      "grad_norm": 0.4739174246788025,
+      "learning_rate": 2.8389958823699103e-05,
+      "loss": 0.3085,
+      "step": 10356
+    },
+    {
+      "epoch": 0.76160011765571,
+      "grad_norm": 0.5467066168785095,
+      "learning_rate": 2.8373337224088536e-05,
+      "loss": 0.3049,
+      "step": 10357
+    },
+    {
+      "epoch": 0.7616736524744466,
+      "grad_norm": 0.425046443939209,
+      "learning_rate": 2.8356719687217882e-05,
+      "loss": 0.3397,
+      "step": 10358
+    },
+    {
+      "epoch": 0.7617471872931834,
+      "grad_norm": 0.5907073616981506,
+      "learning_rate": 2.834010621402966e-05,
+      "loss": 0.3437,
+      "step": 10359
+    },
+    {
+      "epoch": 0.76182072211192,
+      "grad_norm": 0.4360448122024536,
+      "learning_rate": 2.8323496805466276e-05,
+      "loss": 0.3316,
+      "step": 10360
+    },
+    {
+      "epoch": 0.7618942569306567,
+      "grad_norm": 0.5201454758644104,
+      "learning_rate": 2.8306891462469754e-05,
+      "loss": 0.2997,
+      "step": 10361
+    },
+    {
+      "epoch": 0.7619677917493933,
+      "grad_norm": 0.501658022403717,
+      "learning_rate": 2.8290290185982016e-05,
+      "loss": 0.327,
+      "step": 10362
+    },
+    {
+      "epoch": 0.7620413265681301,
+      "grad_norm": 0.6135377287864685,
+      "learning_rate": 2.8273692976944733e-05,
+      "loss": 0.3212,
+      "step": 10363
+    },
+    {
+      "epoch": 0.7621148613868667,
+      "grad_norm": 0.59096759557724,
+      "learning_rate": 2.825709983629924e-05,
+      "loss": 0.318,
+      "step": 10364
+    },
+    {
+      "epoch": 0.7621883962056033,
+      "grad_norm": 0.40102216601371765,
+      "learning_rate": 2.824051076498676e-05,
+      "loss": 0.3207,
+      "step": 10365
+    },
+    {
+      "epoch": 0.76226193102434,
+      "grad_norm": 0.5835230350494385,
+      "learning_rate": 2.8223925763948245e-05,
+      "loss": 0.309,
+      "step": 10366
+    },
+    {
+      "epoch": 0.7623354658430767,
+      "grad_norm": 0.4971928298473358,
+      "learning_rate": 2.820734483412444e-05,
+      "loss": 0.3142,
+      "step": 10367
+    },
+    {
+      "epoch": 0.7624090006618134,
+      "grad_norm": 0.7596021294593811,
+      "learning_rate": 2.8190767976455757e-05,
+      "loss": 0.3198,
+      "step": 10368
+    },
+    {
+      "epoch": 0.76248253548055,
+      "grad_norm": 0.5579007267951965,
+      "learning_rate": 2.8174195191882512e-05,
+      "loss": 0.3155,
+      "step": 10369
+    },
+    {
+      "epoch": 0.7625560702992867,
+      "grad_norm": 0.6370232105255127,
+      "learning_rate": 2.815762648134477e-05,
+      "loss": 0.3324,
+      "step": 10370
+    },
+    {
+      "epoch": 0.7626296051180234,
+      "grad_norm": 0.4573271870613098,
+      "learning_rate": 2.8141061845782212e-05,
+      "loss": 0.323,
+      "step": 10371
+    },
+    {
+      "epoch": 0.7627031399367601,
+      "grad_norm": 0.499529093503952,
+      "learning_rate": 2.8124501286134463e-05,
+      "loss": 0.312,
+      "step": 10372
+    },
+    {
+      "epoch": 0.7627766747554967,
+      "grad_norm": 0.553115725517273,
+      "learning_rate": 2.810794480334089e-05,
+      "loss": 0.2995,
+      "step": 10373
+    },
+    {
+      "epoch": 0.7628502095742334,
+      "grad_norm": 0.4126594662666321,
+      "learning_rate": 2.809139239834052e-05,
+      "loss": 0.3177,
+      "step": 10374
+    },
+    {
+      "epoch": 0.7629237443929701,
+      "grad_norm": 0.5410671830177307,
+      "learning_rate": 2.807484407207227e-05,
+      "loss": 0.3242,
+      "step": 10375
+    },
+    {
+      "epoch": 0.7629972792117068,
+      "grad_norm": 0.5769583582878113,
+      "learning_rate": 2.8058299825474765e-05,
+      "loss": 0.3107,
+      "step": 10376
+    },
+    {
+      "epoch": 0.7630708140304434,
+      "grad_norm": 0.5763788223266602,
+      "learning_rate": 2.8041759659486454e-05,
+      "loss": 0.3238,
+      "step": 10377
+    },
+    {
+      "epoch": 0.76314434884918,
+      "grad_norm": 0.6975780725479126,
+      "learning_rate": 2.8025223575045433e-05,
+      "loss": 0.2869,
+      "step": 10378
+    },
+    {
+      "epoch": 0.7632178836679168,
+      "grad_norm": 0.43763279914855957,
+      "learning_rate": 2.8008691573089696e-05,
+      "loss": 0.2919,
+      "step": 10379
+    },
+    {
+      "epoch": 0.7632914184866534,
+      "grad_norm": 0.46874579787254333,
+      "learning_rate": 2.799216365455698e-05,
+      "loss": 0.3285,
+      "step": 10380
+    },
+    {
+      "epoch": 0.7633649533053901,
+      "grad_norm": 0.5268644094467163,
+      "learning_rate": 2.79756398203847e-05,
+      "loss": 0.3245,
+      "step": 10381
+    },
+    {
+      "epoch": 0.7634384881241267,
+      "grad_norm": 0.7362808585166931,
+      "learning_rate": 2.795912007151017e-05,
+      "loss": 0.3396,
+      "step": 10382
+    },
+    {
+      "epoch": 0.7635120229428635,
+      "grad_norm": 0.36083370447158813,
+      "learning_rate": 2.7942604408870376e-05,
+      "loss": 0.3268,
+      "step": 10383
+    },
+    {
+      "epoch": 0.7635855577616001,
+      "grad_norm": 0.5225012302398682,
+      "learning_rate": 2.7926092833402085e-05,
+      "loss": 0.311,
+      "step": 10384
+    },
+    {
+      "epoch": 0.7636590925803368,
+      "grad_norm": 0.5179669857025146,
+      "learning_rate": 2.7909585346041857e-05,
+      "loss": 0.3029,
+      "step": 10385
+    },
+    {
+      "epoch": 0.7637326273990734,
+      "grad_norm": 0.4565924108028412,
+      "learning_rate": 2.7893081947726073e-05,
+      "loss": 0.3244,
+      "step": 10386
+    },
+    {
+      "epoch": 0.7638061622178102,
+      "grad_norm": 0.5280818939208984,
+      "learning_rate": 2.7876582639390736e-05,
+      "loss": 0.3097,
+      "step": 10387
+    },
+    {
+      "epoch": 0.7638796970365468,
+      "grad_norm": 0.4189249575138092,
+      "learning_rate": 2.786008742197175e-05,
+      "loss": 0.3335,
+      "step": 10388
+    },
+    {
+      "epoch": 0.7639532318552835,
+      "grad_norm": 0.4588470160961151,
+      "learning_rate": 2.7843596296404738e-05,
+      "loss": 0.3267,
+      "step": 10389
+    },
+    {
+      "epoch": 0.7640267666740201,
+      "grad_norm": 0.48860085010528564,
+      "learning_rate": 2.7827109263625128e-05,
+      "loss": 0.3292,
+      "step": 10390
+    },
+    {
+      "epoch": 0.7641003014927569,
+      "grad_norm": 0.46462482213974,
+      "learning_rate": 2.781062632456801e-05,
+      "loss": 0.3288,
+      "step": 10391
+    },
+    {
+      "epoch": 0.7641738363114935,
+      "grad_norm": 0.43235865235328674,
+      "learning_rate": 2.7794147480168352e-05,
+      "loss": 0.3001,
+      "step": 10392
+    },
+    {
+      "epoch": 0.7642473711302301,
+      "grad_norm": 0.44206997752189636,
+      "learning_rate": 2.7777672731360925e-05,
+      "loss": 0.3236,
+      "step": 10393
+    },
+    {
+      "epoch": 0.7643209059489668,
+      "grad_norm": 0.41706404089927673,
+      "learning_rate": 2.776120207908005e-05,
+      "loss": 0.3263,
+      "step": 10394
+    },
+    {
+      "epoch": 0.7643944407677035,
+      "grad_norm": 0.4151836335659027,
+      "learning_rate": 2.7744735524260022e-05,
+      "loss": 0.3331,
+      "step": 10395
+    },
+    {
+      "epoch": 0.7644679755864402,
+      "grad_norm": 0.42352187633514404,
+      "learning_rate": 2.7728273067834887e-05,
+      "loss": 0.3003,
+      "step": 10396
+    },
+    {
+      "epoch": 0.7645415104051768,
+      "grad_norm": 0.4844185411930084,
+      "learning_rate": 2.771181471073835e-05,
+      "loss": 0.3247,
+      "step": 10397
+    },
+    {
+      "epoch": 0.7646150452239135,
+      "grad_norm": 0.3793485462665558,
+      "learning_rate": 2.7695360453903963e-05,
+      "loss": 0.323,
+      "step": 10398
+    },
+    {
+      "epoch": 0.7646885800426502,
+      "grad_norm": 0.4276464581489563,
+      "learning_rate": 2.767891029826504e-05,
+      "loss": 0.3145,
+      "step": 10399
+    },
+    {
+      "epoch": 0.7647621148613869,
+      "grad_norm": 0.48918089270591736,
+      "learning_rate": 2.7662464244754673e-05,
+      "loss": 0.3146,
+      "step": 10400
+    },
+    {
+      "epoch": 0.7648356496801235,
+      "grad_norm": 0.4252111613750458,
+      "learning_rate": 2.7646022294305653e-05,
+      "loss": 0.3321,
+      "step": 10401
+    },
+    {
+      "epoch": 0.7649091844988603,
+      "grad_norm": 0.6087817549705505,
+      "learning_rate": 2.76295844478506e-05,
+      "loss": 0.3199,
+      "step": 10402
+    },
+    {
+      "epoch": 0.7649827193175969,
+      "grad_norm": 0.44099336862564087,
+      "learning_rate": 2.761315070632193e-05,
+      "loss": 0.3328,
+      "step": 10403
+    },
+    {
+      "epoch": 0.7650562541363336,
+      "grad_norm": 0.4412616491317749,
+      "learning_rate": 2.759672107065171e-05,
+      "loss": 0.3279,
+      "step": 10404
+    },
+    {
+      "epoch": 0.7651297889550702,
+      "grad_norm": 0.4717217683792114,
+      "learning_rate": 2.7580295541771916e-05,
+      "loss": 0.3174,
+      "step": 10405
+    },
+    {
+      "epoch": 0.7652033237738068,
+      "grad_norm": 0.4699843227863312,
+      "learning_rate": 2.7563874120614176e-05,
+      "loss": 0.3398,
+      "step": 10406
+    },
+    {
+      "epoch": 0.7652768585925436,
+      "grad_norm": 0.5790001749992371,
+      "learning_rate": 2.7547456808109916e-05,
+      "loss": 0.3193,
+      "step": 10407
+    },
+    {
+      "epoch": 0.7653503934112802,
+      "grad_norm": 0.43881291151046753,
+      "learning_rate": 2.7531043605190376e-05,
+      "loss": 0.3336,
+      "step": 10408
+    },
+    {
+      "epoch": 0.7654239282300169,
+      "grad_norm": 0.7624605298042297,
+      "learning_rate": 2.7514634512786542e-05,
+      "loss": 0.3089,
+      "step": 10409
+    },
+    {
+      "epoch": 0.7654974630487535,
+      "grad_norm": 0.519879162311554,
+      "learning_rate": 2.749822953182911e-05,
+      "loss": 0.327,
+      "step": 10410
+    },
+    {
+      "epoch": 0.7655709978674903,
+      "grad_norm": 0.8033215403556824,
+      "learning_rate": 2.748182866324861e-05,
+      "loss": 0.3229,
+      "step": 10411
+    },
+    {
+      "epoch": 0.7656445326862269,
+      "grad_norm": 0.5232950448989868,
+      "learning_rate": 2.7465431907975326e-05,
+      "loss": 0.3132,
+      "step": 10412
+    },
+    {
+      "epoch": 0.7657180675049636,
+      "grad_norm": 0.6531821489334106,
+      "learning_rate": 2.7449039266939314e-05,
+      "loss": 0.3247,
+      "step": 10413
+    },
+    {
+      "epoch": 0.7657916023237002,
+      "grad_norm": 0.5120433568954468,
+      "learning_rate": 2.7432650741070344e-05,
+      "loss": 0.3311,
+      "step": 10414
+    },
+    {
+      "epoch": 0.765865137142437,
+      "grad_norm": 0.9086534976959229,
+      "learning_rate": 2.741626633129799e-05,
+      "loss": 0.3208,
+      "step": 10415
+    },
+    {
+      "epoch": 0.7659386719611736,
+      "grad_norm": 0.42362770438194275,
+      "learning_rate": 2.7399886038551648e-05,
+      "loss": 0.3491,
+      "step": 10416
+    },
+    {
+      "epoch": 0.7660122067799103,
+      "grad_norm": 0.5862577557563782,
+      "learning_rate": 2.7383509863760392e-05,
+      "loss": 0.3253,
+      "step": 10417
+    },
+    {
+      "epoch": 0.766085741598647,
+      "grad_norm": 0.5606544613838196,
+      "learning_rate": 2.736713780785306e-05,
+      "loss": 0.2956,
+      "step": 10418
+    },
+    {
+      "epoch": 0.7661592764173837,
+      "grad_norm": 0.6087196469306946,
+      "learning_rate": 2.735076987175835e-05,
+      "loss": 0.3266,
+      "step": 10419
+    },
+    {
+      "epoch": 0.7662328112361203,
+      "grad_norm": 0.5703362822532654,
+      "learning_rate": 2.7334406056404615e-05,
+      "loss": 0.3352,
+      "step": 10420
+    },
+    {
+      "epoch": 0.7663063460548569,
+      "grad_norm": 0.5512474179267883,
+      "learning_rate": 2.731804636272006e-05,
+      "loss": 0.3333,
+      "step": 10421
+    },
+    {
+      "epoch": 0.7663798808735937,
+      "grad_norm": 0.611885666847229,
+      "learning_rate": 2.7301690791632618e-05,
+      "loss": 0.312,
+      "step": 10422
+    },
+    {
+      "epoch": 0.7664534156923303,
+      "grad_norm": 0.45404282212257385,
+      "learning_rate": 2.7285339344070016e-05,
+      "loss": 0.3642,
+      "step": 10423
+    },
+    {
+      "epoch": 0.766526950511067,
+      "grad_norm": 0.4696865975856781,
+      "learning_rate": 2.7268992020959682e-05,
+      "loss": 0.3359,
+      "step": 10424
+    },
+    {
+      "epoch": 0.7666004853298036,
+      "grad_norm": 0.4630928039550781,
+      "learning_rate": 2.7252648823228877e-05,
+      "loss": 0.3386,
+      "step": 10425
+    },
+    {
+      "epoch": 0.7666740201485404,
+      "grad_norm": 0.5265793800354004,
+      "learning_rate": 2.7236309751804635e-05,
+      "loss": 0.3472,
+      "step": 10426
+    },
+    {
+      "epoch": 0.766747554967277,
+      "grad_norm": 0.6278029680252075,
+      "learning_rate": 2.721997480761367e-05,
+      "loss": 0.3286,
+      "step": 10427
+    },
+    {
+      "epoch": 0.7668210897860137,
+      "grad_norm": 0.4772230386734009,
+      "learning_rate": 2.720364399158256e-05,
+      "loss": 0.3304,
+      "step": 10428
+    },
+    {
+      "epoch": 0.7668946246047503,
+      "grad_norm": 0.39450034499168396,
+      "learning_rate": 2.7187317304637595e-05,
+      "loss": 0.3132,
+      "step": 10429
+    },
+    {
+      "epoch": 0.7669681594234871,
+      "grad_norm": 0.4757455289363861,
+      "learning_rate": 2.7170994747704803e-05,
+      "loss": 0.3222,
+      "step": 10430
+    },
+    {
+      "epoch": 0.7670416942422237,
+      "grad_norm": 0.4128180146217346,
+      "learning_rate": 2.715467632171005e-05,
+      "loss": 0.343,
+      "step": 10431
+    },
+    {
+      "epoch": 0.7671152290609604,
+      "grad_norm": 0.5278036594390869,
+      "learning_rate": 2.7138362027578966e-05,
+      "loss": 0.3273,
+      "step": 10432
+    },
+    {
+      "epoch": 0.767188763879697,
+      "grad_norm": 0.5343619585037231,
+      "learning_rate": 2.7122051866236843e-05,
+      "loss": 0.3414,
+      "step": 10433
+    },
+    {
+      "epoch": 0.7672622986984337,
+      "grad_norm": 0.4326174259185791,
+      "learning_rate": 2.7105745838608852e-05,
+      "loss": 0.348,
+      "step": 10434
+    },
+    {
+      "epoch": 0.7673358335171704,
+      "grad_norm": 0.5587000250816345,
+      "learning_rate": 2.7089443945619884e-05,
+      "loss": 0.3427,
+      "step": 10435
+    },
+    {
+      "epoch": 0.767409368335907,
+      "grad_norm": 0.44722896814346313,
+      "learning_rate": 2.7073146188194643e-05,
+      "loss": 0.364,
+      "step": 10436
+    },
+    {
+      "epoch": 0.7674829031546437,
+      "grad_norm": 0.5144415497779846,
+      "learning_rate": 2.7056852567257474e-05,
+      "loss": 0.349,
+      "step": 10437
+    },
+    {
+      "epoch": 0.7675564379733804,
+      "grad_norm": 0.5119985938072205,
+      "learning_rate": 2.7040563083732607e-05,
+      "loss": 0.3299,
+      "step": 10438
+    },
+    {
+      "epoch": 0.7676299727921171,
+      "grad_norm": 0.5208997130393982,
+      "learning_rate": 2.7024277738544024e-05,
+      "loss": 0.3417,
+      "step": 10439
+    },
+    {
+      "epoch": 0.7677035076108537,
+      "grad_norm": 0.5305895805358887,
+      "learning_rate": 2.700799653261543e-05,
+      "loss": 0.3659,
+      "step": 10440
+    },
+    {
+      "epoch": 0.7677770424295904,
+      "grad_norm": 0.5141067504882812,
+      "learning_rate": 2.699171946687027e-05,
+      "loss": 0.3197,
+      "step": 10441
+    },
+    {
+      "epoch": 0.7678505772483271,
+      "grad_norm": 0.5821164846420288,
+      "learning_rate": 2.6975446542231863e-05,
+      "loss": 0.3414,
+      "step": 10442
+    },
+    {
+      "epoch": 0.7679241120670638,
+      "grad_norm": 0.44766518473625183,
+      "learning_rate": 2.6959177759623167e-05,
+      "loss": 0.3311,
+      "step": 10443
+    },
+    {
+      "epoch": 0.7679976468858004,
+      "grad_norm": 0.4758276641368866,
+      "learning_rate": 2.6942913119966985e-05,
+      "loss": 0.3542,
+      "step": 10444
+    },
+    {
+      "epoch": 0.7680711817045371,
+      "grad_norm": 0.46002933382987976,
+      "learning_rate": 2.6926652624185866e-05,
+      "loss": 0.3119,
+      "step": 10445
+    },
+    {
+      "epoch": 0.7681447165232738,
+      "grad_norm": 0.4046159088611603,
+      "learning_rate": 2.6910396273202164e-05,
+      "loss": 0.3336,
+      "step": 10446
+    },
+    {
+      "epoch": 0.7682182513420105,
+      "grad_norm": 0.4090314209461212,
+      "learning_rate": 2.6894144067937898e-05,
+      "loss": 0.3245,
+      "step": 10447
+    },
+    {
+      "epoch": 0.7682917861607471,
+      "grad_norm": 0.4216558337211609,
+      "learning_rate": 2.6877896009314918e-05,
+      "loss": 0.3306,
+      "step": 10448
+    },
+    {
+      "epoch": 0.7683653209794837,
+      "grad_norm": 0.6329522728919983,
+      "learning_rate": 2.6861652098254886e-05,
+      "loss": 0.3714,
+      "step": 10449
+    },
+    {
+      "epoch": 0.7684388557982205,
+      "grad_norm": 0.4720131456851959,
+      "learning_rate": 2.6845412335679098e-05,
+      "loss": 0.3532,
+      "step": 10450
+    },
+    {
+      "epoch": 0.7685123906169571,
+      "grad_norm": 0.5507887601852417,
+      "learning_rate": 2.682917672250873e-05,
+      "loss": 0.3517,
+      "step": 10451
+    },
+    {
+      "epoch": 0.7685859254356938,
+      "grad_norm": 0.5100839138031006,
+      "learning_rate": 2.6812945259664735e-05,
+      "loss": 0.3251,
+      "step": 10452
+    },
+    {
+      "epoch": 0.7686594602544304,
+      "grad_norm": 0.4919719696044922,
+      "learning_rate": 2.679671794806766e-05,
+      "loss": 0.3371,
+      "step": 10453
+    },
+    {
+      "epoch": 0.7687329950731672,
+      "grad_norm": 0.476484477519989,
+      "learning_rate": 2.6780494788638e-05,
+      "loss": 0.3479,
+      "step": 10454
+    },
+    {
+      "epoch": 0.7688065298919038,
+      "grad_norm": 0.4278796315193176,
+      "learning_rate": 2.6764275782295976e-05,
+      "loss": 0.3237,
+      "step": 10455
+    },
+    {
+      "epoch": 0.7688800647106405,
+      "grad_norm": 0.4739013612270355,
+      "learning_rate": 2.6748060929961494e-05,
+      "loss": 0.3375,
+      "step": 10456
+    },
+    {
+      "epoch": 0.7689535995293771,
+      "grad_norm": 0.4268258810043335,
+      "learning_rate": 2.6731850232554302e-05,
+      "loss": 0.3618,
+      "step": 10457
+    },
+    {
+      "epoch": 0.7690271343481139,
+      "grad_norm": 0.49023792147636414,
+      "learning_rate": 2.6715643690993876e-05,
+      "loss": 0.3555,
+      "step": 10458
+    },
+    {
+      "epoch": 0.7691006691668505,
+      "grad_norm": 0.6228546500205994,
+      "learning_rate": 2.6699441306199513e-05,
+      "loss": 0.3473,
+      "step": 10459
+    },
+    {
+      "epoch": 0.7691742039855872,
+      "grad_norm": 0.4888075590133667,
+      "learning_rate": 2.668324307909017e-05,
+      "loss": 0.3327,
+      "step": 10460
+    },
+    {
+      "epoch": 0.7692477388043238,
+      "grad_norm": 0.6442314982414246,
+      "learning_rate": 2.6667049010584654e-05,
+      "loss": 0.3761,
+      "step": 10461
+    },
+    {
+      "epoch": 0.7693212736230605,
+      "grad_norm": 0.5703672170639038,
+      "learning_rate": 2.6650859101601545e-05,
+      "loss": 0.3617,
+      "step": 10462
+    },
+    {
+      "epoch": 0.7693948084417972,
+      "grad_norm": 0.5324043035507202,
+      "learning_rate": 2.6634673353059115e-05,
+      "loss": 0.3401,
+      "step": 10463
+    },
+    {
+      "epoch": 0.7694683432605338,
+      "grad_norm": 0.5725683569908142,
+      "learning_rate": 2.6618491765875398e-05,
+      "loss": 0.3466,
+      "step": 10464
+    },
+    {
+      "epoch": 0.7695418780792705,
+      "grad_norm": 0.4390106201171875,
+      "learning_rate": 2.6602314340968305e-05,
+      "loss": 0.3635,
+      "step": 10465
+    },
+    {
+      "epoch": 0.7696154128980072,
+      "grad_norm": 0.4096195101737976,
+      "learning_rate": 2.6586141079255377e-05,
+      "loss": 0.3415,
+      "step": 10466
+    },
+    {
+      "epoch": 0.7696889477167439,
+      "grad_norm": 0.7909799218177795,
+      "learning_rate": 2.6569971981653996e-05,
+      "loss": 0.3488,
+      "step": 10467
+    },
+    {
+      "epoch": 0.7697624825354805,
+      "grad_norm": 0.4990319311618805,
+      "learning_rate": 2.6553807049081303e-05,
+      "loss": 0.3404,
+      "step": 10468
+    },
+    {
+      "epoch": 0.7698360173542173,
+      "grad_norm": 0.6224823594093323,
+      "learning_rate": 2.653764628245422e-05,
+      "loss": 0.3584,
+      "step": 10469
+    },
+    {
+      "epoch": 0.7699095521729539,
+      "grad_norm": 0.46750086545944214,
+      "learning_rate": 2.6521489682689328e-05,
+      "loss": 0.3418,
+      "step": 10470
+    },
+    {
+      "epoch": 0.7699830869916906,
+      "grad_norm": 0.7897912263870239,
+      "learning_rate": 2.6505337250703087e-05,
+      "loss": 0.3227,
+      "step": 10471
+    },
+    {
+      "epoch": 0.7700566218104272,
+      "grad_norm": 0.3904566466808319,
+      "learning_rate": 2.648918898741172e-05,
+      "loss": 0.3256,
+      "step": 10472
+    },
+    {
+      "epoch": 0.770130156629164,
+      "grad_norm": 0.5131181478500366,
+      "learning_rate": 2.6473044893731093e-05,
+      "loss": 0.3574,
+      "step": 10473
+    },
+    {
+      "epoch": 0.7702036914479006,
+      "grad_norm": 0.5680312514305115,
+      "learning_rate": 2.6456904970576956e-05,
+      "loss": 0.3457,
+      "step": 10474
+    },
+    {
+      "epoch": 0.7702772262666373,
+      "grad_norm": 0.4939415752887726,
+      "learning_rate": 2.6440769218864848e-05,
+      "loss": 0.333,
+      "step": 10475
+    },
+    {
+      "epoch": 0.7703507610853739,
+      "grad_norm": 0.6999942660331726,
+      "learning_rate": 2.6424637639509874e-05,
+      "loss": 0.3512,
+      "step": 10476
+    },
+    {
+      "epoch": 0.7704242959041105,
+      "grad_norm": 0.5418015122413635,
+      "learning_rate": 2.64085102334271e-05,
+      "loss": 0.3383,
+      "step": 10477
+    },
+    {
+      "epoch": 0.7704978307228473,
+      "grad_norm": 0.6885183453559875,
+      "learning_rate": 2.6392387001531316e-05,
+      "loss": 0.3335,
+      "step": 10478
+    },
+    {
+      "epoch": 0.7705713655415839,
+      "grad_norm": 0.5580676198005676,
+      "learning_rate": 2.6376267944736998e-05,
+      "loss": 0.3344,
+      "step": 10479
+    },
+    {
+      "epoch": 0.7706449003603206,
+      "grad_norm": 0.43484506011009216,
+      "learning_rate": 2.636015306395847e-05,
+      "loss": 0.3178,
+      "step": 10480
+    },
+    {
+      "epoch": 0.7707184351790572,
+      "grad_norm": 0.4733402132987976,
+      "learning_rate": 2.6344042360109768e-05,
+      "loss": 0.3328,
+      "step": 10481
+    },
+    {
+      "epoch": 0.770791969997794,
+      "grad_norm": 0.587882936000824,
+      "learning_rate": 2.6327935834104732e-05,
+      "loss": 0.3268,
+      "step": 10482
+    },
+    {
+      "epoch": 0.7708655048165306,
+      "grad_norm": 0.46432220935821533,
+      "learning_rate": 2.6311833486856908e-05,
+      "loss": 0.3258,
+      "step": 10483
+    },
+    {
+      "epoch": 0.7709390396352673,
+      "grad_norm": 0.5584412813186646,
+      "learning_rate": 2.6295735319279656e-05,
+      "loss": 0.3835,
+      "step": 10484
+    },
+    {
+      "epoch": 0.771012574454004,
+      "grad_norm": 0.5605945587158203,
+      "learning_rate": 2.627964133228611e-05,
+      "loss": 0.3318,
+      "step": 10485
+    },
+    {
+      "epoch": 0.7710861092727407,
+      "grad_norm": 0.4266977310180664,
+      "learning_rate": 2.6263551526789075e-05,
+      "loss": 0.3558,
+      "step": 10486
+    },
+    {
+      "epoch": 0.7711596440914773,
+      "grad_norm": 0.5973663926124573,
+      "learning_rate": 2.624746590370124e-05,
+      "loss": 0.3247,
+      "step": 10487
+    },
+    {
+      "epoch": 0.771233178910214,
+      "grad_norm": 0.5472822189331055,
+      "learning_rate": 2.6231384463934982e-05,
+      "loss": 0.3568,
+      "step": 10488
+    },
+    {
+      "epoch": 0.7713067137289507,
+      "grad_norm": 0.567816436290741,
+      "learning_rate": 2.6215307208402416e-05,
+      "loss": 0.3446,
+      "step": 10489
+    },
+    {
+      "epoch": 0.7713802485476873,
+      "grad_norm": 0.5225076675415039,
+      "learning_rate": 2.6199234138015482e-05,
+      "loss": 0.3308,
+      "step": 10490
+    },
+    {
+      "epoch": 0.771453783366424,
+      "grad_norm": 0.5865317583084106,
+      "learning_rate": 2.6183165253685915e-05,
+      "loss": 0.3449,
+      "step": 10491
+    },
+    {
+      "epoch": 0.7715273181851606,
+      "grad_norm": 0.4881812334060669,
+      "learning_rate": 2.6167100556325086e-05,
+      "loss": 0.3729,
+      "step": 10492
+    },
+    {
+      "epoch": 0.7716008530038974,
+      "grad_norm": 0.7177110314369202,
+      "learning_rate": 2.6151040046844232e-05,
+      "loss": 0.3563,
+      "step": 10493
+    },
+    {
+      "epoch": 0.771674387822634,
+      "grad_norm": 0.5399317741394043,
+      "learning_rate": 2.6134983726154315e-05,
+      "loss": 0.3662,
+      "step": 10494
+    },
+    {
+      "epoch": 0.7717479226413707,
+      "grad_norm": 0.4344121515750885,
+      "learning_rate": 2.6118931595166118e-05,
+      "loss": 0.3605,
+      "step": 10495
+    },
+    {
+      "epoch": 0.7718214574601073,
+      "grad_norm": 0.5767824649810791,
+      "learning_rate": 2.6102883654790045e-05,
+      "loss": 0.3664,
+      "step": 10496
+    },
+    {
+      "epoch": 0.7718949922788441,
+      "grad_norm": 0.5842704176902771,
+      "learning_rate": 2.608683990593641e-05,
+      "loss": 0.3532,
+      "step": 10497
+    },
+    {
+      "epoch": 0.7719685270975807,
+      "grad_norm": 0.6042270660400391,
+      "learning_rate": 2.607080034951527e-05,
+      "loss": 0.3865,
+      "step": 10498
+    },
+    {
+      "epoch": 0.7720420619163174,
+      "grad_norm": 0.5573521852493286,
+      "learning_rate": 2.6054764986436296e-05,
+      "loss": 0.3847,
+      "step": 10499
+    },
+    {
+      "epoch": 0.772115596735054,
+      "grad_norm": 0.5584924221038818,
+      "learning_rate": 2.6038733817609095e-05,
+      "loss": 0.3615,
+      "step": 10500
+    },
+    {
+      "epoch": 0.7721891315537908,
+      "grad_norm": 0.450714647769928,
+      "learning_rate": 2.6022706843943002e-05,
+      "loss": 0.3429,
+      "step": 10501
+    },
+    {
+      "epoch": 0.7722626663725274,
+      "grad_norm": 0.41809117794036865,
+      "learning_rate": 2.600668406634701e-05,
+      "loss": 0.3636,
+      "step": 10502
+    },
+    {
+      "epoch": 0.7723362011912641,
+      "grad_norm": 0.5408306121826172,
+      "learning_rate": 2.5990665485729982e-05,
+      "loss": 0.363,
+      "step": 10503
+    },
+    {
+      "epoch": 0.7724097360100007,
+      "grad_norm": 0.4608887732028961,
+      "learning_rate": 2.5974651103000514e-05,
+      "loss": 0.3622,
+      "step": 10504
+    },
+    {
+      "epoch": 0.7724832708287374,
+      "grad_norm": 0.6336163878440857,
+      "learning_rate": 2.595864091906699e-05,
+      "loss": 0.369,
+      "step": 10505
+    },
+    {
+      "epoch": 0.7725568056474741,
+      "grad_norm": 0.4158175587654114,
+      "learning_rate": 2.594263493483746e-05,
+      "loss": 0.3355,
+      "step": 10506
+    },
+    {
+      "epoch": 0.7726303404662107,
+      "grad_norm": 0.468377947807312,
+      "learning_rate": 2.5926633151219837e-05,
+      "loss": 0.3696,
+      "step": 10507
+    },
+    {
+      "epoch": 0.7727038752849474,
+      "grad_norm": 0.43340668082237244,
+      "learning_rate": 2.5910635569121788e-05,
+      "loss": 0.3877,
+      "step": 10508
+    },
+    {
+      "epoch": 0.7727774101036841,
+      "grad_norm": 0.4462980031967163,
+      "learning_rate": 2.5894642189450634e-05,
+      "loss": 0.3468,
+      "step": 10509
+    },
+    {
+      "epoch": 0.7728509449224208,
+      "grad_norm": 0.3873593211174011,
+      "learning_rate": 2.587865301311363e-05,
+      "loss": 0.3358,
+      "step": 10510
+    },
+    {
+      "epoch": 0.7729244797411574,
+      "grad_norm": 0.4418613016605377,
+      "learning_rate": 2.5862668041017636e-05,
+      "loss": 0.3409,
+      "step": 10511
+    },
+    {
+      "epoch": 0.7729980145598941,
+      "grad_norm": 0.5012530088424683,
+      "learning_rate": 2.5846687274069327e-05,
+      "loss": 0.3497,
+      "step": 10512
+    },
+    {
+      "epoch": 0.7730715493786308,
+      "grad_norm": 0.4301334619522095,
+      "learning_rate": 2.5830710713175166e-05,
+      "loss": 0.3642,
+      "step": 10513
+    },
+    {
+      "epoch": 0.7731450841973675,
+      "grad_norm": 0.47628211975097656,
+      "learning_rate": 2.5814738359241407e-05,
+      "loss": 0.3471,
+      "step": 10514
+    },
+    {
+      "epoch": 0.7732186190161041,
+      "grad_norm": 0.5010542869567871,
+      "learning_rate": 2.579877021317395e-05,
+      "loss": 0.3621,
+      "step": 10515
+    },
+    {
+      "epoch": 0.7732921538348408,
+      "grad_norm": 0.40070685744285583,
+      "learning_rate": 2.5782806275878546e-05,
+      "loss": 0.3663,
+      "step": 10516
+    },
+    {
+      "epoch": 0.7733656886535775,
+      "grad_norm": 0.5417672395706177,
+      "learning_rate": 2.5766846548260704e-05,
+      "loss": 0.3485,
+      "step": 10517
+    },
+    {
+      "epoch": 0.7734392234723141,
+      "grad_norm": 0.4417475759983063,
+      "learning_rate": 2.5750891031225686e-05,
+      "loss": 0.3319,
+      "step": 10518
+    },
+    {
+      "epoch": 0.7735127582910508,
+      "grad_norm": 0.4826425611972809,
+      "learning_rate": 2.5734939725678466e-05,
+      "loss": 0.3564,
+      "step": 10519
+    },
+    {
+      "epoch": 0.7735862931097874,
+      "grad_norm": 0.586991012096405,
+      "learning_rate": 2.571899263252383e-05,
+      "loss": 0.3404,
+      "step": 10520
+    },
+    {
+      "epoch": 0.7736598279285242,
+      "grad_norm": 0.49973341822624207,
+      "learning_rate": 2.5703049752666364e-05,
+      "loss": 0.3309,
+      "step": 10521
+    },
+    {
+      "epoch": 0.7737333627472608,
+      "grad_norm": 0.654922604560852,
+      "learning_rate": 2.5687111087010318e-05,
+      "loss": 0.3608,
+      "step": 10522
+    },
+    {
+      "epoch": 0.7738068975659975,
+      "grad_norm": 0.5102683305740356,
+      "learning_rate": 2.5671176636459727e-05,
+      "loss": 0.337,
+      "step": 10523
+    },
+    {
+      "epoch": 0.7738804323847341,
+      "grad_norm": 0.7918392419815063,
+      "learning_rate": 2.5655246401918476e-05,
+      "loss": 0.3614,
+      "step": 10524
+    },
+    {
+      "epoch": 0.7739539672034709,
+      "grad_norm": 0.5741273760795593,
+      "learning_rate": 2.5639320384290068e-05,
+      "loss": 0.3567,
+      "step": 10525
+    },
+    {
+      "epoch": 0.7740275020222075,
+      "grad_norm": 0.7393288612365723,
+      "learning_rate": 2.5623398584477898e-05,
+      "loss": 0.3301,
+      "step": 10526
+    },
+    {
+      "epoch": 0.7741010368409442,
+      "grad_norm": 0.45109057426452637,
+      "learning_rate": 2.5607481003385038e-05,
+      "loss": 0.3538,
+      "step": 10527
+    },
+    {
+      "epoch": 0.7741745716596808,
+      "grad_norm": 0.7306743264198303,
+      "learning_rate": 2.5591567641914406e-05,
+      "loss": 0.3602,
+      "step": 10528
+    },
+    {
+      "epoch": 0.7742481064784176,
+      "grad_norm": 0.5221695303916931,
+      "learning_rate": 2.5575658500968548e-05,
+      "loss": 0.3418,
+      "step": 10529
+    },
+    {
+      "epoch": 0.7743216412971542,
+      "grad_norm": 0.5424720048904419,
+      "learning_rate": 2.5559753581449887e-05,
+      "loss": 0.3479,
+      "step": 10530
+    },
+    {
+      "epoch": 0.7743951761158909,
+      "grad_norm": 0.6017113327980042,
+      "learning_rate": 2.5543852884260587e-05,
+      "loss": 0.3233,
+      "step": 10531
+    },
+    {
+      "epoch": 0.7744687109346275,
+      "grad_norm": 0.6378235220909119,
+      "learning_rate": 2.5527956410302502e-05,
+      "loss": 0.3601,
+      "step": 10532
+    },
+    {
+      "epoch": 0.7745422457533642,
+      "grad_norm": 0.5907344818115234,
+      "learning_rate": 2.551206416047732e-05,
+      "loss": 0.3605,
+      "step": 10533
+    },
+    {
+      "epoch": 0.7746157805721009,
+      "grad_norm": 0.5745229721069336,
+      "learning_rate": 2.5496176135686532e-05,
+      "loss": 0.3274,
+      "step": 10534
+    },
+    {
+      "epoch": 0.7746893153908375,
+      "grad_norm": 0.6289781332015991,
+      "learning_rate": 2.5480292336831202e-05,
+      "loss": 0.3472,
+      "step": 10535
+    },
+    {
+      "epoch": 0.7747628502095742,
+      "grad_norm": 0.7013591527938843,
+      "learning_rate": 2.546441276481233e-05,
+      "loss": 0.3552,
+      "step": 10536
+    },
+    {
+      "epoch": 0.7748363850283109,
+      "grad_norm": 0.8467854857444763,
+      "learning_rate": 2.544853742053066e-05,
+      "loss": 0.3468,
+      "step": 10537
+    },
+    {
+      "epoch": 0.7749099198470476,
+      "grad_norm": 0.7176987528800964,
+      "learning_rate": 2.5432666304886587e-05,
+      "loss": 0.3188,
+      "step": 10538
+    },
+    {
+      "epoch": 0.7749834546657842,
+      "grad_norm": 0.49663016200065613,
+      "learning_rate": 2.5416799418780378e-05,
+      "loss": 0.3356,
+      "step": 10539
+    },
+    {
+      "epoch": 0.775056989484521,
+      "grad_norm": 0.6700572371482849,
+      "learning_rate": 2.5400936763112028e-05,
+      "loss": 0.3434,
+      "step": 10540
+    },
+    {
+      "epoch": 0.7751305243032576,
+      "grad_norm": 0.5532013177871704,
+      "learning_rate": 2.53850783387813e-05,
+      "loss": 0.3486,
+      "step": 10541
+    },
+    {
+      "epoch": 0.7752040591219943,
+      "grad_norm": 0.43522295355796814,
+      "learning_rate": 2.536922414668763e-05,
+      "loss": 0.3306,
+      "step": 10542
+    },
+    {
+      "epoch": 0.7752775939407309,
+      "grad_norm": 0.5382030010223389,
+      "learning_rate": 2.535337418773034e-05,
+      "loss": 0.3403,
+      "step": 10543
+    },
+    {
+      "epoch": 0.7753511287594677,
+      "grad_norm": 0.5040439367294312,
+      "learning_rate": 2.5337528462808468e-05,
+      "loss": 0.3623,
+      "step": 10544
+    },
+    {
+      "epoch": 0.7754246635782043,
+      "grad_norm": 0.6155710220336914,
+      "learning_rate": 2.532168697282077e-05,
+      "loss": 0.3781,
+      "step": 10545
+    },
+    {
+      "epoch": 0.7754981983969409,
+      "grad_norm": 0.5336854457855225,
+      "learning_rate": 2.5305849718665777e-05,
+      "loss": 0.3308,
+      "step": 10546
+    },
+    {
+      "epoch": 0.7755717332156776,
+      "grad_norm": 0.4097786545753479,
+      "learning_rate": 2.5290016701241847e-05,
+      "loss": 0.3412,
+      "step": 10547
+    },
+    {
+      "epoch": 0.7756452680344142,
+      "grad_norm": 0.42500802874565125,
+      "learning_rate": 2.5274187921446977e-05,
+      "loss": 0.3474,
+      "step": 10548
+    },
+    {
+      "epoch": 0.775718802853151,
+      "grad_norm": 0.5126277208328247,
+      "learning_rate": 2.525836338017904e-05,
+      "loss": 0.333,
+      "step": 10549
+    },
+    {
+      "epoch": 0.7757923376718876,
+      "grad_norm": 0.568162202835083,
+      "learning_rate": 2.52425430783356e-05,
+      "loss": 0.3511,
+      "step": 10550
+    },
+    {
+      "epoch": 0.7758658724906243,
+      "grad_norm": 0.4916384518146515,
+      "learning_rate": 2.5226727016814045e-05,
+      "loss": 0.3475,
+      "step": 10551
+    },
+    {
+      "epoch": 0.775939407309361,
+      "grad_norm": 0.5338225364685059,
+      "learning_rate": 2.5210915196511407e-05,
+      "loss": 0.3468,
+      "step": 10552
+    },
+    {
+      "epoch": 0.7760129421280977,
+      "grad_norm": 0.559208869934082,
+      "learning_rate": 2.51951076183246e-05,
+      "loss": 0.3772,
+      "step": 10553
+    },
+    {
+      "epoch": 0.7760864769468343,
+      "grad_norm": 0.4503602981567383,
+      "learning_rate": 2.5179304283150262e-05,
+      "loss": 0.359,
+      "step": 10554
+    },
+    {
+      "epoch": 0.776160011765571,
+      "grad_norm": 0.6765434741973877,
+      "learning_rate": 2.5163505191884707e-05,
+      "loss": 0.3601,
+      "step": 10555
+    },
+    {
+      "epoch": 0.7762335465843077,
+      "grad_norm": 0.46824198961257935,
+      "learning_rate": 2.5147710345424126e-05,
+      "loss": 0.3798,
+      "step": 10556
+    },
+    {
+      "epoch": 0.7763070814030444,
+      "grad_norm": 0.4608836770057678,
+      "learning_rate": 2.5131919744664466e-05,
+      "loss": 0.3426,
+      "step": 10557
+    },
+    {
+      "epoch": 0.776380616221781,
+      "grad_norm": 0.5819476246833801,
+      "learning_rate": 2.5116133390501262e-05,
+      "loss": 0.3678,
+      "step": 10558
+    },
+    {
+      "epoch": 0.7764541510405177,
+      "grad_norm": 0.36589133739471436,
+      "learning_rate": 2.5100351283830004e-05,
+      "loss": 0.3627,
+      "step": 10559
+    },
+    {
+      "epoch": 0.7765276858592544,
+      "grad_norm": 0.5381962060928345,
+      "learning_rate": 2.50845734255459e-05,
+      "loss": 0.3661,
+      "step": 10560
+    },
+    {
+      "epoch": 0.776601220677991,
+      "grad_norm": 0.41784292459487915,
+      "learning_rate": 2.506879981654382e-05,
+      "loss": 0.3546,
+      "step": 10561
+    },
+    {
+      "epoch": 0.7766747554967277,
+      "grad_norm": 0.6255543828010559,
+      "learning_rate": 2.5053030457718496e-05,
+      "loss": 0.3759,
+      "step": 10562
+    },
+    {
+      "epoch": 0.7767482903154643,
+      "grad_norm": 0.460704505443573,
+      "learning_rate": 2.5037265349964377e-05,
+      "loss": 0.3355,
+      "step": 10563
+    },
+    {
+      "epoch": 0.7768218251342011,
+      "grad_norm": 0.4873126745223999,
+      "learning_rate": 2.5021504494175706e-05,
+      "loss": 0.3648,
+      "step": 10564
+    },
+    {
+      "epoch": 0.7768953599529377,
+      "grad_norm": 0.5173463225364685,
+      "learning_rate": 2.50057478912464e-05,
+      "loss": 0.3694,
+      "step": 10565
+    },
+    {
+      "epoch": 0.7769688947716744,
+      "grad_norm": 0.481698602437973,
+      "learning_rate": 2.4989995542070223e-05,
+      "loss": 0.3519,
+      "step": 10566
+    },
+    {
+      "epoch": 0.777042429590411,
+      "grad_norm": 0.4419515132904053,
+      "learning_rate": 2.4974247447540677e-05,
+      "loss": 0.3554,
+      "step": 10567
+    },
+    {
+      "epoch": 0.7771159644091478,
+      "grad_norm": 0.45656564831733704,
+      "learning_rate": 2.4958503608550977e-05,
+      "loss": 0.3704,
+      "step": 10568
+    },
+    {
+      "epoch": 0.7771894992278844,
+      "grad_norm": 0.4777006208896637,
+      "learning_rate": 2.494276402599417e-05,
+      "loss": 0.3619,
+      "step": 10569
+    },
+    {
+      "epoch": 0.7772630340466211,
+      "grad_norm": 0.4395066201686859,
+      "learning_rate": 2.4927028700763e-05,
+      "loss": 0.3374,
+      "step": 10570
+    },
+    {
+      "epoch": 0.7773365688653577,
+      "grad_norm": 0.47275543212890625,
+      "learning_rate": 2.4911297633749965e-05,
+      "loss": 0.3366,
+      "step": 10571
+    },
+    {
+      "epoch": 0.7774101036840945,
+      "grad_norm": 0.5044657588005066,
+      "learning_rate": 2.489557082584736e-05,
+      "loss": 0.3723,
+      "step": 10572
+    },
+    {
+      "epoch": 0.7774836385028311,
+      "grad_norm": 0.45943716168403625,
+      "learning_rate": 2.4879848277947248e-05,
+      "loss": 0.3631,
+      "step": 10573
+    },
+    {
+      "epoch": 0.7775571733215677,
+      "grad_norm": 0.5474114418029785,
+      "learning_rate": 2.4864129990941454e-05,
+      "loss": 0.3531,
+      "step": 10574
+    },
+    {
+      "epoch": 0.7776307081403044,
+      "grad_norm": 0.5281530022621155,
+      "learning_rate": 2.4848415965721462e-05,
+      "loss": 0.3667,
+      "step": 10575
+    },
+    {
+      "epoch": 0.7777042429590411,
+      "grad_norm": 0.537030041217804,
+      "learning_rate": 2.4832706203178623e-05,
+      "loss": 0.3666,
+      "step": 10576
+    },
+    {
+      "epoch": 0.7777777777777778,
+      "grad_norm": 0.6168251037597656,
+      "learning_rate": 2.4817000704204053e-05,
+      "loss": 0.3445,
+      "step": 10577
+    },
+    {
+      "epoch": 0.7778513125965144,
+      "grad_norm": 0.5184450745582581,
+      "learning_rate": 2.480129946968852e-05,
+      "loss": 0.3516,
+      "step": 10578
+    },
+    {
+      "epoch": 0.7779248474152511,
+      "grad_norm": 0.6287129521369934,
+      "learning_rate": 2.4785602500522632e-05,
+      "loss": 0.3542,
+      "step": 10579
+    },
+    {
+      "epoch": 0.7779983822339878,
+      "grad_norm": 0.47351688146591187,
+      "learning_rate": 2.4769909797596814e-05,
+      "loss": 0.3521,
+      "step": 10580
+    },
+    {
+      "epoch": 0.7780719170527245,
+      "grad_norm": 0.47180116176605225,
+      "learning_rate": 2.4754221361801045e-05,
+      "loss": 0.3419,
+      "step": 10581
+    },
+    {
+      "epoch": 0.7781454518714611,
+      "grad_norm": 0.6352446675300598,
+      "learning_rate": 2.4738537194025248e-05,
+      "loss": 0.3162,
+      "step": 10582
+    },
+    {
+      "epoch": 0.7782189866901978,
+      "grad_norm": 0.4767645001411438,
+      "learning_rate": 2.4722857295159075e-05,
+      "loss": 0.363,
+      "step": 10583
+    },
+    {
+      "epoch": 0.7782925215089345,
+      "grad_norm": 0.567162811756134,
+      "learning_rate": 2.470718166609185e-05,
+      "loss": 0.3609,
+      "step": 10584
+    },
+    {
+      "epoch": 0.7783660563276712,
+      "grad_norm": 0.5194567441940308,
+      "learning_rate": 2.469151030771275e-05,
+      "loss": 0.3493,
+      "step": 10585
+    },
+    {
+      "epoch": 0.7784395911464078,
+      "grad_norm": 0.6889566779136658,
+      "learning_rate": 2.4675843220910654e-05,
+      "loss": 0.3678,
+      "step": 10586
+    },
+    {
+      "epoch": 0.7785131259651445,
+      "grad_norm": 0.7491114735603333,
+      "learning_rate": 2.4660180406574253e-05,
+      "loss": 0.3647,
+      "step": 10587
+    },
+    {
+      "epoch": 0.7785866607838812,
+      "grad_norm": 0.5491151213645935,
+      "learning_rate": 2.464452186559191e-05,
+      "loss": 0.3604,
+      "step": 10588
+    },
+    {
+      "epoch": 0.7786601956026178,
+      "grad_norm": 0.8120251297950745,
+      "learning_rate": 2.462886759885179e-05,
+      "loss": 0.373,
+      "step": 10589
+    },
+    {
+      "epoch": 0.7787337304213545,
+      "grad_norm": 0.5274749994277954,
+      "learning_rate": 2.461321760724189e-05,
+      "loss": 0.3542,
+      "step": 10590
+    },
+    {
+      "epoch": 0.7788072652400911,
+      "grad_norm": 0.7074584364891052,
+      "learning_rate": 2.4597571891649805e-05,
+      "loss": 0.3583,
+      "step": 10591
+    },
+    {
+      "epoch": 0.7788808000588279,
+      "grad_norm": 0.48851463198661804,
+      "learning_rate": 2.4581930452963064e-05,
+      "loss": 0.3535,
+      "step": 10592
+    },
+    {
+      "epoch": 0.7789543348775645,
+      "grad_norm": 0.5534822940826416,
+      "learning_rate": 2.4566293292068798e-05,
+      "loss": 0.3554,
+      "step": 10593
+    },
+    {
+      "epoch": 0.7790278696963012,
+      "grad_norm": 0.4347151219844818,
+      "learning_rate": 2.4550660409853977e-05,
+      "loss": 0.353,
+      "step": 10594
+    },
+    {
+      "epoch": 0.7791014045150378,
+      "grad_norm": 0.5010644793510437,
+      "learning_rate": 2.4535031807205312e-05,
+      "loss": 0.3738,
+      "step": 10595
+    },
+    {
+      "epoch": 0.7791749393337746,
+      "grad_norm": 0.605037271976471,
+      "learning_rate": 2.4519407485009284e-05,
+      "loss": 0.3624,
+      "step": 10596
+    },
+    {
+      "epoch": 0.7792484741525112,
+      "grad_norm": 0.4438419044017792,
+      "learning_rate": 2.450378744415215e-05,
+      "loss": 0.3485,
+      "step": 10597
+    },
+    {
+      "epoch": 0.7793220089712479,
+      "grad_norm": 0.39747974276542664,
+      "learning_rate": 2.4488171685519835e-05,
+      "loss": 0.327,
+      "step": 10598
+    },
+    {
+      "epoch": 0.7793955437899845,
+      "grad_norm": 0.4034651815891266,
+      "learning_rate": 2.447256020999812e-05,
+      "loss": 0.3745,
+      "step": 10599
+    },
+    {
+      "epoch": 0.7794690786087213,
+      "grad_norm": 0.4465847313404083,
+      "learning_rate": 2.4456953018472516e-05,
+      "loss": 0.3546,
+      "step": 10600
+    },
+    {
+      "epoch": 0.7795426134274579,
+      "grad_norm": 0.39490798115730286,
+      "learning_rate": 2.4441350111828233e-05,
+      "loss": 0.3409,
+      "step": 10601
+    },
+    {
+      "epoch": 0.7796161482461945,
+      "grad_norm": 0.4343103766441345,
+      "learning_rate": 2.442575149095032e-05,
+      "loss": 0.3517,
+      "step": 10602
+    },
+    {
+      "epoch": 0.7796896830649312,
+      "grad_norm": 0.49761685729026794,
+      "learning_rate": 2.4410157156723567e-05,
+      "loss": 0.3666,
+      "step": 10603
+    },
+    {
+      "epoch": 0.7797632178836679,
+      "grad_norm": 0.47463440895080566,
+      "learning_rate": 2.4394567110032475e-05,
+      "loss": 0.3583,
+      "step": 10604
+    },
+    {
+      "epoch": 0.7798367527024046,
+      "grad_norm": 0.49345290660858154,
+      "learning_rate": 2.4378981351761287e-05,
+      "loss": 0.3454,
+      "step": 10605
+    },
+    {
+      "epoch": 0.7799102875211412,
+      "grad_norm": 0.46642109751701355,
+      "learning_rate": 2.4363399882794114e-05,
+      "loss": 0.3717,
+      "step": 10606
+    },
+    {
+      "epoch": 0.779983822339878,
+      "grad_norm": 0.5842941403388977,
+      "learning_rate": 2.43478227040147e-05,
+      "loss": 0.3773,
+      "step": 10607
+    },
+    {
+      "epoch": 0.7800573571586146,
+      "grad_norm": 0.5384358763694763,
+      "learning_rate": 2.433224981630662e-05,
+      "loss": 0.3506,
+      "step": 10608
+    },
+    {
+      "epoch": 0.7801308919773513,
+      "grad_norm": 0.6730407476425171,
+      "learning_rate": 2.4316681220553185e-05,
+      "loss": 0.3658,
+      "step": 10609
+    },
+    {
+      "epoch": 0.7802044267960879,
+      "grad_norm": 0.4407198131084442,
+      "learning_rate": 2.430111691763749e-05,
+      "loss": 0.3547,
+      "step": 10610
+    },
+    {
+      "epoch": 0.7802779616148247,
+      "grad_norm": 0.525498628616333,
+      "learning_rate": 2.4285556908442297e-05,
+      "loss": 0.355,
+      "step": 10611
+    },
+    {
+      "epoch": 0.7803514964335613,
+      "grad_norm": 0.5791994333267212,
+      "learning_rate": 2.427000119385022e-05,
+      "loss": 0.377,
+      "step": 10612
+    },
+    {
+      "epoch": 0.780425031252298,
+      "grad_norm": 0.41634297370910645,
+      "learning_rate": 2.425444977474364e-05,
+      "loss": 0.362,
+      "step": 10613
+    },
+    {
+      "epoch": 0.7804985660710346,
+      "grad_norm": 0.6777190566062927,
+      "learning_rate": 2.4238902652004568e-05,
+      "loss": 0.3506,
+      "step": 10614
+    },
+    {
+      "epoch": 0.7805721008897714,
+      "grad_norm": 0.448273241519928,
+      "learning_rate": 2.4223359826514923e-05,
+      "loss": 0.352,
+      "step": 10615
+    },
+    {
+      "epoch": 0.780645635708508,
+      "grad_norm": 0.45261919498443604,
+      "learning_rate": 2.4207821299156275e-05,
+      "loss": 0.3417,
+      "step": 10616
+    },
+    {
+      "epoch": 0.7807191705272446,
+      "grad_norm": 0.4605804681777954,
+      "learning_rate": 2.4192287070809972e-05,
+      "loss": 0.339,
+      "step": 10617
+    },
+    {
+      "epoch": 0.7807927053459813,
+      "grad_norm": 0.6142990589141846,
+      "learning_rate": 2.4176757142357132e-05,
+      "loss": 0.3382,
+      "step": 10618
+    },
+    {
+      "epoch": 0.780866240164718,
+      "grad_norm": 0.42942506074905396,
+      "learning_rate": 2.4161231514678682e-05,
+      "loss": 0.3555,
+      "step": 10619
+    },
+    {
+      "epoch": 0.7809397749834547,
+      "grad_norm": 0.5398353338241577,
+      "learning_rate": 2.4145710188655192e-05,
+      "loss": 0.3606,
+      "step": 10620
+    },
+    {
+      "epoch": 0.7810133098021913,
+      "grad_norm": 0.4520362913608551,
+      "learning_rate": 2.413019316516707e-05,
+      "loss": 0.3875,
+      "step": 10621
+    },
+    {
+      "epoch": 0.781086844620928,
+      "grad_norm": 0.5517558455467224,
+      "learning_rate": 2.4114680445094462e-05,
+      "loss": 0.3546,
+      "step": 10622
+    },
+    {
+      "epoch": 0.7811603794396647,
+      "grad_norm": 0.5241686701774597,
+      "learning_rate": 2.4099172029317296e-05,
+      "loss": 0.3484,
+      "step": 10623
+    },
+    {
+      "epoch": 0.7812339142584014,
+      "grad_norm": 0.4716179072856903,
+      "learning_rate": 2.4083667918715158e-05,
+      "loss": 0.3661,
+      "step": 10624
+    },
+    {
+      "epoch": 0.781307449077138,
+      "grad_norm": 0.47592198848724365,
+      "learning_rate": 2.406816811416749e-05,
+      "loss": 0.3641,
+      "step": 10625
+    },
+    {
+      "epoch": 0.7813809838958747,
+      "grad_norm": 0.4388347864151001,
+      "learning_rate": 2.4052672616553494e-05,
+      "loss": 0.3813,
+      "step": 10626
+    },
+    {
+      "epoch": 0.7814545187146114,
+      "grad_norm": 0.5450957417488098,
+      "learning_rate": 2.4037181426752064e-05,
+      "loss": 0.3644,
+      "step": 10627
+    },
+    {
+      "epoch": 0.7815280535333481,
+      "grad_norm": 0.4556047320365906,
+      "learning_rate": 2.4021694545641837e-05,
+      "loss": 0.3728,
+      "step": 10628
+    },
+    {
+      "epoch": 0.7816015883520847,
+      "grad_norm": 0.5777156352996826,
+      "learning_rate": 2.40062119741013e-05,
+      "loss": 0.3549,
+      "step": 10629
+    },
+    {
+      "epoch": 0.7816751231708213,
+      "grad_norm": 0.40971630811691284,
+      "learning_rate": 2.3990733713008595e-05,
+      "loss": 0.3666,
+      "step": 10630
+    },
+    {
+      "epoch": 0.7817486579895581,
+      "grad_norm": 0.5470374822616577,
+      "learning_rate": 2.39752597632417e-05,
+      "loss": 0.3871,
+      "step": 10631
+    },
+    {
+      "epoch": 0.7818221928082947,
+      "grad_norm": 0.4534602463245392,
+      "learning_rate": 2.3959790125678283e-05,
+      "loss": 0.3805,
+      "step": 10632
+    },
+    {
+      "epoch": 0.7818957276270314,
+      "grad_norm": 0.5198513269424438,
+      "learning_rate": 2.394432480119586e-05,
+      "loss": 0.3495,
+      "step": 10633
+    },
+    {
+      "epoch": 0.781969262445768,
+      "grad_norm": 0.39684823155403137,
+      "learning_rate": 2.3928863790671574e-05,
+      "loss": 0.3623,
+      "step": 10634
+    },
+    {
+      "epoch": 0.7820427972645048,
+      "grad_norm": 0.45460066199302673,
+      "learning_rate": 2.391340709498241e-05,
+      "loss": 0.3555,
+      "step": 10635
+    },
+    {
+      "epoch": 0.7821163320832414,
+      "grad_norm": 0.4652835428714752,
+      "learning_rate": 2.3897954715005124e-05,
+      "loss": 0.3587,
+      "step": 10636
+    },
+    {
+      "epoch": 0.7821898669019781,
+      "grad_norm": 0.5056663751602173,
+      "learning_rate": 2.388250665161612e-05,
+      "loss": 0.3627,
+      "step": 10637
+    },
+    {
+      "epoch": 0.7822634017207147,
+      "grad_norm": 0.48958760499954224,
+      "learning_rate": 2.3867062905691683e-05,
+      "loss": 0.3611,
+      "step": 10638
+    },
+    {
+      "epoch": 0.7823369365394515,
+      "grad_norm": 0.5001001954078674,
+      "learning_rate": 2.3851623478107832e-05,
+      "loss": 0.3661,
+      "step": 10639
+    },
+    {
+      "epoch": 0.7824104713581881,
+      "grad_norm": 0.4423164129257202,
+      "learning_rate": 2.383618836974021e-05,
+      "loss": 0.3344,
+      "step": 10640
+    },
+    {
+      "epoch": 0.7824840061769248,
+      "grad_norm": 0.4037863612174988,
+      "learning_rate": 2.382075758146436e-05,
+      "loss": 0.3522,
+      "step": 10641
+    },
+    {
+      "epoch": 0.7825575409956614,
+      "grad_norm": 0.41805246472358704,
+      "learning_rate": 2.380533111415556e-05,
+      "loss": 0.3554,
+      "step": 10642
+    },
+    {
+      "epoch": 0.7826310758143982,
+      "grad_norm": 0.5237528085708618,
+      "learning_rate": 2.378990896868877e-05,
+      "loss": 0.3484,
+      "step": 10643
+    },
+    {
+      "epoch": 0.7827046106331348,
+      "grad_norm": 0.4074634313583374,
+      "learning_rate": 2.377449114593876e-05,
+      "loss": 0.3852,
+      "step": 10644
+    },
+    {
+      "epoch": 0.7827781454518714,
+      "grad_norm": 0.4543856382369995,
+      "learning_rate": 2.375907764678007e-05,
+      "loss": 0.3721,
+      "step": 10645
+    },
+    {
+      "epoch": 0.7828516802706081,
+      "grad_norm": 0.4262408912181854,
+      "learning_rate": 2.3743668472086976e-05,
+      "loss": 0.3438,
+      "step": 10646
+    },
+    {
+      "epoch": 0.7829252150893448,
+      "grad_norm": 0.4564390778541565,
+      "learning_rate": 2.372826362273345e-05,
+      "loss": 0.3613,
+      "step": 10647
+    },
+    {
+      "epoch": 0.7829987499080815,
+      "grad_norm": 0.45421263575553894,
+      "learning_rate": 2.3712863099593307e-05,
+      "loss": 0.3976,
+      "step": 10648
+    },
+    {
+      "epoch": 0.7830722847268181,
+      "grad_norm": 0.38535910844802856,
+      "learning_rate": 2.369746690354011e-05,
+      "loss": 0.3747,
+      "step": 10649
+    },
+    {
+      "epoch": 0.7831458195455548,
+      "grad_norm": 0.4935294985771179,
+      "learning_rate": 2.3682075035447116e-05,
+      "loss": 0.35,
+      "step": 10650
+    },
+    {
+      "epoch": 0.7832193543642915,
+      "grad_norm": 0.4412441551685333,
+      "learning_rate": 2.3666687496187346e-05,
+      "loss": 0.3673,
+      "step": 10651
+    },
+    {
+      "epoch": 0.7832928891830282,
+      "grad_norm": 0.512861430644989,
+      "learning_rate": 2.3651304286633635e-05,
+      "loss": 0.382,
+      "step": 10652
+    },
+    {
+      "epoch": 0.7833664240017648,
+      "grad_norm": 0.464360773563385,
+      "learning_rate": 2.363592540765849e-05,
+      "loss": 0.3524,
+      "step": 10653
+    },
+    {
+      "epoch": 0.7834399588205015,
+      "grad_norm": 0.49043965339660645,
+      "learning_rate": 2.362055086013425e-05,
+      "loss": 0.3562,
+      "step": 10654
+    },
+    {
+      "epoch": 0.7835134936392382,
+      "grad_norm": 0.4541548490524292,
+      "learning_rate": 2.3605180644932977e-05,
+      "loss": 0.3784,
+      "step": 10655
+    },
+    {
+      "epoch": 0.7835870284579749,
+      "grad_norm": 0.6173545718193054,
+      "learning_rate": 2.35898147629265e-05,
+      "loss": 0.3523,
+      "step": 10656
+    },
+    {
+      "epoch": 0.7836605632767115,
+      "grad_norm": 0.6138259768486023,
+      "learning_rate": 2.3574453214986348e-05,
+      "loss": 0.3611,
+      "step": 10657
+    },
+    {
+      "epoch": 0.7837340980954481,
+      "grad_norm": 0.4531562030315399,
+      "learning_rate": 2.355909600198386e-05,
+      "loss": 0.3379,
+      "step": 10658
+    },
+    {
+      "epoch": 0.7838076329141849,
+      "grad_norm": 0.6939131617546082,
+      "learning_rate": 2.354374312479014e-05,
+      "loss": 0.3616,
+      "step": 10659
+    },
+    {
+      "epoch": 0.7838811677329215,
+      "grad_norm": 0.4573412537574768,
+      "learning_rate": 2.3528394584275982e-05,
+      "loss": 0.3757,
+      "step": 10660
+    },
+    {
+      "epoch": 0.7839547025516582,
+      "grad_norm": 0.659293532371521,
+      "learning_rate": 2.3513050381311987e-05,
+      "loss": 0.3373,
+      "step": 10661
+    },
+    {
+      "epoch": 0.7840282373703948,
+      "grad_norm": 0.47627297043800354,
+      "learning_rate": 2.3497710516768534e-05,
+      "loss": 0.3691,
+      "step": 10662
+    },
+    {
+      "epoch": 0.7841017721891316,
+      "grad_norm": 0.7074831128120422,
+      "learning_rate": 2.3482374991515632e-05,
+      "loss": 0.337,
+      "step": 10663
+    },
+    {
+      "epoch": 0.7841753070078682,
+      "grad_norm": 0.4223443567752838,
+      "learning_rate": 2.346704380642317e-05,
+      "loss": 0.3669,
+      "step": 10664
+    },
+    {
+      "epoch": 0.7842488418266049,
+      "grad_norm": 0.6387078762054443,
+      "learning_rate": 2.345171696236078e-05,
+      "loss": 0.326,
+      "step": 10665
+    },
+    {
+      "epoch": 0.7843223766453415,
+      "grad_norm": 0.5832350850105286,
+      "learning_rate": 2.3436394460197762e-05,
+      "loss": 0.3724,
+      "step": 10666
+    },
+    {
+      "epoch": 0.7843959114640783,
+      "grad_norm": 0.591362714767456,
+      "learning_rate": 2.3421076300803256e-05,
+      "loss": 0.3581,
+      "step": 10667
+    },
+    {
+      "epoch": 0.7844694462828149,
+      "grad_norm": 0.4242936372756958,
+      "learning_rate": 2.340576248504611e-05,
+      "loss": 0.3718,
+      "step": 10668
+    },
+    {
+      "epoch": 0.7845429811015516,
+      "grad_norm": 0.49627089500427246,
+      "learning_rate": 2.339045301379499e-05,
+      "loss": 0.3521,
+      "step": 10669
+    },
+    {
+      "epoch": 0.7846165159202882,
+      "grad_norm": 0.45726504921913147,
+      "learning_rate": 2.3375147887918182e-05,
+      "loss": 0.3414,
+      "step": 10670
+    },
+    {
+      "epoch": 0.784690050739025,
+      "grad_norm": 0.425422728061676,
+      "learning_rate": 2.335984710828386e-05,
+      "loss": 0.3593,
+      "step": 10671
+    },
+    {
+      "epoch": 0.7847635855577616,
+      "grad_norm": 0.6551979184150696,
+      "learning_rate": 2.3344550675759914e-05,
+      "loss": 0.372,
+      "step": 10672
+    },
+    {
+      "epoch": 0.7848371203764982,
+      "grad_norm": 0.5363835692405701,
+      "learning_rate": 2.3329258591213932e-05,
+      "loss": 0.3436,
+      "step": 10673
+    },
+    {
+      "epoch": 0.784910655195235,
+      "grad_norm": 0.37946343421936035,
+      "learning_rate": 2.3313970855513335e-05,
+      "loss": 0.3411,
+      "step": 10674
+    },
+    {
+      "epoch": 0.7849841900139716,
+      "grad_norm": 0.47457852959632874,
+      "learning_rate": 2.329868746952524e-05,
+      "loss": 0.3664,
+      "step": 10675
+    },
+    {
+      "epoch": 0.7850577248327083,
+      "grad_norm": 0.4603305459022522,
+      "learning_rate": 2.3283408434116515e-05,
+      "loss": 0.3298,
+      "step": 10676
+    },
+    {
+      "epoch": 0.7851312596514449,
+      "grad_norm": 0.46417316794395447,
+      "learning_rate": 2.326813375015383e-05,
+      "loss": 0.3449,
+      "step": 10677
+    },
+    {
+      "epoch": 0.7852047944701817,
+      "grad_norm": 0.47684624791145325,
+      "learning_rate": 2.3252863418503568e-05,
+      "loss": 0.3338,
+      "step": 10678
+    },
+    {
+      "epoch": 0.7852783292889183,
+      "grad_norm": 0.5281082987785339,
+      "learning_rate": 2.3237597440031923e-05,
+      "loss": 0.3431,
+      "step": 10679
+    },
+    {
+      "epoch": 0.785351864107655,
+      "grad_norm": 0.4218333065509796,
+      "learning_rate": 2.3222335815604723e-05,
+      "loss": 0.3443,
+      "step": 10680
+    },
+    {
+      "epoch": 0.7854253989263916,
+      "grad_norm": 0.61531662940979,
+      "learning_rate": 2.3207078546087667e-05,
+      "loss": 0.3671,
+      "step": 10681
+    },
+    {
+      "epoch": 0.7854989337451284,
+      "grad_norm": 0.4415034055709839,
+      "learning_rate": 2.3191825632346198e-05,
+      "loss": 0.3473,
+      "step": 10682
+    },
+    {
+      "epoch": 0.785572468563865,
+      "grad_norm": 0.43520331382751465,
+      "learning_rate": 2.3176577075245397e-05,
+      "loss": 0.3538,
+      "step": 10683
+    },
+    {
+      "epoch": 0.7856460033826017,
+      "grad_norm": 0.4117346405982971,
+      "learning_rate": 2.3161332875650222e-05,
+      "loss": 0.3645,
+      "step": 10684
+    },
+    {
+      "epoch": 0.7857195382013383,
+      "grad_norm": 0.4212453067302704,
+      "learning_rate": 2.3146093034425397e-05,
+      "loss": 0.3585,
+      "step": 10685
+    },
+    {
+      "epoch": 0.785793073020075,
+      "grad_norm": 0.46074140071868896,
+      "learning_rate": 2.3130857552435225e-05,
+      "loss": 0.3419,
+      "step": 10686
+    },
+    {
+      "epoch": 0.7858666078388117,
+      "grad_norm": 0.4538916051387787,
+      "learning_rate": 2.3115626430543934e-05,
+      "loss": 0.3949,
+      "step": 10687
+    },
+    {
+      "epoch": 0.7859401426575483,
+      "grad_norm": 0.48705732822418213,
+      "learning_rate": 2.310039966961548e-05,
+      "loss": 0.3493,
+      "step": 10688
+    },
+    {
+      "epoch": 0.786013677476285,
+      "grad_norm": 0.4243554174900055,
+      "learning_rate": 2.308517727051348e-05,
+      "loss": 0.3548,
+      "step": 10689
+    },
+    {
+      "epoch": 0.7860872122950217,
+      "grad_norm": 0.4443734884262085,
+      "learning_rate": 2.3069959234101403e-05,
+      "loss": 0.3642,
+      "step": 10690
+    },
+    {
+      "epoch": 0.7861607471137584,
+      "grad_norm": 0.42824774980545044,
+      "learning_rate": 2.305474556124243e-05,
+      "loss": 0.3662,
+      "step": 10691
+    },
+    {
+      "epoch": 0.786234281932495,
+      "grad_norm": 0.3882371187210083,
+      "learning_rate": 2.303953625279951e-05,
+      "loss": 0.3679,
+      "step": 10692
+    },
+    {
+      "epoch": 0.7863078167512317,
+      "grad_norm": 0.4676635265350342,
+      "learning_rate": 2.3024331309635283e-05,
+      "loss": 0.3454,
+      "step": 10693
+    },
+    {
+      "epoch": 0.7863813515699684,
+      "grad_norm": 0.4969755709171295,
+      "learning_rate": 2.300913073261223e-05,
+      "loss": 0.3564,
+      "step": 10694
+    },
+    {
+      "epoch": 0.7864548863887051,
+      "grad_norm": 0.4461078941822052,
+      "learning_rate": 2.2993934522592565e-05,
+      "loss": 0.3949,
+      "step": 10695
+    },
+    {
+      "epoch": 0.7865284212074417,
+      "grad_norm": 0.46380406618118286,
+      "learning_rate": 2.297874268043816e-05,
+      "loss": 0.3354,
+      "step": 10696
+    },
+    {
+      "epoch": 0.7866019560261784,
+      "grad_norm": 0.4022988975048065,
+      "learning_rate": 2.296355520701079e-05,
+      "loss": 0.3929,
+      "step": 10697
+    },
+    {
+      "epoch": 0.7866754908449151,
+      "grad_norm": 0.43358346819877625,
+      "learning_rate": 2.2948372103171868e-05,
+      "loss": 0.3786,
+      "step": 10698
+    },
+    {
+      "epoch": 0.7867490256636518,
+      "grad_norm": 0.6392577886581421,
+      "learning_rate": 2.2933193369782567e-05,
+      "loss": 0.3983,
+      "step": 10699
+    },
+    {
+      "epoch": 0.7868225604823884,
+      "grad_norm": 0.48019087314605713,
+      "learning_rate": 2.291801900770386e-05,
+      "loss": 0.3625,
+      "step": 10700
+    },
+    {
+      "epoch": 0.786896095301125,
+      "grad_norm": 0.6047197580337524,
+      "learning_rate": 2.290284901779648e-05,
+      "loss": 0.3743,
+      "step": 10701
+    },
+    {
+      "epoch": 0.7869696301198618,
+      "grad_norm": 0.48192471265792847,
+      "learning_rate": 2.288768340092089e-05,
+      "loss": 0.3335,
+      "step": 10702
+    },
+    {
+      "epoch": 0.7870431649385984,
+      "grad_norm": 0.7216126918792725,
+      "learning_rate": 2.2872522157937237e-05,
+      "loss": 0.3575,
+      "step": 10703
+    },
+    {
+      "epoch": 0.7871166997573351,
+      "grad_norm": 0.5395556092262268,
+      "learning_rate": 2.2857365289705536e-05,
+      "loss": 0.356,
+      "step": 10704
+    },
+    {
+      "epoch": 0.7871902345760717,
+      "grad_norm": 0.7421590089797974,
+      "learning_rate": 2.2842212797085517e-05,
+      "loss": 0.3266,
+      "step": 10705
+    },
+    {
+      "epoch": 0.7872637693948085,
+      "grad_norm": 0.5836133360862732,
+      "learning_rate": 2.282706468093658e-05,
+      "loss": 0.3587,
+      "step": 10706
+    },
+    {
+      "epoch": 0.7873373042135451,
+      "grad_norm": 0.46477508544921875,
+      "learning_rate": 2.281192094211798e-05,
+      "loss": 0.3621,
+      "step": 10707
+    },
+    {
+      "epoch": 0.7874108390322818,
+      "grad_norm": 0.6598595976829529,
+      "learning_rate": 2.2796781581488714e-05,
+      "loss": 0.3865,
+      "step": 10708
+    },
+    {
+      "epoch": 0.7874843738510184,
+      "grad_norm": 0.48838961124420166,
+      "learning_rate": 2.278164659990748e-05,
+      "loss": 0.3828,
+      "step": 10709
+    },
+    {
+      "epoch": 0.7875579086697552,
+      "grad_norm": 0.5893398523330688,
+      "learning_rate": 2.2766515998232706e-05,
+      "loss": 0.3356,
+      "step": 10710
+    },
+    {
+      "epoch": 0.7876314434884918,
+      "grad_norm": 0.4404858648777008,
+      "learning_rate": 2.2751389777322685e-05,
+      "loss": 0.3488,
+      "step": 10711
+    },
+    {
+      "epoch": 0.7877049783072285,
+      "grad_norm": 0.42267709970474243,
+      "learning_rate": 2.273626793803533e-05,
+      "loss": 0.3456,
+      "step": 10712
+    },
+    {
+      "epoch": 0.7877785131259651,
+      "grad_norm": 0.45640406012535095,
+      "learning_rate": 2.272115048122839e-05,
+      "loss": 0.3351,
+      "step": 10713
+    },
+    {
+      "epoch": 0.7878520479447018,
+      "grad_norm": 0.4079216420650482,
+      "learning_rate": 2.270603740775935e-05,
+      "loss": 0.3394,
+      "step": 10714
+    },
+    {
+      "epoch": 0.7879255827634385,
+      "grad_norm": 0.5509805083274841,
+      "learning_rate": 2.2690928718485472e-05,
+      "loss": 0.3499,
+      "step": 10715
+    },
+    {
+      "epoch": 0.7879991175821751,
+      "grad_norm": 0.6195715069770813,
+      "learning_rate": 2.267582441426368e-05,
+      "loss": 0.3753,
+      "step": 10716
+    },
+    {
+      "epoch": 0.7880726524009118,
+      "grad_norm": 0.4716539680957794,
+      "learning_rate": 2.2660724495950737e-05,
+      "loss": 0.3588,
+      "step": 10717
+    },
+    {
+      "epoch": 0.7881461872196485,
+      "grad_norm": 0.5240958333015442,
+      "learning_rate": 2.264562896440313e-05,
+      "loss": 0.4115,
+      "step": 10718
+    },
+    {
+      "epoch": 0.7882197220383852,
+      "grad_norm": 0.5998661518096924,
+      "learning_rate": 2.2630537820477073e-05,
+      "loss": 0.3511,
+      "step": 10719
+    },
+    {
+      "epoch": 0.7882932568571218,
+      "grad_norm": 0.43058568239212036,
+      "learning_rate": 2.2615451065028583e-05,
+      "loss": 0.3654,
+      "step": 10720
+    },
+    {
+      "epoch": 0.7883667916758585,
+      "grad_norm": 0.5165978670120239,
+      "learning_rate": 2.2600368698913367e-05,
+      "loss": 0.3741,
+      "step": 10721
+    },
+    {
+      "epoch": 0.7884403264945952,
+      "grad_norm": 0.44582700729370117,
+      "learning_rate": 2.2585290722986908e-05,
+      "loss": 0.3994,
+      "step": 10722
+    },
+    {
+      "epoch": 0.7885138613133319,
+      "grad_norm": 0.43858012557029724,
+      "learning_rate": 2.2570217138104442e-05,
+      "loss": 0.3332,
+      "step": 10723
+    },
+    {
+      "epoch": 0.7885873961320685,
+      "grad_norm": 0.49110448360443115,
+      "learning_rate": 2.255514794512099e-05,
+      "loss": 0.3384,
+      "step": 10724
+    },
+    {
+      "epoch": 0.7886609309508053,
+      "grad_norm": 0.6330487728118896,
+      "learning_rate": 2.25400831448913e-05,
+      "loss": 0.3471,
+      "step": 10725
+    },
+    {
+      "epoch": 0.7887344657695419,
+      "grad_norm": 0.43700110912323,
+      "learning_rate": 2.252502273826981e-05,
+      "loss": 0.3366,
+      "step": 10726
+    },
+    {
+      "epoch": 0.7888080005882786,
+      "grad_norm": 0.5929771661758423,
+      "learning_rate": 2.2509966726110798e-05,
+      "loss": 0.3473,
+      "step": 10727
+    },
+    {
+      "epoch": 0.7888815354070152,
+      "grad_norm": 0.7596095204353333,
+      "learning_rate": 2.2494915109268278e-05,
+      "loss": 0.3557,
+      "step": 10728
+    },
+    {
+      "epoch": 0.7889550702257518,
+      "grad_norm": 0.43711820244789124,
+      "learning_rate": 2.247986788859594e-05,
+      "loss": 0.3435,
+      "step": 10729
+    },
+    {
+      "epoch": 0.7890286050444886,
+      "grad_norm": 0.5477190017700195,
+      "learning_rate": 2.2464825064947302e-05,
+      "loss": 0.3303,
+      "step": 10730
+    },
+    {
+      "epoch": 0.7891021398632252,
+      "grad_norm": 0.46568411588668823,
+      "learning_rate": 2.244978663917564e-05,
+      "loss": 0.3383,
+      "step": 10731
+    },
+    {
+      "epoch": 0.7891756746819619,
+      "grad_norm": 0.4371599555015564,
+      "learning_rate": 2.2434752612133926e-05,
+      "loss": 0.341,
+      "step": 10732
+    },
+    {
+      "epoch": 0.7892492095006985,
+      "grad_norm": 0.5708867907524109,
+      "learning_rate": 2.2419722984674875e-05,
+      "loss": 0.3213,
+      "step": 10733
+    },
+    {
+      "epoch": 0.7893227443194353,
+      "grad_norm": 0.5263074040412903,
+      "learning_rate": 2.2404697757651026e-05,
+      "loss": 0.3581,
+      "step": 10734
+    },
+    {
+      "epoch": 0.7893962791381719,
+      "grad_norm": 0.48730191588401794,
+      "learning_rate": 2.2389676931914594e-05,
+      "loss": 0.3668,
+      "step": 10735
+    },
+    {
+      "epoch": 0.7894698139569086,
+      "grad_norm": 0.5281966328620911,
+      "learning_rate": 2.2374660508317592e-05,
+      "loss": 0.3691,
+      "step": 10736
+    },
+    {
+      "epoch": 0.7895433487756452,
+      "grad_norm": 0.6719037294387817,
+      "learning_rate": 2.235964848771176e-05,
+      "loss": 0.3459,
+      "step": 10737
+    },
+    {
+      "epoch": 0.789616883594382,
+      "grad_norm": 0.44478800892829895,
+      "learning_rate": 2.234464087094863e-05,
+      "loss": 0.3406,
+      "step": 10738
+    },
+    {
+      "epoch": 0.7896904184131186,
+      "grad_norm": 0.433542937040329,
+      "learning_rate": 2.232963765887941e-05,
+      "loss": 0.3471,
+      "step": 10739
+    },
+    {
+      "epoch": 0.7897639532318553,
+      "grad_norm": 0.47773218154907227,
+      "learning_rate": 2.2314638852355098e-05,
+      "loss": 0.3524,
+      "step": 10740
+    },
+    {
+      "epoch": 0.789837488050592,
+      "grad_norm": 0.5175096988677979,
+      "learning_rate": 2.2299644452226485e-05,
+      "loss": 0.3537,
+      "step": 10741
+    },
+    {
+      "epoch": 0.7899110228693286,
+      "grad_norm": 0.5012705326080322,
+      "learning_rate": 2.2284654459344013e-05,
+      "loss": 0.3408,
+      "step": 10742
+    },
+    {
+      "epoch": 0.7899845576880653,
+      "grad_norm": 0.44458675384521484,
+      "learning_rate": 2.2269668874557947e-05,
+      "loss": 0.3705,
+      "step": 10743
+    },
+    {
+      "epoch": 0.7900580925068019,
+      "grad_norm": 0.48008912801742554,
+      "learning_rate": 2.2254687698718356e-05,
+      "loss": 0.3306,
+      "step": 10744
+    },
+    {
+      "epoch": 0.7901316273255387,
+      "grad_norm": 0.6235682964324951,
+      "learning_rate": 2.2239710932674874e-05,
+      "loss": 0.3416,
+      "step": 10745
+    },
+    {
+      "epoch": 0.7902051621442753,
+      "grad_norm": 0.4517041742801666,
+      "learning_rate": 2.2224738577277048e-05,
+      "loss": 0.3456,
+      "step": 10746
+    },
+    {
+      "epoch": 0.790278696963012,
+      "grad_norm": 0.5755457282066345,
+      "learning_rate": 2.2209770633374138e-05,
+      "loss": 0.3545,
+      "step": 10747
+    },
+    {
+      "epoch": 0.7903522317817486,
+      "grad_norm": 0.5316616892814636,
+      "learning_rate": 2.219480710181515e-05,
+      "loss": 0.3465,
+      "step": 10748
+    },
+    {
+      "epoch": 0.7904257666004854,
+      "grad_norm": 0.49133729934692383,
+      "learning_rate": 2.21798479834488e-05,
+      "loss": 0.3981,
+      "step": 10749
+    },
+    {
+      "epoch": 0.790499301419222,
+      "grad_norm": 0.7161216139793396,
+      "learning_rate": 2.216489327912359e-05,
+      "loss": 0.3491,
+      "step": 10750
+    },
+    {
+      "epoch": 0.7905728362379587,
+      "grad_norm": 0.5482704639434814,
+      "learning_rate": 2.21499429896878e-05,
+      "loss": 0.3673,
+      "step": 10751
+    },
+    {
+      "epoch": 0.7906463710566953,
+      "grad_norm": 0.44159454107284546,
+      "learning_rate": 2.2134997115989374e-05,
+      "loss": 0.3706,
+      "step": 10752
+    },
+    {
+      "epoch": 0.7907199058754321,
+      "grad_norm": 0.4768686890602112,
+      "learning_rate": 2.2120055658876093e-05,
+      "loss": 0.3748,
+      "step": 10753
+    },
+    {
+      "epoch": 0.7907934406941687,
+      "grad_norm": 0.43020251393318176,
+      "learning_rate": 2.2105118619195476e-05,
+      "loss": 0.3242,
+      "step": 10754
+    },
+    {
+      "epoch": 0.7908669755129054,
+      "grad_norm": 0.455568790435791,
+      "learning_rate": 2.209018599779472e-05,
+      "loss": 0.3599,
+      "step": 10755
+    },
+    {
+      "epoch": 0.790940510331642,
+      "grad_norm": 0.5528205633163452,
+      "learning_rate": 2.207525779552082e-05,
+      "loss": 0.3533,
+      "step": 10756
+    },
+    {
+      "epoch": 0.7910140451503787,
+      "grad_norm": 0.6108226180076599,
+      "learning_rate": 2.2060334013220564e-05,
+      "loss": 0.3755,
+      "step": 10757
+    },
+    {
+      "epoch": 0.7910875799691154,
+      "grad_norm": 0.44464942812919617,
+      "learning_rate": 2.2045414651740383e-05,
+      "loss": 0.3369,
+      "step": 10758
+    },
+    {
+      "epoch": 0.791161114787852,
+      "grad_norm": 0.4897739291191101,
+      "learning_rate": 2.2030499711926556e-05,
+      "loss": 0.3566,
+      "step": 10759
+    },
+    {
+      "epoch": 0.7912346496065887,
+      "grad_norm": 0.5176183581352234,
+      "learning_rate": 2.2015589194625064e-05,
+      "loss": 0.3607,
+      "step": 10760
+    },
+    {
+      "epoch": 0.7913081844253254,
+      "grad_norm": 0.406965047121048,
+      "learning_rate": 2.2000683100681685e-05,
+      "loss": 0.3604,
+      "step": 10761
+    },
+    {
+      "epoch": 0.7913817192440621,
+      "grad_norm": 0.5040404796600342,
+      "learning_rate": 2.1985781430941853e-05,
+      "loss": 0.3445,
+      "step": 10762
+    },
+    {
+      "epoch": 0.7914552540627987,
+      "grad_norm": 0.41719886660575867,
+      "learning_rate": 2.197088418625083e-05,
+      "loss": 0.3539,
+      "step": 10763
+    },
+    {
+      "epoch": 0.7915287888815354,
+      "grad_norm": 0.48041337728500366,
+      "learning_rate": 2.1955991367453623e-05,
+      "loss": 0.3674,
+      "step": 10764
+    },
+    {
+      "epoch": 0.7916023237002721,
+      "grad_norm": 0.4860018193721771,
+      "learning_rate": 2.1941102975394923e-05,
+      "loss": 0.3462,
+      "step": 10765
+    },
+    {
+      "epoch": 0.7916758585190088,
+      "grad_norm": 0.5481411218643188,
+      "learning_rate": 2.1926219010919236e-05,
+      "loss": 0.3503,
+      "step": 10766
+    },
+    {
+      "epoch": 0.7917493933377454,
+      "grad_norm": 0.43504613637924194,
+      "learning_rate": 2.1911339474870873e-05,
+      "loss": 0.3297,
+      "step": 10767
+    },
+    {
+      "epoch": 0.7918229281564821,
+      "grad_norm": 0.5144774317741394,
+      "learning_rate": 2.1896464368093672e-05,
+      "loss": 0.3927,
+      "step": 10768
+    },
+    {
+      "epoch": 0.7918964629752188,
+      "grad_norm": 0.41995304822921753,
+      "learning_rate": 2.1881593691431445e-05,
+      "loss": 0.3331,
+      "step": 10769
+    },
+    {
+      "epoch": 0.7919699977939554,
+      "grad_norm": 0.49739235639572144,
+      "learning_rate": 2.1866727445727696e-05,
+      "loss": 0.3367,
+      "step": 10770
+    },
+    {
+      "epoch": 0.7920435326126921,
+      "grad_norm": 0.5234074592590332,
+      "learning_rate": 2.1851865631825597e-05,
+      "loss": 0.3201,
+      "step": 10771
+    },
+    {
+      "epoch": 0.7921170674314287,
+      "grad_norm": 0.3766542673110962,
+      "learning_rate": 2.1837008250568148e-05,
+      "loss": 0.3263,
+      "step": 10772
+    },
+    {
+      "epoch": 0.7921906022501655,
+      "grad_norm": 0.42998138070106506,
+      "learning_rate": 2.1822155302798086e-05,
+      "loss": 0.3518,
+      "step": 10773
+    },
+    {
+      "epoch": 0.7922641370689021,
+      "grad_norm": 0.49912336468696594,
+      "learning_rate": 2.1807306789357895e-05,
+      "loss": 0.349,
+      "step": 10774
+    },
+    {
+      "epoch": 0.7923376718876388,
+      "grad_norm": 0.4731800854206085,
+      "learning_rate": 2.1792462711089768e-05,
+      "loss": 0.3684,
+      "step": 10775
+    },
+    {
+      "epoch": 0.7924112067063754,
+      "grad_norm": 0.5459535717964172,
+      "learning_rate": 2.1777623068835683e-05,
+      "loss": 0.3341,
+      "step": 10776
+    },
+    {
+      "epoch": 0.7924847415251122,
+      "grad_norm": 0.5040522813796997,
+      "learning_rate": 2.1762787863437396e-05,
+      "loss": 0.3626,
+      "step": 10777
+    },
+    {
+      "epoch": 0.7925582763438488,
+      "grad_norm": 0.5024466514587402,
+      "learning_rate": 2.1747957095736327e-05,
+      "loss": 0.3686,
+      "step": 10778
+    },
+    {
+      "epoch": 0.7926318111625855,
+      "grad_norm": 0.5288118720054626,
+      "learning_rate": 2.173313076657374e-05,
+      "loss": 0.3548,
+      "step": 10779
+    },
+    {
+      "epoch": 0.7927053459813221,
+      "grad_norm": 0.4898412227630615,
+      "learning_rate": 2.171830887679057e-05,
+      "loss": 0.3369,
+      "step": 10780
+    },
+    {
+      "epoch": 0.7927788808000589,
+      "grad_norm": 0.509179949760437,
+      "learning_rate": 2.1703491427227517e-05,
+      "loss": 0.3501,
+      "step": 10781
+    },
+    {
+      "epoch": 0.7928524156187955,
+      "grad_norm": 0.5358036756515503,
+      "learning_rate": 2.168867841872505e-05,
+      "loss": 0.3446,
+      "step": 10782
+    },
+    {
+      "epoch": 0.7929259504375322,
+      "grad_norm": 0.5126795172691345,
+      "learning_rate": 2.1673869852123396e-05,
+      "loss": 0.3535,
+      "step": 10783
+    },
+    {
+      "epoch": 0.7929994852562688,
+      "grad_norm": 0.43633612990379333,
+      "learning_rate": 2.1659065728262528e-05,
+      "loss": 0.3822,
+      "step": 10784
+    },
+    {
+      "epoch": 0.7930730200750055,
+      "grad_norm": 0.4468674063682556,
+      "learning_rate": 2.1644266047982098e-05,
+      "loss": 0.3905,
+      "step": 10785
+    },
+    {
+      "epoch": 0.7931465548937422,
+      "grad_norm": 0.4696578085422516,
+      "learning_rate": 2.1629470812121588e-05,
+      "loss": 0.3582,
+      "step": 10786
+    },
+    {
+      "epoch": 0.7932200897124788,
+      "grad_norm": 0.4404720067977905,
+      "learning_rate": 2.161468002152024e-05,
+      "loss": 0.3484,
+      "step": 10787
+    },
+    {
+      "epoch": 0.7932936245312155,
+      "grad_norm": 0.41348743438720703,
+      "learning_rate": 2.1599893677016925e-05,
+      "loss": 0.3476,
+      "step": 10788
+    },
+    {
+      "epoch": 0.7933671593499522,
+      "grad_norm": 0.40103912353515625,
+      "learning_rate": 2.1585111779450385e-05,
+      "loss": 0.3577,
+      "step": 10789
+    },
+    {
+      "epoch": 0.7934406941686889,
+      "grad_norm": 0.48954305052757263,
+      "learning_rate": 2.1570334329659093e-05,
+      "loss": 0.3738,
+      "step": 10790
+    },
+    {
+      "epoch": 0.7935142289874255,
+      "grad_norm": 0.4784267544746399,
+      "learning_rate": 2.15555613284812e-05,
+      "loss": 0.3619,
+      "step": 10791
+    },
+    {
+      "epoch": 0.7935877638061622,
+      "grad_norm": 0.4663011431694031,
+      "learning_rate": 2.1540792776754627e-05,
+      "loss": 0.3369,
+      "step": 10792
+    },
+    {
+      "epoch": 0.7936612986248989,
+      "grad_norm": 0.4678775668144226,
+      "learning_rate": 2.1526028675317123e-05,
+      "loss": 0.3682,
+      "step": 10793
+    },
+    {
+      "epoch": 0.7937348334436356,
+      "grad_norm": 0.4626762866973877,
+      "learning_rate": 2.1511269025006053e-05,
+      "loss": 0.352,
+      "step": 10794
+    },
+    {
+      "epoch": 0.7938083682623722,
+      "grad_norm": 0.5031585097312927,
+      "learning_rate": 2.1496513826658636e-05,
+      "loss": 0.3714,
+      "step": 10795
+    },
+    {
+      "epoch": 0.793881903081109,
+      "grad_norm": 0.5217744708061218,
+      "learning_rate": 2.1481763081111804e-05,
+      "loss": 0.3905,
+      "step": 10796
+    },
+    {
+      "epoch": 0.7939554378998456,
+      "grad_norm": 0.6557087898254395,
+      "learning_rate": 2.1467016789202264e-05,
+      "loss": 0.3774,
+      "step": 10797
+    },
+    {
+      "epoch": 0.7940289727185822,
+      "grad_norm": 0.41613227128982544,
+      "learning_rate": 2.1452274951766393e-05,
+      "loss": 0.346,
+      "step": 10798
+    },
+    {
+      "epoch": 0.7941025075373189,
+      "grad_norm": 0.46782737970352173,
+      "learning_rate": 2.1437537569640364e-05,
+      "loss": 0.3752,
+      "step": 10799
+    },
+    {
+      "epoch": 0.7941760423560555,
+      "grad_norm": 0.6465039253234863,
+      "learning_rate": 2.1422804643660156e-05,
+      "loss": 0.3311,
+      "step": 10800
+    },
+    {
+      "epoch": 0.7942495771747923,
+      "grad_norm": 0.4048122465610504,
+      "learning_rate": 2.1408076174661362e-05,
+      "loss": 0.3518,
+      "step": 10801
+    },
+    {
+      "epoch": 0.7943231119935289,
+      "grad_norm": 0.6633783578872681,
+      "learning_rate": 2.1393352163479452e-05,
+      "loss": 0.355,
+      "step": 10802
+    },
+    {
+      "epoch": 0.7943966468122656,
+      "grad_norm": 0.5392841696739197,
+      "learning_rate": 2.1378632610949567e-05,
+      "loss": 0.3524,
+      "step": 10803
+    },
+    {
+      "epoch": 0.7944701816310022,
+      "grad_norm": 0.877730667591095,
+      "learning_rate": 2.1363917517906583e-05,
+      "loss": 0.3368,
+      "step": 10804
+    },
+    {
+      "epoch": 0.794543716449739,
+      "grad_norm": 0.5043714046478271,
+      "learning_rate": 2.134920688518519e-05,
+      "loss": 0.3492,
+      "step": 10805
+    },
+    {
+      "epoch": 0.7946172512684756,
+      "grad_norm": 0.574603796005249,
+      "learning_rate": 2.1334500713619787e-05,
+      "loss": 0.3413,
+      "step": 10806
+    },
+    {
+      "epoch": 0.7946907860872123,
+      "grad_norm": 0.44500866532325745,
+      "learning_rate": 2.1319799004044538e-05,
+      "loss": 0.3646,
+      "step": 10807
+    },
+    {
+      "epoch": 0.794764320905949,
+      "grad_norm": 0.41401588916778564,
+      "learning_rate": 2.130510175729331e-05,
+      "loss": 0.3447,
+      "step": 10808
+    },
+    {
+      "epoch": 0.7948378557246857,
+      "grad_norm": 0.5507619380950928,
+      "learning_rate": 2.129040897419976e-05,
+      "loss": 0.3549,
+      "step": 10809
+    },
+    {
+      "epoch": 0.7949113905434223,
+      "grad_norm": 0.4341209828853607,
+      "learning_rate": 2.12757206555973e-05,
+      "loss": 0.3382,
+      "step": 10810
+    },
+    {
+      "epoch": 0.794984925362159,
+      "grad_norm": 0.5574464797973633,
+      "learning_rate": 2.1261036802319013e-05,
+      "loss": 0.3579,
+      "step": 10811
+    },
+    {
+      "epoch": 0.7950584601808957,
+      "grad_norm": 0.46902135014533997,
+      "learning_rate": 2.1246357415197827e-05,
+      "loss": 0.3744,
+      "step": 10812
+    },
+    {
+      "epoch": 0.7951319949996323,
+      "grad_norm": 0.43430522084236145,
+      "learning_rate": 2.1231682495066386e-05,
+      "loss": 0.3421,
+      "step": 10813
+    },
+    {
+      "epoch": 0.795205529818369,
+      "grad_norm": 0.40884140133857727,
+      "learning_rate": 2.1217012042757034e-05,
+      "loss": 0.3699,
+      "step": 10814
+    },
+    {
+      "epoch": 0.7952790646371056,
+      "grad_norm": 0.3931480646133423,
+      "learning_rate": 2.1202346059101874e-05,
+      "loss": 0.3457,
+      "step": 10815
+    },
+    {
+      "epoch": 0.7953525994558424,
+      "grad_norm": 0.407873272895813,
+      "learning_rate": 2.118768454493284e-05,
+      "loss": 0.3501,
+      "step": 10816
+    },
+    {
+      "epoch": 0.795426134274579,
+      "grad_norm": 0.4011022746562958,
+      "learning_rate": 2.117302750108148e-05,
+      "loss": 0.3631,
+      "step": 10817
+    },
+    {
+      "epoch": 0.7954996690933157,
+      "grad_norm": 0.36964112520217896,
+      "learning_rate": 2.115837492837919e-05,
+      "loss": 0.3336,
+      "step": 10818
+    },
+    {
+      "epoch": 0.7955732039120523,
+      "grad_norm": 0.5676648616790771,
+      "learning_rate": 2.1143726827657086e-05,
+      "loss": 0.3456,
+      "step": 10819
+    },
+    {
+      "epoch": 0.7956467387307891,
+      "grad_norm": 0.4463241398334503,
+      "learning_rate": 2.1129083199746037e-05,
+      "loss": 0.3364,
+      "step": 10820
+    },
+    {
+      "epoch": 0.7957202735495257,
+      "grad_norm": 0.7097681164741516,
+      "learning_rate": 2.111444404547659e-05,
+      "loss": 0.3287,
+      "step": 10821
+    },
+    {
+      "epoch": 0.7957938083682624,
+      "grad_norm": 0.4706067740917206,
+      "learning_rate": 2.1099809365679135e-05,
+      "loss": 0.3681,
+      "step": 10822
+    },
+    {
+      "epoch": 0.795867343186999,
+      "grad_norm": 0.5834869146347046,
+      "learning_rate": 2.108517916118379e-05,
+      "loss": 0.357,
+      "step": 10823
+    },
+    {
+      "epoch": 0.7959408780057358,
+      "grad_norm": 0.5249218940734863,
+      "learning_rate": 2.107055343282034e-05,
+      "loss": 0.3477,
+      "step": 10824
+    },
+    {
+      "epoch": 0.7960144128244724,
+      "grad_norm": 0.4673477113246918,
+      "learning_rate": 2.105593218141839e-05,
+      "loss": 0.3632,
+      "step": 10825
+    },
+    {
+      "epoch": 0.796087947643209,
+      "grad_norm": 0.4353119134902954,
+      "learning_rate": 2.104131540780734e-05,
+      "loss": 0.3502,
+      "step": 10826
+    },
+    {
+      "epoch": 0.7961614824619457,
+      "grad_norm": 0.5071302652359009,
+      "learning_rate": 2.1026703112816147e-05,
+      "loss": 0.3486,
+      "step": 10827
+    },
+    {
+      "epoch": 0.7962350172806824,
+      "grad_norm": 0.6832529306411743,
+      "learning_rate": 2.1012095297273705e-05,
+      "loss": 0.3725,
+      "step": 10828
+    },
+    {
+      "epoch": 0.7963085520994191,
+      "grad_norm": 0.6428468823432922,
+      "learning_rate": 2.0997491962008565e-05,
+      "loss": 0.3399,
+      "step": 10829
+    },
+    {
+      "epoch": 0.7963820869181557,
+      "grad_norm": 0.6495316028594971,
+      "learning_rate": 2.098289310784909e-05,
+      "loss": 0.331,
+      "step": 10830
+    },
+    {
+      "epoch": 0.7964556217368924,
+      "grad_norm": 0.8337749242782593,
+      "learning_rate": 2.096829873562328e-05,
+      "loss": 0.3565,
+      "step": 10831
+    },
+    {
+      "epoch": 0.7965291565556291,
+      "grad_norm": 0.48656418919563293,
+      "learning_rate": 2.095370884615897e-05,
+      "loss": 0.3539,
+      "step": 10832
+    },
+    {
+      "epoch": 0.7966026913743658,
+      "grad_norm": 0.7225539088249207,
+      "learning_rate": 2.093912344028375e-05,
+      "loss": 0.3517,
+      "step": 10833
+    },
+    {
+      "epoch": 0.7966762261931024,
+      "grad_norm": 0.5531941652297974,
+      "learning_rate": 2.0924542518824852e-05,
+      "loss": 0.3732,
+      "step": 10834
+    },
+    {
+      "epoch": 0.7967497610118391,
+      "grad_norm": 0.660987913608551,
+      "learning_rate": 2.0909966082609356e-05,
+      "loss": 0.3732,
+      "step": 10835
+    },
+    {
+      "epoch": 0.7968232958305758,
+      "grad_norm": 0.5453696846961975,
+      "learning_rate": 2.0895394132464074e-05,
+      "loss": 0.3518,
+      "step": 10836
+    },
+    {
+      "epoch": 0.7968968306493125,
+      "grad_norm": 0.43511661887168884,
+      "learning_rate": 2.0880826669215527e-05,
+      "loss": 0.3561,
+      "step": 10837
+    },
+    {
+      "epoch": 0.7969703654680491,
+      "grad_norm": 0.4939083456993103,
+      "learning_rate": 2.0866263693689947e-05,
+      "loss": 0.3555,
+      "step": 10838
+    },
+    {
+      "epoch": 0.7970439002867858,
+      "grad_norm": 0.5375283360481262,
+      "learning_rate": 2.0851705206713447e-05,
+      "loss": 0.349,
+      "step": 10839
+    },
+    {
+      "epoch": 0.7971174351055225,
+      "grad_norm": 0.5203908085823059,
+      "learning_rate": 2.0837151209111717e-05,
+      "loss": 0.3719,
+      "step": 10840
+    },
+    {
+      "epoch": 0.7971909699242591,
+      "grad_norm": 0.4507681131362915,
+      "learning_rate": 2.082260170171032e-05,
+      "loss": 0.3645,
+      "step": 10841
+    },
+    {
+      "epoch": 0.7972645047429958,
+      "grad_norm": 0.6112905144691467,
+      "learning_rate": 2.080805668533452e-05,
+      "loss": 0.3441,
+      "step": 10842
+    },
+    {
+      "epoch": 0.7973380395617324,
+      "grad_norm": 0.5238696336746216,
+      "learning_rate": 2.0793516160809334e-05,
+      "loss": 0.3443,
+      "step": 10843
+    },
+    {
+      "epoch": 0.7974115743804692,
+      "grad_norm": 0.7093983292579651,
+      "learning_rate": 2.077898012895948e-05,
+      "loss": 0.363,
+      "step": 10844
+    },
+    {
+      "epoch": 0.7974851091992058,
+      "grad_norm": 0.48335742950439453,
+      "learning_rate": 2.0764448590609476e-05,
+      "loss": 0.3254,
+      "step": 10845
+    },
+    {
+      "epoch": 0.7975586440179425,
+      "grad_norm": 0.4154163897037506,
+      "learning_rate": 2.0749921546583595e-05,
+      "loss": 0.3355,
+      "step": 10846
+    },
+    {
+      "epoch": 0.7976321788366791,
+      "grad_norm": 0.6152236461639404,
+      "learning_rate": 2.073539899770578e-05,
+      "loss": 0.3439,
+      "step": 10847
+    },
+    {
+      "epoch": 0.7977057136554159,
+      "grad_norm": 0.42453157901763916,
+      "learning_rate": 2.0720880944799782e-05,
+      "loss": 0.3617,
+      "step": 10848
+    },
+    {
+      "epoch": 0.7977792484741525,
+      "grad_norm": 0.6702137589454651,
+      "learning_rate": 2.070636738868914e-05,
+      "loss": 0.3688,
+      "step": 10849
+    },
+    {
+      "epoch": 0.7978527832928892,
+      "grad_norm": 0.5599128603935242,
+      "learning_rate": 2.0691858330196956e-05,
+      "loss": 0.3575,
+      "step": 10850
+    },
+    {
+      "epoch": 0.7979263181116258,
+      "grad_norm": 0.4546409249305725,
+      "learning_rate": 2.067735377014628e-05,
+      "loss": 0.346,
+      "step": 10851
+    },
+    {
+      "epoch": 0.7979998529303626,
+      "grad_norm": 0.622424840927124,
+      "learning_rate": 2.0662853709359796e-05,
+      "loss": 0.369,
+      "step": 10852
+    },
+    {
+      "epoch": 0.7980733877490992,
+      "grad_norm": 0.5847616791725159,
+      "learning_rate": 2.0648358148660017e-05,
+      "loss": 0.3513,
+      "step": 10853
+    },
+    {
+      "epoch": 0.7981469225678358,
+      "grad_norm": 0.5710198283195496,
+      "learning_rate": 2.0633867088869073e-05,
+      "loss": 0.35,
+      "step": 10854
+    },
+    {
+      "epoch": 0.7982204573865725,
+      "grad_norm": 0.8056076169013977,
+      "learning_rate": 2.061938053080894e-05,
+      "loss": 0.3434,
+      "step": 10855
+    },
+    {
+      "epoch": 0.7982939922053092,
+      "grad_norm": 0.42400720715522766,
+      "learning_rate": 2.060489847530136e-05,
+      "loss": 0.363,
+      "step": 10856
+    },
+    {
+      "epoch": 0.7983675270240459,
+      "grad_norm": 0.8351950645446777,
+      "learning_rate": 2.0590420923167687e-05,
+      "loss": 0.3684,
+      "step": 10857
+    },
+    {
+      "epoch": 0.7984410618427825,
+      "grad_norm": 0.4312412142753601,
+      "learning_rate": 2.0575947875229152e-05,
+      "loss": 0.343,
+      "step": 10858
+    },
+    {
+      "epoch": 0.7985145966615192,
+      "grad_norm": 0.5448611378669739,
+      "learning_rate": 2.0561479332306697e-05,
+      "loss": 0.3542,
+      "step": 10859
+    },
+    {
+      "epoch": 0.7985881314802559,
+      "grad_norm": 0.4074912965297699,
+      "learning_rate": 2.054701529522094e-05,
+      "loss": 0.3525,
+      "step": 10860
+    },
+    {
+      "epoch": 0.7986616662989926,
+      "grad_norm": 0.42238250374794006,
+      "learning_rate": 2.053255576479235e-05,
+      "loss": 0.3763,
+      "step": 10861
+    },
+    {
+      "epoch": 0.7987352011177292,
+      "grad_norm": 0.4764612317085266,
+      "learning_rate": 2.0518100741841073e-05,
+      "loss": 0.3589,
+      "step": 10862
+    },
+    {
+      "epoch": 0.798808735936466,
+      "grad_norm": 0.5535637736320496,
+      "learning_rate": 2.0503650227186967e-05,
+      "loss": 0.3506,
+      "step": 10863
+    },
+    {
+      "epoch": 0.7988822707552026,
+      "grad_norm": 0.43044835329055786,
+      "learning_rate": 2.0489204221649716e-05,
+      "loss": 0.3834,
+      "step": 10864
+    },
+    {
+      "epoch": 0.7989558055739393,
+      "grad_norm": 0.43789342045783997,
+      "learning_rate": 2.0474762726048713e-05,
+      "loss": 0.3479,
+      "step": 10865
+    },
+    {
+      "epoch": 0.7990293403926759,
+      "grad_norm": 0.4516502618789673,
+      "learning_rate": 2.0460325741203122e-05,
+      "loss": 0.3631,
+      "step": 10866
+    },
+    {
+      "epoch": 0.7991028752114127,
+      "grad_norm": 0.4062371850013733,
+      "learning_rate": 2.044589326793177e-05,
+      "loss": 0.3382,
+      "step": 10867
+    },
+    {
+      "epoch": 0.7991764100301493,
+      "grad_norm": 0.5089420676231384,
+      "learning_rate": 2.043146530705331e-05,
+      "loss": 0.3317,
+      "step": 10868
+    },
+    {
+      "epoch": 0.7992499448488859,
+      "grad_norm": 0.4246264100074768,
+      "learning_rate": 2.0417041859386144e-05,
+      "loss": 0.3664,
+      "step": 10869
+    },
+    {
+      "epoch": 0.7993234796676226,
+      "grad_norm": 0.558800995349884,
+      "learning_rate": 2.0402622925748305e-05,
+      "loss": 0.3362,
+      "step": 10870
+    },
+    {
+      "epoch": 0.7993970144863592,
+      "grad_norm": 0.511026918888092,
+      "learning_rate": 2.0388208506957707e-05,
+      "loss": 0.3574,
+      "step": 10871
+    },
+    {
+      "epoch": 0.799470549305096,
+      "grad_norm": 0.5230469703674316,
+      "learning_rate": 2.0373798603831996e-05,
+      "loss": 0.366,
+      "step": 10872
+    },
+    {
+      "epoch": 0.7995440841238326,
+      "grad_norm": 0.6382681727409363,
+      "learning_rate": 2.0359393217188404e-05,
+      "loss": 0.3588,
+      "step": 10873
+    },
+    {
+      "epoch": 0.7996176189425693,
+      "grad_norm": 0.4676513373851776,
+      "learning_rate": 2.0344992347844082e-05,
+      "loss": 0.3506,
+      "step": 10874
+    },
+    {
+      "epoch": 0.799691153761306,
+      "grad_norm": 0.579765260219574,
+      "learning_rate": 2.0330595996615853e-05,
+      "loss": 0.3538,
+      "step": 10875
+    },
+    {
+      "epoch": 0.7997646885800427,
+      "grad_norm": 0.46723949909210205,
+      "learning_rate": 2.031620416432033e-05,
+      "loss": 0.3354,
+      "step": 10876
+    },
+    {
+      "epoch": 0.7998382233987793,
+      "grad_norm": 0.5441309809684753,
+      "learning_rate": 2.030181685177376e-05,
+      "loss": 0.3913,
+      "step": 10877
+    },
+    {
+      "epoch": 0.799911758217516,
+      "grad_norm": 0.4317614436149597,
+      "learning_rate": 2.0287434059792254e-05,
+      "loss": 0.3311,
+      "step": 10878
+    },
+    {
+      "epoch": 0.7999852930362527,
+      "grad_norm": 0.5575008988380432,
+      "learning_rate": 2.0273055789191643e-05,
+      "loss": 0.3497,
+      "step": 10879
+    },
+    {
+      "epoch": 0.8000588278549894,
+      "grad_norm": 0.4615150988101959,
+      "learning_rate": 2.0258682040787425e-05,
+      "loss": 0.3398,
+      "step": 10880
+    },
+    {
+      "epoch": 0.800132362673726,
+      "grad_norm": 0.5521947145462036,
+      "learning_rate": 2.0244312815394906e-05,
+      "loss": 0.3306,
+      "step": 10881
+    },
+    {
+      "epoch": 0.8002058974924627,
+      "grad_norm": 0.47826722264289856,
+      "learning_rate": 2.0229948113829155e-05,
+      "loss": 0.3357,
+      "step": 10882
+    },
+    {
+      "epoch": 0.8002794323111994,
+      "grad_norm": 0.5014266967773438,
+      "learning_rate": 2.0215587936904913e-05,
+      "loss": 0.3413,
+      "step": 10883
+    },
+    {
+      "epoch": 0.800352967129936,
+      "grad_norm": 0.4684417247772217,
+      "learning_rate": 2.0201232285436754e-05,
+      "loss": 0.3291,
+      "step": 10884
+    },
+    {
+      "epoch": 0.8004265019486727,
+      "grad_norm": 0.4146701991558075,
+      "learning_rate": 2.0186881160238914e-05,
+      "loss": 0.3507,
+      "step": 10885
+    },
+    {
+      "epoch": 0.8005000367674093,
+      "grad_norm": 0.5923832058906555,
+      "learning_rate": 2.0172534562125365e-05,
+      "loss": 0.3193,
+      "step": 10886
+    },
+    {
+      "epoch": 0.8005735715861461,
+      "grad_norm": 0.40932294726371765,
+      "learning_rate": 2.015819249190991e-05,
+      "loss": 0.3367,
+      "step": 10887
+    },
+    {
+      "epoch": 0.8006471064048827,
+      "grad_norm": 0.5283913612365723,
+      "learning_rate": 2.0143854950406037e-05,
+      "loss": 0.3484,
+      "step": 10888
+    },
+    {
+      "epoch": 0.8007206412236194,
+      "grad_norm": 0.44643381237983704,
+      "learning_rate": 2.012952193842702e-05,
+      "loss": 0.3324,
+      "step": 10889
+    },
+    {
+      "epoch": 0.800794176042356,
+      "grad_norm": 0.4640791416168213,
+      "learning_rate": 2.0115193456785774e-05,
+      "loss": 0.332,
+      "step": 10890
+    },
+    {
+      "epoch": 0.8008677108610928,
+      "grad_norm": 0.4750983715057373,
+      "learning_rate": 2.010086950629506e-05,
+      "loss": 0.3427,
+      "step": 10891
+    },
+    {
+      "epoch": 0.8009412456798294,
+      "grad_norm": 0.6511127352714539,
+      "learning_rate": 2.0086550087767387e-05,
+      "loss": 0.3493,
+      "step": 10892
+    },
+    {
+      "epoch": 0.8010147804985661,
+      "grad_norm": 0.4364197850227356,
+      "learning_rate": 2.0072235202014888e-05,
+      "loss": 0.335,
+      "step": 10893
+    },
+    {
+      "epoch": 0.8010883153173027,
+      "grad_norm": 0.4158484935760498,
+      "learning_rate": 2.0057924849849564e-05,
+      "loss": 0.3399,
+      "step": 10894
+    },
+    {
+      "epoch": 0.8011618501360395,
+      "grad_norm": 0.5283784866333008,
+      "learning_rate": 2.004361903208314e-05,
+      "loss": 0.3535,
+      "step": 10895
+    },
+    {
+      "epoch": 0.8012353849547761,
+      "grad_norm": 0.4309467375278473,
+      "learning_rate": 2.0029317749527022e-05,
+      "loss": 0.342,
+      "step": 10896
+    },
+    {
+      "epoch": 0.8013089197735127,
+      "grad_norm": 0.42502155900001526,
+      "learning_rate": 2.0015021002992372e-05,
+      "loss": 0.3464,
+      "step": 10897
+    },
+    {
+      "epoch": 0.8013824545922494,
+      "grad_norm": 0.5712305307388306,
+      "learning_rate": 2.000072879329017e-05,
+      "loss": 0.3548,
+      "step": 10898
+    },
+    {
+      "epoch": 0.8014559894109861,
+      "grad_norm": 0.4334051311016083,
+      "learning_rate": 1.9986441121231026e-05,
+      "loss": 0.3204,
+      "step": 10899
+    },
+    {
+      "epoch": 0.8015295242297228,
+      "grad_norm": 0.3720455765724182,
+      "learning_rate": 1.9972157987625394e-05,
+      "loss": 0.3355,
+      "step": 10900
+    },
+    {
+      "epoch": 0.8016030590484594,
+      "grad_norm": 0.4397164285182953,
+      "learning_rate": 1.9957879393283408e-05,
+      "loss": 0.3409,
+      "step": 10901
+    },
+    {
+      "epoch": 0.8016765938671961,
+      "grad_norm": 0.47064533829689026,
+      "learning_rate": 1.9943605339015003e-05,
+      "loss": 0.3151,
+      "step": 10902
+    },
+    {
+      "epoch": 0.8017501286859328,
+      "grad_norm": 0.461748868227005,
+      "learning_rate": 1.9929335825629758e-05,
+      "loss": 0.3311,
+      "step": 10903
+    },
+    {
+      "epoch": 0.8018236635046695,
+      "grad_norm": 0.4765840768814087,
+      "learning_rate": 1.9915070853937102e-05,
+      "loss": 0.3311,
+      "step": 10904
+    },
+    {
+      "epoch": 0.8018971983234061,
+      "grad_norm": 0.40356311202049255,
+      "learning_rate": 1.9900810424746154e-05,
+      "loss": 0.3406,
+      "step": 10905
+    },
+    {
+      "epoch": 0.8019707331421428,
+      "grad_norm": 0.5461817383766174,
+      "learning_rate": 1.9886554538865754e-05,
+      "loss": 0.3155,
+      "step": 10906
+    },
+    {
+      "epoch": 0.8020442679608795,
+      "grad_norm": 0.727736234664917,
+      "learning_rate": 1.9872303197104548e-05,
+      "loss": 0.3414,
+      "step": 10907
+    },
+    {
+      "epoch": 0.8021178027796162,
+      "grad_norm": 0.518928587436676,
+      "learning_rate": 1.9858056400270876e-05,
+      "loss": 0.3461,
+      "step": 10908
+    },
+    {
+      "epoch": 0.8021913375983528,
+      "grad_norm": 0.7223975658416748,
+      "learning_rate": 1.984381414917279e-05,
+      "loss": 0.3274,
+      "step": 10909
+    },
+    {
+      "epoch": 0.8022648724170895,
+      "grad_norm": 0.4334070086479187,
+      "learning_rate": 1.982957644461816e-05,
+      "loss": 0.3464,
+      "step": 10910
+    },
+    {
+      "epoch": 0.8023384072358262,
+      "grad_norm": 0.5305948257446289,
+      "learning_rate": 1.9815343287414556e-05,
+      "loss": 0.3152,
+      "step": 10911
+    },
+    {
+      "epoch": 0.8024119420545628,
+      "grad_norm": 0.5527846813201904,
+      "learning_rate": 1.9801114678369335e-05,
+      "loss": 0.3371,
+      "step": 10912
+    },
+    {
+      "epoch": 0.8024854768732995,
+      "grad_norm": 0.4818758964538574,
+      "learning_rate": 1.97868906182895e-05,
+      "loss": 0.346,
+      "step": 10913
+    },
+    {
+      "epoch": 0.8025590116920361,
+      "grad_norm": 0.4886718988418579,
+      "learning_rate": 1.9772671107981876e-05,
+      "loss": 0.3215,
+      "step": 10914
+    },
+    {
+      "epoch": 0.8026325465107729,
+      "grad_norm": 0.44308575987815857,
+      "learning_rate": 1.9758456148253046e-05,
+      "loss": 0.3408,
+      "step": 10915
+    },
+    {
+      "epoch": 0.8027060813295095,
+      "grad_norm": 0.6115341782569885,
+      "learning_rate": 1.974424573990923e-05,
+      "loss": 0.3396,
+      "step": 10916
+    },
+    {
+      "epoch": 0.8027796161482462,
+      "grad_norm": 0.5149179100990295,
+      "learning_rate": 1.97300398837565e-05,
+      "loss": 0.3455,
+      "step": 10917
+    },
+    {
+      "epoch": 0.8028531509669828,
+      "grad_norm": 0.48370596766471863,
+      "learning_rate": 1.9715838580600654e-05,
+      "loss": 0.3397,
+      "step": 10918
+    },
+    {
+      "epoch": 0.8029266857857196,
+      "grad_norm": 0.5771750807762146,
+      "learning_rate": 1.9701641831247164e-05,
+      "loss": 0.325,
+      "step": 10919
+    },
+    {
+      "epoch": 0.8030002206044562,
+      "grad_norm": 0.5844594240188599,
+      "learning_rate": 1.9687449636501253e-05,
+      "loss": 0.3554,
+      "step": 10920
+    },
+    {
+      "epoch": 0.8030737554231929,
+      "grad_norm": 0.5265563726425171,
+      "learning_rate": 1.9673261997168002e-05,
+      "loss": 0.3237,
+      "step": 10921
+    },
+    {
+      "epoch": 0.8031472902419295,
+      "grad_norm": 0.8114467263221741,
+      "learning_rate": 1.9659078914052064e-05,
+      "loss": 0.3245,
+      "step": 10922
+    },
+    {
+      "epoch": 0.8032208250606663,
+      "grad_norm": 0.432705283164978,
+      "learning_rate": 1.9644900387957965e-05,
+      "loss": 0.3128,
+      "step": 10923
+    },
+    {
+      "epoch": 0.8032943598794029,
+      "grad_norm": 0.8459081053733826,
+      "learning_rate": 1.9630726419689917e-05,
+      "loss": 0.3361,
+      "step": 10924
+    },
+    {
+      "epoch": 0.8033678946981395,
+      "grad_norm": 0.44931939244270325,
+      "learning_rate": 1.9616557010051906e-05,
+      "loss": 0.3348,
+      "step": 10925
+    },
+    {
+      "epoch": 0.8034414295168762,
+      "grad_norm": 0.3795832395553589,
+      "learning_rate": 1.960239215984758e-05,
+      "loss": 0.3455,
+      "step": 10926
+    },
+    {
+      "epoch": 0.8035149643356129,
+      "grad_norm": 0.6084975600242615,
+      "learning_rate": 1.9588231869880426e-05,
+      "loss": 0.3394,
+      "step": 10927
+    },
+    {
+      "epoch": 0.8035884991543496,
+      "grad_norm": 0.46658581495285034,
+      "learning_rate": 1.9574076140953656e-05,
+      "loss": 0.3526,
+      "step": 10928
+    },
+    {
+      "epoch": 0.8036620339730862,
+      "grad_norm": 0.6149799823760986,
+      "learning_rate": 1.9559924973870125e-05,
+      "loss": 0.3131,
+      "step": 10929
+    },
+    {
+      "epoch": 0.803735568791823,
+      "grad_norm": 0.5155419707298279,
+      "learning_rate": 1.9545778369432544e-05,
+      "loss": 0.3543,
+      "step": 10930
+    },
+    {
+      "epoch": 0.8038091036105596,
+      "grad_norm": 0.6638898253440857,
+      "learning_rate": 1.953163632844338e-05,
+      "loss": 0.3492,
+      "step": 10931
+    },
+    {
+      "epoch": 0.8038826384292963,
+      "grad_norm": 0.3761219382286072,
+      "learning_rate": 1.9517498851704665e-05,
+      "loss": 0.3447,
+      "step": 10932
+    },
+    {
+      "epoch": 0.8039561732480329,
+      "grad_norm": 0.5690950155258179,
+      "learning_rate": 1.950336594001836e-05,
+      "loss": 0.3522,
+      "step": 10933
+    },
+    {
+      "epoch": 0.8040297080667697,
+      "grad_norm": 0.5329024195671082,
+      "learning_rate": 1.948923759418608e-05,
+      "loss": 0.3349,
+      "step": 10934
+    },
+    {
+      "epoch": 0.8041032428855063,
+      "grad_norm": 0.49591439962387085,
+      "learning_rate": 1.947511381500924e-05,
+      "loss": 0.3474,
+      "step": 10935
+    },
+    {
+      "epoch": 0.804176777704243,
+      "grad_norm": 0.4026503264904022,
+      "learning_rate": 1.9460994603288906e-05,
+      "loss": 0.343,
+      "step": 10936
+    },
+    {
+      "epoch": 0.8042503125229796,
+      "grad_norm": 0.4858540892601013,
+      "learning_rate": 1.9446879959825946e-05,
+      "loss": 0.321,
+      "step": 10937
+    },
+    {
+      "epoch": 0.8043238473417164,
+      "grad_norm": 0.4066292941570282,
+      "learning_rate": 1.9432769885421e-05,
+      "loss": 0.3321,
+      "step": 10938
+    },
+    {
+      "epoch": 0.804397382160453,
+      "grad_norm": 0.41638341546058655,
+      "learning_rate": 1.941866438087434e-05,
+      "loss": 0.3198,
+      "step": 10939
+    },
+    {
+      "epoch": 0.8044709169791896,
+      "grad_norm": 0.4801656901836395,
+      "learning_rate": 1.940456344698608e-05,
+      "loss": 0.31,
+      "step": 10940
+    },
+    {
+      "epoch": 0.8045444517979263,
+      "grad_norm": 0.5162137746810913,
+      "learning_rate": 1.9390467084556053e-05,
+      "loss": 0.3376,
+      "step": 10941
+    },
+    {
+      "epoch": 0.804617986616663,
+      "grad_norm": 0.4318986237049103,
+      "learning_rate": 1.937637529438382e-05,
+      "loss": 0.3464,
+      "step": 10942
+    },
+    {
+      "epoch": 0.8046915214353997,
+      "grad_norm": 0.42976412177085876,
+      "learning_rate": 1.9362288077268632e-05,
+      "loss": 0.3208,
+      "step": 10943
+    },
+    {
+      "epoch": 0.8047650562541363,
+      "grad_norm": 0.41965100169181824,
+      "learning_rate": 1.934820543400959e-05,
+      "loss": 0.3834,
+      "step": 10944
+    },
+    {
+      "epoch": 0.804838591072873,
+      "grad_norm": 0.497853547334671,
+      "learning_rate": 1.9334127365405418e-05,
+      "loss": 0.3112,
+      "step": 10945
+    },
+    {
+      "epoch": 0.8049121258916097,
+      "grad_norm": 0.4713118374347687,
+      "learning_rate": 1.9320053872254684e-05,
+      "loss": 0.3143,
+      "step": 10946
+    },
+    {
+      "epoch": 0.8049856607103464,
+      "grad_norm": 0.36717909574508667,
+      "learning_rate": 1.9305984955355638e-05,
+      "loss": 0.3831,
+      "step": 10947
+    },
+    {
+      "epoch": 0.805059195529083,
+      "grad_norm": 0.37926188111305237,
+      "learning_rate": 1.92919206155063e-05,
+      "loss": 0.3151,
+      "step": 10948
+    },
+    {
+      "epoch": 0.8051327303478197,
+      "grad_norm": 0.36992645263671875,
+      "learning_rate": 1.9277860853504382e-05,
+      "loss": 0.3218,
+      "step": 10949
+    },
+    {
+      "epoch": 0.8052062651665564,
+      "grad_norm": 0.4896516799926758,
+      "learning_rate": 1.9263805670147383e-05,
+      "loss": 0.3227,
+      "step": 10950
+    },
+    {
+      "epoch": 0.8052797999852931,
+      "grad_norm": 0.41516584157943726,
+      "learning_rate": 1.9249755066232566e-05,
+      "loss": 0.3508,
+      "step": 10951
+    },
+    {
+      "epoch": 0.8053533348040297,
+      "grad_norm": 0.41349515318870544,
+      "learning_rate": 1.9235709042556827e-05,
+      "loss": 0.3358,
+      "step": 10952
+    },
+    {
+      "epoch": 0.8054268696227663,
+      "grad_norm": 0.44470712542533875,
+      "learning_rate": 1.9221667599916905e-05,
+      "loss": 0.3534,
+      "step": 10953
+    },
+    {
+      "epoch": 0.8055004044415031,
+      "grad_norm": 0.44769707322120667,
+      "learning_rate": 1.9207630739109296e-05,
+      "loss": 0.3385,
+      "step": 10954
+    },
+    {
+      "epoch": 0.8055739392602397,
+      "grad_norm": 0.5376906394958496,
+      "learning_rate": 1.9193598460930085e-05,
+      "loss": 0.3454,
+      "step": 10955
+    },
+    {
+      "epoch": 0.8056474740789764,
+      "grad_norm": 0.48433709144592285,
+      "learning_rate": 1.9179570766175246e-05,
+      "loss": 0.3405,
+      "step": 10956
+    },
+    {
+      "epoch": 0.805721008897713,
+      "grad_norm": 0.4141180217266083,
+      "learning_rate": 1.916554765564045e-05,
+      "loss": 0.3281,
+      "step": 10957
+    },
+    {
+      "epoch": 0.8057945437164498,
+      "grad_norm": 0.7254524827003479,
+      "learning_rate": 1.9151529130121127e-05,
+      "loss": 0.3337,
+      "step": 10958
+    },
+    {
+      "epoch": 0.8058680785351864,
+      "grad_norm": 0.4328545928001404,
+      "learning_rate": 1.9137515190412358e-05,
+      "loss": 0.3331,
+      "step": 10959
+    },
+    {
+      "epoch": 0.8059416133539231,
+      "grad_norm": 0.6690646409988403,
+      "learning_rate": 1.9123505837309075e-05,
+      "loss": 0.348,
+      "step": 10960
+    },
+    {
+      "epoch": 0.8060151481726597,
+      "grad_norm": 0.4203026592731476,
+      "learning_rate": 1.910950107160592e-05,
+      "loss": 0.3266,
+      "step": 10961
+    },
+    {
+      "epoch": 0.8060886829913965,
+      "grad_norm": 0.808388888835907,
+      "learning_rate": 1.9095500894097195e-05,
+      "loss": 0.3535,
+      "step": 10962
+    },
+    {
+      "epoch": 0.8061622178101331,
+      "grad_norm": 0.4638335704803467,
+      "learning_rate": 1.908150530557704e-05,
+      "loss": 0.317,
+      "step": 10963
+    },
+    {
+      "epoch": 0.8062357526288698,
+      "grad_norm": 0.5674444437026978,
+      "learning_rate": 1.906751430683934e-05,
+      "loss": 0.3219,
+      "step": 10964
+    },
+    {
+      "epoch": 0.8063092874476064,
+      "grad_norm": 0.5202406048774719,
+      "learning_rate": 1.9053527898677603e-05,
+      "loss": 0.3348,
+      "step": 10965
+    },
+    {
+      "epoch": 0.8063828222663432,
+      "grad_norm": 0.4369564652442932,
+      "learning_rate": 1.903954608188523e-05,
+      "loss": 0.3525,
+      "step": 10966
+    },
+    {
+      "epoch": 0.8064563570850798,
+      "grad_norm": 0.44035983085632324,
+      "learning_rate": 1.9025568857255228e-05,
+      "loss": 0.3406,
+      "step": 10967
+    },
+    {
+      "epoch": 0.8065298919038164,
+      "grad_norm": 0.4577168822288513,
+      "learning_rate": 1.9011596225580397e-05,
+      "loss": 0.3278,
+      "step": 10968
+    },
+    {
+      "epoch": 0.8066034267225531,
+      "grad_norm": 0.42473283410072327,
+      "learning_rate": 1.8997628187653305e-05,
+      "loss": 0.3218,
+      "step": 10969
+    },
+    {
+      "epoch": 0.8066769615412898,
+      "grad_norm": 0.5735015273094177,
+      "learning_rate": 1.8983664744266215e-05,
+      "loss": 0.3391,
+      "step": 10970
+    },
+    {
+      "epoch": 0.8067504963600265,
+      "grad_norm": 0.5525845289230347,
+      "learning_rate": 1.89697058962112e-05,
+      "loss": 0.3169,
+      "step": 10971
+    },
+    {
+      "epoch": 0.8068240311787631,
+      "grad_norm": 0.3774339258670807,
+      "learning_rate": 1.895575164427994e-05,
+      "loss": 0.3063,
+      "step": 10972
+    },
+    {
+      "epoch": 0.8068975659974998,
+      "grad_norm": 0.5126948356628418,
+      "learning_rate": 1.8941801989263984e-05,
+      "loss": 0.3466,
+      "step": 10973
+    },
+    {
+      "epoch": 0.8069711008162365,
+      "grad_norm": 0.474386990070343,
+      "learning_rate": 1.8927856931954592e-05,
+      "loss": 0.3382,
+      "step": 10974
+    },
+    {
+      "epoch": 0.8070446356349732,
+      "grad_norm": 0.45995262265205383,
+      "learning_rate": 1.891391647314268e-05,
+      "loss": 0.31,
+      "step": 10975
+    },
+    {
+      "epoch": 0.8071181704537098,
+      "grad_norm": 0.5102444887161255,
+      "learning_rate": 1.8899980613619006e-05,
+      "loss": 0.3588,
+      "step": 10976
+    },
+    {
+      "epoch": 0.8071917052724465,
+      "grad_norm": 0.4607975482940674,
+      "learning_rate": 1.8886049354174073e-05,
+      "loss": 0.3422,
+      "step": 10977
+    },
+    {
+      "epoch": 0.8072652400911832,
+      "grad_norm": 0.4243262708187103,
+      "learning_rate": 1.887212269559797e-05,
+      "loss": 0.3141,
+      "step": 10978
+    },
+    {
+      "epoch": 0.8073387749099199,
+      "grad_norm": 0.593819797039032,
+      "learning_rate": 1.885820063868068e-05,
+      "loss": 0.3177,
+      "step": 10979
+    },
+    {
+      "epoch": 0.8074123097286565,
+      "grad_norm": 0.3828634023666382,
+      "learning_rate": 1.8844283184211887e-05,
+      "loss": 0.3146,
+      "step": 10980
+    },
+    {
+      "epoch": 0.8074858445473931,
+      "grad_norm": 0.44546839594841003,
+      "learning_rate": 1.8830370332981028e-05,
+      "loss": 0.3252,
+      "step": 10981
+    },
+    {
+      "epoch": 0.8075593793661299,
+      "grad_norm": 0.4659217894077301,
+      "learning_rate": 1.8816462085777188e-05,
+      "loss": 0.3283,
+      "step": 10982
+    },
+    {
+      "epoch": 0.8076329141848665,
+      "grad_norm": 0.4231095016002655,
+      "learning_rate": 1.8802558443389307e-05,
+      "loss": 0.3195,
+      "step": 10983
+    },
+    {
+      "epoch": 0.8077064490036032,
+      "grad_norm": 0.43271490931510925,
+      "learning_rate": 1.8788659406606023e-05,
+      "loss": 0.3486,
+      "step": 10984
+    },
+    {
+      "epoch": 0.8077799838223398,
+      "grad_norm": 0.6037747263908386,
+      "learning_rate": 1.8774764976215653e-05,
+      "loss": 0.3269,
+      "step": 10985
+    },
+    {
+      "epoch": 0.8078535186410766,
+      "grad_norm": 0.4363400936126709,
+      "learning_rate": 1.876087515300634e-05,
+      "loss": 0.3586,
+      "step": 10986
+    },
+    {
+      "epoch": 0.8079270534598132,
+      "grad_norm": 0.4194544553756714,
+      "learning_rate": 1.8746989937765946e-05,
+      "loss": 0.3398,
+      "step": 10987
+    },
+    {
+      "epoch": 0.8080005882785499,
+      "grad_norm": 0.36093461513519287,
+      "learning_rate": 1.8733109331282016e-05,
+      "loss": 0.3513,
+      "step": 10988
+    },
+    {
+      "epoch": 0.8080741230972865,
+      "grad_norm": 0.4603554308414459,
+      "learning_rate": 1.87192333343419e-05,
+      "loss": 0.3236,
+      "step": 10989
+    },
+    {
+      "epoch": 0.8081476579160233,
+      "grad_norm": 0.3937876224517822,
+      "learning_rate": 1.8705361947732658e-05,
+      "loss": 0.3253,
+      "step": 10990
+    },
+    {
+      "epoch": 0.8082211927347599,
+      "grad_norm": 0.4551372230052948,
+      "learning_rate": 1.8691495172241048e-05,
+      "loss": 0.3361,
+      "step": 10991
+    },
+    {
+      "epoch": 0.8082947275534966,
+      "grad_norm": 0.43749892711639404,
+      "learning_rate": 1.867763300865365e-05,
+      "loss": 0.3231,
+      "step": 10992
+    },
+    {
+      "epoch": 0.8083682623722332,
+      "grad_norm": 0.3816970884799957,
+      "learning_rate": 1.8663775457756728e-05,
+      "loss": 0.3505,
+      "step": 10993
+    },
+    {
+      "epoch": 0.80844179719097,
+      "grad_norm": 0.5059309005737305,
+      "learning_rate": 1.8649922520336316e-05,
+      "loss": 0.3207,
+      "step": 10994
+    },
+    {
+      "epoch": 0.8085153320097066,
+      "grad_norm": 0.5367390513420105,
+      "learning_rate": 1.863607419717813e-05,
+      "loss": 0.3338,
+      "step": 10995
+    },
+    {
+      "epoch": 0.8085888668284432,
+      "grad_norm": 0.40589308738708496,
+      "learning_rate": 1.8622230489067684e-05,
+      "loss": 0.338,
+      "step": 10996
+    },
+    {
+      "epoch": 0.80866240164718,
+      "grad_norm": 0.45322003960609436,
+      "learning_rate": 1.8608391396790224e-05,
+      "loss": 0.3101,
+      "step": 10997
+    },
+    {
+      "epoch": 0.8087359364659166,
+      "grad_norm": 0.3881501853466034,
+      "learning_rate": 1.859455692113069e-05,
+      "loss": 0.3293,
+      "step": 10998
+    },
+    {
+      "epoch": 0.8088094712846533,
+      "grad_norm": 0.5156011581420898,
+      "learning_rate": 1.8580727062873792e-05,
+      "loss": 0.3323,
+      "step": 10999
+    },
+    {
+      "epoch": 0.8088830061033899,
+      "grad_norm": 0.46907544136047363,
+      "learning_rate": 1.8566901822803996e-05,
+      "loss": 0.3233,
+      "step": 11000
+    },
+    {
+      "epoch": 0.8089565409221267,
+      "grad_norm": 0.5685644745826721,
+      "learning_rate": 1.855308120170548e-05,
+      "loss": 0.3143,
+      "step": 11001
+    },
+    {
+      "epoch": 0.8090300757408633,
+      "grad_norm": 0.45486757159233093,
+      "learning_rate": 1.853926520036212e-05,
+      "loss": 0.3353,
+      "step": 11002
+    },
+    {
+      "epoch": 0.8091036105596,
+      "grad_norm": 0.5674640536308289,
+      "learning_rate": 1.8525453819557613e-05,
+      "loss": 0.3202,
+      "step": 11003
+    },
+    {
+      "epoch": 0.8091771453783366,
+      "grad_norm": 0.3881370723247528,
+      "learning_rate": 1.8511647060075364e-05,
+      "loss": 0.3284,
+      "step": 11004
+    },
+    {
+      "epoch": 0.8092506801970734,
+      "grad_norm": 0.5592933297157288,
+      "learning_rate": 1.849784492269847e-05,
+      "loss": 0.3505,
+      "step": 11005
+    },
+    {
+      "epoch": 0.80932421501581,
+      "grad_norm": 0.47598549723625183,
+      "learning_rate": 1.8484047408209836e-05,
+      "loss": 0.3299,
+      "step": 11006
+    },
+    {
+      "epoch": 0.8093977498345467,
+      "grad_norm": 0.40640246868133545,
+      "learning_rate": 1.847025451739207e-05,
+      "loss": 0.314,
+      "step": 11007
+    },
+    {
+      "epoch": 0.8094712846532833,
+      "grad_norm": 0.712538480758667,
+      "learning_rate": 1.8456466251027493e-05,
+      "loss": 0.3263,
+      "step": 11008
+    },
+    {
+      "epoch": 0.80954481947202,
+      "grad_norm": 0.4194779098033905,
+      "learning_rate": 1.844268260989821e-05,
+      "loss": 0.342,
+      "step": 11009
+    },
+    {
+      "epoch": 0.8096183542907567,
+      "grad_norm": 0.6050736904144287,
+      "learning_rate": 1.8428903594786062e-05,
+      "loss": 0.3529,
+      "step": 11010
+    },
+    {
+      "epoch": 0.8096918891094933,
+      "grad_norm": 0.5593645572662354,
+      "learning_rate": 1.841512920647256e-05,
+      "loss": 0.3285,
+      "step": 11011
+    },
+    {
+      "epoch": 0.80976542392823,
+      "grad_norm": 0.5934981107711792,
+      "learning_rate": 1.840135944573903e-05,
+      "loss": 0.3427,
+      "step": 11012
+    },
+    {
+      "epoch": 0.8098389587469667,
+      "grad_norm": 0.43723246455192566,
+      "learning_rate": 1.838759431336655e-05,
+      "loss": 0.3066,
+      "step": 11013
+    },
+    {
+      "epoch": 0.8099124935657034,
+      "grad_norm": 0.5419145226478577,
+      "learning_rate": 1.837383381013581e-05,
+      "loss": 0.3141,
+      "step": 11014
+    },
+    {
+      "epoch": 0.80998602838444,
+      "grad_norm": 0.42314332723617554,
+      "learning_rate": 1.8360077936827358e-05,
+      "loss": 0.3477,
+      "step": 11015
+    },
+    {
+      "epoch": 0.8100595632031767,
+      "grad_norm": 0.45625045895576477,
+      "learning_rate": 1.834632669422144e-05,
+      "loss": 0.3186,
+      "step": 11016
+    },
+    {
+      "epoch": 0.8101330980219134,
+      "grad_norm": 0.6726362109184265,
+      "learning_rate": 1.8332580083098072e-05,
+      "loss": 0.3545,
+      "step": 11017
+    },
+    {
+      "epoch": 0.8102066328406501,
+      "grad_norm": 0.39735639095306396,
+      "learning_rate": 1.8318838104236924e-05,
+      "loss": 0.3307,
+      "step": 11018
+    },
+    {
+      "epoch": 0.8102801676593867,
+      "grad_norm": 0.5294108390808105,
+      "learning_rate": 1.830510075841747e-05,
+      "loss": 0.3343,
+      "step": 11019
+    },
+    {
+      "epoch": 0.8103537024781234,
+      "grad_norm": 0.5115461349487305,
+      "learning_rate": 1.8291368046418955e-05,
+      "loss": 0.3386,
+      "step": 11020
+    },
+    {
+      "epoch": 0.8104272372968601,
+      "grad_norm": 0.42215582728385925,
+      "learning_rate": 1.8277639969020243e-05,
+      "loss": 0.3501,
+      "step": 11021
+    },
+    {
+      "epoch": 0.8105007721155968,
+      "grad_norm": 0.6786707639694214,
+      "learning_rate": 1.8263916527000048e-05,
+      "loss": 0.3418,
+      "step": 11022
+    },
+    {
+      "epoch": 0.8105743069343334,
+      "grad_norm": 0.3896918296813965,
+      "learning_rate": 1.8250197721136775e-05,
+      "loss": 0.3153,
+      "step": 11023
+    },
+    {
+      "epoch": 0.81064784175307,
+      "grad_norm": 0.5291802883148193,
+      "learning_rate": 1.8236483552208572e-05,
+      "loss": 0.3314,
+      "step": 11024
+    },
+    {
+      "epoch": 0.8107213765718068,
+      "grad_norm": 0.4590732157230377,
+      "learning_rate": 1.822277402099328e-05,
+      "loss": 0.3414,
+      "step": 11025
+    },
+    {
+      "epoch": 0.8107949113905434,
+      "grad_norm": 0.4871167540550232,
+      "learning_rate": 1.8209069128268564e-05,
+      "loss": 0.3608,
+      "step": 11026
+    },
+    {
+      "epoch": 0.8108684462092801,
+      "grad_norm": 0.6986081600189209,
+      "learning_rate": 1.819536887481177e-05,
+      "loss": 0.3292,
+      "step": 11027
+    },
+    {
+      "epoch": 0.8109419810280167,
+      "grad_norm": 0.44056424498558044,
+      "learning_rate": 1.8181673261399967e-05,
+      "loss": 0.3122,
+      "step": 11028
+    },
+    {
+      "epoch": 0.8110155158467535,
+      "grad_norm": 0.5087268352508545,
+      "learning_rate": 1.8167982288810005e-05,
+      "loss": 0.3701,
+      "step": 11029
+    },
+    {
+      "epoch": 0.8110890506654901,
+      "grad_norm": 0.425006628036499,
+      "learning_rate": 1.815429595781848e-05,
+      "loss": 0.3266,
+      "step": 11030
+    },
+    {
+      "epoch": 0.8111625854842268,
+      "grad_norm": 0.4288313388824463,
+      "learning_rate": 1.8140614269201628e-05,
+      "loss": 0.3027,
+      "step": 11031
+    },
+    {
+      "epoch": 0.8112361203029634,
+      "grad_norm": 0.44218122959136963,
+      "learning_rate": 1.812693722373553e-05,
+      "loss": 0.3271,
+      "step": 11032
+    },
+    {
+      "epoch": 0.8113096551217002,
+      "grad_norm": 0.38486379384994507,
+      "learning_rate": 1.8113264822195976e-05,
+      "loss": 0.3254,
+      "step": 11033
+    },
+    {
+      "epoch": 0.8113831899404368,
+      "grad_norm": 0.4492456912994385,
+      "learning_rate": 1.809959706535843e-05,
+      "loss": 0.3455,
+      "step": 11034
+    },
+    {
+      "epoch": 0.8114567247591735,
+      "grad_norm": 0.5656858086585999,
+      "learning_rate": 1.808593395399818e-05,
+      "loss": 0.3194,
+      "step": 11035
+    },
+    {
+      "epoch": 0.8115302595779101,
+      "grad_norm": 0.44327181577682495,
+      "learning_rate": 1.8072275488890254e-05,
+      "loss": 0.3521,
+      "step": 11036
+    },
+    {
+      "epoch": 0.8116037943966468,
+      "grad_norm": 0.5163567662239075,
+      "learning_rate": 1.805862167080927e-05,
+      "loss": 0.3262,
+      "step": 11037
+    },
+    {
+      "epoch": 0.8116773292153835,
+      "grad_norm": 0.5490378141403198,
+      "learning_rate": 1.8044972500529744e-05,
+      "loss": 0.3285,
+      "step": 11038
+    },
+    {
+      "epoch": 0.8117508640341201,
+      "grad_norm": 0.5784122943878174,
+      "learning_rate": 1.8031327978825875e-05,
+      "loss": 0.3504,
+      "step": 11039
+    },
+    {
+      "epoch": 0.8118243988528568,
+      "grad_norm": 0.6582843065261841,
+      "learning_rate": 1.8017688106471597e-05,
+      "loss": 0.3204,
+      "step": 11040
+    },
+    {
+      "epoch": 0.8118979336715935,
+      "grad_norm": 0.55445796251297,
+      "learning_rate": 1.800405288424056e-05,
+      "loss": 0.3298,
+      "step": 11041
+    },
+    {
+      "epoch": 0.8119714684903302,
+      "grad_norm": 0.40344715118408203,
+      "learning_rate": 1.7990422312906165e-05,
+      "loss": 0.3107,
+      "step": 11042
+    },
+    {
+      "epoch": 0.8120450033090668,
+      "grad_norm": 0.5185624957084656,
+      "learning_rate": 1.79767963932416e-05,
+      "loss": 0.3307,
+      "step": 11043
+    },
+    {
+      "epoch": 0.8121185381278035,
+      "grad_norm": 0.47096648812294006,
+      "learning_rate": 1.796317512601967e-05,
+      "loss": 0.3431,
+      "step": 11044
+    },
+    {
+      "epoch": 0.8121920729465402,
+      "grad_norm": 0.43855616450309753,
+      "learning_rate": 1.7949558512013033e-05,
+      "loss": 0.3238,
+      "step": 11045
+    },
+    {
+      "epoch": 0.8122656077652769,
+      "grad_norm": 0.45170557498931885,
+      "learning_rate": 1.7935946551994053e-05,
+      "loss": 0.3151,
+      "step": 11046
+    },
+    {
+      "epoch": 0.8123391425840135,
+      "grad_norm": 0.46497127413749695,
+      "learning_rate": 1.7922339246734754e-05,
+      "loss": 0.3517,
+      "step": 11047
+    },
+    {
+      "epoch": 0.8124126774027502,
+      "grad_norm": 0.4055930972099304,
+      "learning_rate": 1.7908736597007024e-05,
+      "loss": 0.325,
+      "step": 11048
+    },
+    {
+      "epoch": 0.8124862122214869,
+      "grad_norm": 0.39922842383384705,
+      "learning_rate": 1.7895138603582395e-05,
+      "loss": 0.3183,
+      "step": 11049
+    },
+    {
+      "epoch": 0.8125597470402236,
+      "grad_norm": 0.4097026586532593,
+      "learning_rate": 1.7881545267232123e-05,
+      "loss": 0.3262,
+      "step": 11050
+    },
+    {
+      "epoch": 0.8126332818589602,
+      "grad_norm": 0.3719673454761505,
+      "learning_rate": 1.7867956588727264e-05,
+      "loss": 0.3178,
+      "step": 11051
+    },
+    {
+      "epoch": 0.8127068166776968,
+      "grad_norm": 0.41076770424842834,
+      "learning_rate": 1.7854372568838596e-05,
+      "loss": 0.3701,
+      "step": 11052
+    },
+    {
+      "epoch": 0.8127803514964336,
+      "grad_norm": 0.5449963808059692,
+      "learning_rate": 1.7840793208336625e-05,
+      "loss": 0.3489,
+      "step": 11053
+    },
+    {
+      "epoch": 0.8128538863151702,
+      "grad_norm": 0.48015257716178894,
+      "learning_rate": 1.7827218507991537e-05,
+      "loss": 0.3152,
+      "step": 11054
+    },
+    {
+      "epoch": 0.8129274211339069,
+      "grad_norm": 0.40048718452453613,
+      "learning_rate": 1.781364846857334e-05,
+      "loss": 0.3375,
+      "step": 11055
+    },
+    {
+      "epoch": 0.8130009559526435,
+      "grad_norm": 0.39918673038482666,
+      "learning_rate": 1.780008309085177e-05,
+      "loss": 0.3399,
+      "step": 11056
+    },
+    {
+      "epoch": 0.8130744907713803,
+      "grad_norm": 0.5204900503158569,
+      "learning_rate": 1.7786522375596205e-05,
+      "loss": 0.3087,
+      "step": 11057
+    },
+    {
+      "epoch": 0.8131480255901169,
+      "grad_norm": 0.4242802560329437,
+      "learning_rate": 1.7772966323575857e-05,
+      "loss": 0.3593,
+      "step": 11058
+    },
+    {
+      "epoch": 0.8132215604088536,
+      "grad_norm": 0.4021637439727783,
+      "learning_rate": 1.775941493555968e-05,
+      "loss": 0.3256,
+      "step": 11059
+    },
+    {
+      "epoch": 0.8132950952275902,
+      "grad_norm": 0.4773935377597809,
+      "learning_rate": 1.774586821231623e-05,
+      "loss": 0.3384,
+      "step": 11060
+    },
+    {
+      "epoch": 0.813368630046327,
+      "grad_norm": 0.46989884972572327,
+      "learning_rate": 1.7732326154613953e-05,
+      "loss": 0.3341,
+      "step": 11061
+    },
+    {
+      "epoch": 0.8134421648650636,
+      "grad_norm": 0.5555069446563721,
+      "learning_rate": 1.771878876322095e-05,
+      "loss": 0.3376,
+      "step": 11062
+    },
+    {
+      "epoch": 0.8135156996838003,
+      "grad_norm": 0.48696523904800415,
+      "learning_rate": 1.7705256038905105e-05,
+      "loss": 0.333,
+      "step": 11063
+    },
+    {
+      "epoch": 0.813589234502537,
+      "grad_norm": 0.559050440788269,
+      "learning_rate": 1.7691727982433958e-05,
+      "loss": 0.3322,
+      "step": 11064
+    },
+    {
+      "epoch": 0.8136627693212736,
+      "grad_norm": 0.5116819143295288,
+      "learning_rate": 1.7678204594574876e-05,
+      "loss": 0.3446,
+      "step": 11065
+    },
+    {
+      "epoch": 0.8137363041400103,
+      "grad_norm": 0.42290541529655457,
+      "learning_rate": 1.766468587609492e-05,
+      "loss": 0.3252,
+      "step": 11066
+    },
+    {
+      "epoch": 0.8138098389587469,
+      "grad_norm": 0.5960253477096558,
+      "learning_rate": 1.765117182776085e-05,
+      "loss": 0.3429,
+      "step": 11067
+    },
+    {
+      "epoch": 0.8138833737774837,
+      "grad_norm": 0.5587530732154846,
+      "learning_rate": 1.763766245033923e-05,
+      "loss": 0.3518,
+      "step": 11068
+    },
+    {
+      "epoch": 0.8139569085962203,
+      "grad_norm": 0.6187414526939392,
+      "learning_rate": 1.7624157744596325e-05,
+      "loss": 0.3534,
+      "step": 11069
+    },
+    {
+      "epoch": 0.814030443414957,
+      "grad_norm": 0.47143328189849854,
+      "learning_rate": 1.7610657711298116e-05,
+      "loss": 0.3675,
+      "step": 11070
+    },
+    {
+      "epoch": 0.8141039782336936,
+      "grad_norm": 0.5516613125801086,
+      "learning_rate": 1.759716235121037e-05,
+      "loss": 0.3538,
+      "step": 11071
+    },
+    {
+      "epoch": 0.8141775130524304,
+      "grad_norm": 0.4317902624607086,
+      "learning_rate": 1.7583671665098544e-05,
+      "loss": 0.3546,
+      "step": 11072
+    },
+    {
+      "epoch": 0.814251047871167,
+      "grad_norm": 0.4995206892490387,
+      "learning_rate": 1.757018565372781e-05,
+      "loss": 0.3513,
+      "step": 11073
+    },
+    {
+      "epoch": 0.8143245826899037,
+      "grad_norm": 0.528494119644165,
+      "learning_rate": 1.755670431786314e-05,
+      "loss": 0.3304,
+      "step": 11074
+    },
+    {
+      "epoch": 0.8143981175086403,
+      "grad_norm": 0.4684215188026428,
+      "learning_rate": 1.75432276582692e-05,
+      "loss": 0.321,
+      "step": 11075
+    },
+    {
+      "epoch": 0.8144716523273771,
+      "grad_norm": 0.6749570965766907,
+      "learning_rate": 1.7529755675710437e-05,
+      "loss": 0.3521,
+      "step": 11076
+    },
+    {
+      "epoch": 0.8145451871461137,
+      "grad_norm": 0.4968467354774475,
+      "learning_rate": 1.7516288370950938e-05,
+      "loss": 0.3542,
+      "step": 11077
+    },
+    {
+      "epoch": 0.8146187219648504,
+      "grad_norm": 0.4229936897754669,
+      "learning_rate": 1.7502825744754613e-05,
+      "loss": 0.3577,
+      "step": 11078
+    },
+    {
+      "epoch": 0.814692256783587,
+      "grad_norm": 0.7255519032478333,
+      "learning_rate": 1.7489367797885103e-05,
+      "loss": 0.3622,
+      "step": 11079
+    },
+    {
+      "epoch": 0.8147657916023237,
+      "grad_norm": 0.5713162422180176,
+      "learning_rate": 1.7475914531105695e-05,
+      "loss": 0.3179,
+      "step": 11080
+    },
+    {
+      "epoch": 0.8148393264210604,
+      "grad_norm": 0.4820696711540222,
+      "learning_rate": 1.7462465945179517e-05,
+      "loss": 0.343,
+      "step": 11081
+    },
+    {
+      "epoch": 0.814912861239797,
+      "grad_norm": 0.48673170804977417,
+      "learning_rate": 1.744902204086939e-05,
+      "loss": 0.3334,
+      "step": 11082
+    },
+    {
+      "epoch": 0.8149863960585337,
+      "grad_norm": 0.5993697047233582,
+      "learning_rate": 1.7435582818937857e-05,
+      "loss": 0.3712,
+      "step": 11083
+    },
+    {
+      "epoch": 0.8150599308772704,
+      "grad_norm": 0.5683297514915466,
+      "learning_rate": 1.7422148280147177e-05,
+      "loss": 0.3361,
+      "step": 11084
+    },
+    {
+      "epoch": 0.8151334656960071,
+      "grad_norm": 0.43301406502723694,
+      "learning_rate": 1.7408718425259397e-05,
+      "loss": 0.3496,
+      "step": 11085
+    },
+    {
+      "epoch": 0.8152070005147437,
+      "grad_norm": 0.6033175587654114,
+      "learning_rate": 1.7395293255036294e-05,
+      "loss": 0.3252,
+      "step": 11086
+    },
+    {
+      "epoch": 0.8152805353334804,
+      "grad_norm": 0.5760914087295532,
+      "learning_rate": 1.7381872770239304e-05,
+      "loss": 0.3383,
+      "step": 11087
+    },
+    {
+      "epoch": 0.8153540701522171,
+      "grad_norm": 0.6514500975608826,
+      "learning_rate": 1.736845697162969e-05,
+      "loss": 0.3429,
+      "step": 11088
+    },
+    {
+      "epoch": 0.8154276049709538,
+      "grad_norm": 0.5669245719909668,
+      "learning_rate": 1.735504585996841e-05,
+      "loss": 0.3731,
+      "step": 11089
+    },
+    {
+      "epoch": 0.8155011397896904,
+      "grad_norm": 0.5554004907608032,
+      "learning_rate": 1.7341639436016132e-05,
+      "loss": 0.3513,
+      "step": 11090
+    },
+    {
+      "epoch": 0.8155746746084271,
+      "grad_norm": 0.49474555253982544,
+      "learning_rate": 1.73282377005333e-05,
+      "loss": 0.3257,
+      "step": 11091
+    },
+    {
+      "epoch": 0.8156482094271638,
+      "grad_norm": 0.5017703771591187,
+      "learning_rate": 1.7314840654280095e-05,
+      "loss": 0.3624,
+      "step": 11092
+    },
+    {
+      "epoch": 0.8157217442459004,
+      "grad_norm": 0.6098153591156006,
+      "learning_rate": 1.7301448298016364e-05,
+      "loss": 0.3636,
+      "step": 11093
+    },
+    {
+      "epoch": 0.8157952790646371,
+      "grad_norm": 0.5311233997344971,
+      "learning_rate": 1.728806063250179e-05,
+      "loss": 0.3596,
+      "step": 11094
+    },
+    {
+      "epoch": 0.8158688138833737,
+      "grad_norm": 0.41358035802841187,
+      "learning_rate": 1.7274677658495698e-05,
+      "loss": 0.3511,
+      "step": 11095
+    },
+    {
+      "epoch": 0.8159423487021105,
+      "grad_norm": 0.49936962127685547,
+      "learning_rate": 1.7261299376757166e-05,
+      "loss": 0.3348,
+      "step": 11096
+    },
+    {
+      "epoch": 0.8160158835208471,
+      "grad_norm": 0.5369852185249329,
+      "learning_rate": 1.724792578804506e-05,
+      "loss": 0.3441,
+      "step": 11097
+    },
+    {
+      "epoch": 0.8160894183395838,
+      "grad_norm": 0.37910252809524536,
+      "learning_rate": 1.7234556893117935e-05,
+      "loss": 0.3916,
+      "step": 11098
+    },
+    {
+      "epoch": 0.8161629531583204,
+      "grad_norm": 0.54303377866745,
+      "learning_rate": 1.722119269273411e-05,
+      "loss": 0.3365,
+      "step": 11099
+    },
+    {
+      "epoch": 0.8162364879770572,
+      "grad_norm": 0.4249476492404938,
+      "learning_rate": 1.720783318765157e-05,
+      "loss": 0.3276,
+      "step": 11100
+    },
+    {
+      "epoch": 0.8163100227957938,
+      "grad_norm": 0.49701496958732605,
+      "learning_rate": 1.719447837862811e-05,
+      "loss": 0.3503,
+      "step": 11101
+    },
+    {
+      "epoch": 0.8163835576145305,
+      "grad_norm": 0.45838066935539246,
+      "learning_rate": 1.718112826642124e-05,
+      "loss": 0.3516,
+      "step": 11102
+    },
+    {
+      "epoch": 0.8164570924332671,
+      "grad_norm": 0.3982459306716919,
+      "learning_rate": 1.7167782851788162e-05,
+      "loss": 0.3734,
+      "step": 11103
+    },
+    {
+      "epoch": 0.8165306272520039,
+      "grad_norm": 0.44120457768440247,
+      "learning_rate": 1.715444213548586e-05,
+      "loss": 0.3477,
+      "step": 11104
+    },
+    {
+      "epoch": 0.8166041620707405,
+      "grad_norm": 0.4356992840766907,
+      "learning_rate": 1.7141106118271045e-05,
+      "loss": 0.365,
+      "step": 11105
+    },
+    {
+      "epoch": 0.8166776968894772,
+      "grad_norm": 0.44425472617149353,
+      "learning_rate": 1.7127774800900143e-05,
+      "loss": 0.3443,
+      "step": 11106
+    },
+    {
+      "epoch": 0.8167512317082138,
+      "grad_norm": 0.44816604256629944,
+      "learning_rate": 1.7114448184129284e-05,
+      "loss": 0.375,
+      "step": 11107
+    },
+    {
+      "epoch": 0.8168247665269505,
+      "grad_norm": 0.6002014875411987,
+      "learning_rate": 1.7101126268714395e-05,
+      "loss": 0.3345,
+      "step": 11108
+    },
+    {
+      "epoch": 0.8168983013456872,
+      "grad_norm": 0.4437272548675537,
+      "learning_rate": 1.708780905541114e-05,
+      "loss": 0.3391,
+      "step": 11109
+    },
+    {
+      "epoch": 0.8169718361644238,
+      "grad_norm": 0.4155541956424713,
+      "learning_rate": 1.7074496544974828e-05,
+      "loss": 0.3523,
+      "step": 11110
+    },
+    {
+      "epoch": 0.8170453709831605,
+      "grad_norm": 0.7729811668395996,
+      "learning_rate": 1.7061188738160583e-05,
+      "loss": 0.3823,
+      "step": 11111
+    },
+    {
+      "epoch": 0.8171189058018972,
+      "grad_norm": 0.39900532364845276,
+      "learning_rate": 1.704788563572327e-05,
+      "loss": 0.3451,
+      "step": 11112
+    },
+    {
+      "epoch": 0.8171924406206339,
+      "grad_norm": 0.6786566972732544,
+      "learning_rate": 1.703458723841741e-05,
+      "loss": 0.3597,
+      "step": 11113
+    },
+    {
+      "epoch": 0.8172659754393705,
+      "grad_norm": 0.6522272229194641,
+      "learning_rate": 1.702129354699731e-05,
+      "loss": 0.4,
+      "step": 11114
+    },
+    {
+      "epoch": 0.8173395102581072,
+      "grad_norm": 0.5306702256202698,
+      "learning_rate": 1.700800456221704e-05,
+      "loss": 0.3414,
+      "step": 11115
+    },
+    {
+      "epoch": 0.8174130450768439,
+      "grad_norm": 0.7218397855758667,
+      "learning_rate": 1.6994720284830302e-05,
+      "loss": 0.3436,
+      "step": 11116
+    },
+    {
+      "epoch": 0.8174865798955806,
+      "grad_norm": 0.5853484272956848,
+      "learning_rate": 1.6981440715590635e-05,
+      "loss": 0.3718,
+      "step": 11117
+    },
+    {
+      "epoch": 0.8175601147143172,
+      "grad_norm": 0.5214657187461853,
+      "learning_rate": 1.6968165855251317e-05,
+      "loss": 0.349,
+      "step": 11118
+    },
+    {
+      "epoch": 0.817633649533054,
+      "grad_norm": 0.6052377223968506,
+      "learning_rate": 1.69548957045652e-05,
+      "loss": 0.3862,
+      "step": 11119
+    },
+    {
+      "epoch": 0.8177071843517906,
+      "grad_norm": 0.6135154962539673,
+      "learning_rate": 1.6941630264285046e-05,
+      "loss": 0.3273,
+      "step": 11120
+    },
+    {
+      "epoch": 0.8177807191705272,
+      "grad_norm": 0.4679131507873535,
+      "learning_rate": 1.692836953516328e-05,
+      "loss": 0.3428,
+      "step": 11121
+    },
+    {
+      "epoch": 0.8178542539892639,
+      "grad_norm": 0.5340070128440857,
+      "learning_rate": 1.6915113517952084e-05,
+      "loss": 0.3537,
+      "step": 11122
+    },
+    {
+      "epoch": 0.8179277888080005,
+      "grad_norm": 0.6074815392494202,
+      "learning_rate": 1.6901862213403308e-05,
+      "loss": 0.3456,
+      "step": 11123
+    },
+    {
+      "epoch": 0.8180013236267373,
+      "grad_norm": 0.4394057095050812,
+      "learning_rate": 1.68886156222686e-05,
+      "loss": 0.3403,
+      "step": 11124
+    },
+    {
+      "epoch": 0.8180748584454739,
+      "grad_norm": 0.5302613377571106,
+      "learning_rate": 1.687537374529936e-05,
+      "loss": 0.3677,
+      "step": 11125
+    },
+    {
+      "epoch": 0.8181483932642106,
+      "grad_norm": 0.4705594480037689,
+      "learning_rate": 1.6862136583246614e-05,
+      "loss": 0.3555,
+      "step": 11126
+    },
+    {
+      "epoch": 0.8182219280829472,
+      "grad_norm": 0.5950855016708374,
+      "learning_rate": 1.684890413686122e-05,
+      "loss": 0.3478,
+      "step": 11127
+    },
+    {
+      "epoch": 0.818295462901684,
+      "grad_norm": 0.5309828519821167,
+      "learning_rate": 1.6835676406893764e-05,
+      "loss": 0.3511,
+      "step": 11128
+    },
+    {
+      "epoch": 0.8183689977204206,
+      "grad_norm": 0.522314727306366,
+      "learning_rate": 1.6822453394094505e-05,
+      "loss": 0.3614,
+      "step": 11129
+    },
+    {
+      "epoch": 0.8184425325391573,
+      "grad_norm": 0.6715375781059265,
+      "learning_rate": 1.680923509921345e-05,
+      "loss": 0.3342,
+      "step": 11130
+    },
+    {
+      "epoch": 0.818516067357894,
+      "grad_norm": 0.5179555416107178,
+      "learning_rate": 1.6796021523000372e-05,
+      "loss": 0.3681,
+      "step": 11131
+    },
+    {
+      "epoch": 0.8185896021766307,
+      "grad_norm": 0.4664762318134308,
+      "learning_rate": 1.6782812666204783e-05,
+      "loss": 0.3791,
+      "step": 11132
+    },
+    {
+      "epoch": 0.8186631369953673,
+      "grad_norm": 0.37638646364212036,
+      "learning_rate": 1.6769608529575866e-05,
+      "loss": 0.3413,
+      "step": 11133
+    },
+    {
+      "epoch": 0.818736671814104,
+      "grad_norm": 0.4652269780635834,
+      "learning_rate": 1.6756409113862582e-05,
+      "loss": 0.3423,
+      "step": 11134
+    },
+    {
+      "epoch": 0.8188102066328407,
+      "grad_norm": 0.49063822627067566,
+      "learning_rate": 1.6743214419813656e-05,
+      "loss": 0.3371,
+      "step": 11135
+    },
+    {
+      "epoch": 0.8188837414515773,
+      "grad_norm": 0.41432222723960876,
+      "learning_rate": 1.6730024448177438e-05,
+      "loss": 0.3503,
+      "step": 11136
+    },
+    {
+      "epoch": 0.818957276270314,
+      "grad_norm": 0.6136830449104309,
+      "learning_rate": 1.6716839199702118e-05,
+      "loss": 0.3532,
+      "step": 11137
+    },
+    {
+      "epoch": 0.8190308110890506,
+      "grad_norm": 0.4015312194824219,
+      "learning_rate": 1.6703658675135592e-05,
+      "loss": 0.3241,
+      "step": 11138
+    },
+    {
+      "epoch": 0.8191043459077874,
+      "grad_norm": 0.47476908564567566,
+      "learning_rate": 1.6690482875225432e-05,
+      "loss": 0.3583,
+      "step": 11139
+    },
+    {
+      "epoch": 0.819177880726524,
+      "grad_norm": 0.4482343792915344,
+      "learning_rate": 1.667731180071901e-05,
+      "loss": 0.3492,
+      "step": 11140
+    },
+    {
+      "epoch": 0.8192514155452607,
+      "grad_norm": 0.6667893528938293,
+      "learning_rate": 1.6664145452363433e-05,
+      "loss": 0.3566,
+      "step": 11141
+    },
+    {
+      "epoch": 0.8193249503639973,
+      "grad_norm": 0.5049400925636292,
+      "learning_rate": 1.665098383090543e-05,
+      "loss": 0.3396,
+      "step": 11142
+    },
+    {
+      "epoch": 0.8193984851827341,
+      "grad_norm": 0.4905174672603607,
+      "learning_rate": 1.6637826937091584e-05,
+      "loss": 0.3467,
+      "step": 11143
+    },
+    {
+      "epoch": 0.8194720200014707,
+      "grad_norm": 0.4845331311225891,
+      "learning_rate": 1.662467477166817e-05,
+      "loss": 0.335,
+      "step": 11144
+    },
+    {
+      "epoch": 0.8195455548202074,
+      "grad_norm": 0.5871942639350891,
+      "learning_rate": 1.6611527335381217e-05,
+      "loss": 0.3529,
+      "step": 11145
+    },
+    {
+      "epoch": 0.819619089638944,
+      "grad_norm": 0.3966774642467499,
+      "learning_rate": 1.6598384628976404e-05,
+      "loss": 0.3487,
+      "step": 11146
+    },
+    {
+      "epoch": 0.8196926244576808,
+      "grad_norm": 0.5612010359764099,
+      "learning_rate": 1.658524665319925e-05,
+      "loss": 0.3681,
+      "step": 11147
+    },
+    {
+      "epoch": 0.8197661592764174,
+      "grad_norm": 0.47630494832992554,
+      "learning_rate": 1.6572113408794955e-05,
+      "loss": 0.3339,
+      "step": 11148
+    },
+    {
+      "epoch": 0.819839694095154,
+      "grad_norm": 0.4134984314441681,
+      "learning_rate": 1.65589848965084e-05,
+      "loss": 0.3521,
+      "step": 11149
+    },
+    {
+      "epoch": 0.8199132289138907,
+      "grad_norm": 0.44248199462890625,
+      "learning_rate": 1.6545861117084284e-05,
+      "loss": 0.3435,
+      "step": 11150
+    },
+    {
+      "epoch": 0.8199867637326274,
+      "grad_norm": 0.4866008162498474,
+      "learning_rate": 1.6532742071267028e-05,
+      "loss": 0.3326,
+      "step": 11151
+    },
+    {
+      "epoch": 0.8200602985513641,
+      "grad_norm": 0.39689430594444275,
+      "learning_rate": 1.6519627759800694e-05,
+      "loss": 0.3208,
+      "step": 11152
+    },
+    {
+      "epoch": 0.8201338333701007,
+      "grad_norm": 0.5208266377449036,
+      "learning_rate": 1.650651818342921e-05,
+      "loss": 0.3427,
+      "step": 11153
+    },
+    {
+      "epoch": 0.8202073681888374,
+      "grad_norm": 0.41167744994163513,
+      "learning_rate": 1.649341334289609e-05,
+      "loss": 0.3424,
+      "step": 11154
+    },
+    {
+      "epoch": 0.8202809030075741,
+      "grad_norm": 0.3832330107688904,
+      "learning_rate": 1.6480313238944733e-05,
+      "loss": 0.315,
+      "step": 11155
+    },
+    {
+      "epoch": 0.8203544378263108,
+      "grad_norm": 0.5252886414527893,
+      "learning_rate": 1.6467217872318118e-05,
+      "loss": 0.3451,
+      "step": 11156
+    },
+    {
+      "epoch": 0.8204279726450474,
+      "grad_norm": 0.56669682264328,
+      "learning_rate": 1.6454127243759066e-05,
+      "loss": 0.3367,
+      "step": 11157
+    },
+    {
+      "epoch": 0.8205015074637841,
+      "grad_norm": 0.5294145941734314,
+      "learning_rate": 1.64410413540101e-05,
+      "loss": 0.3485,
+      "step": 11158
+    },
+    {
+      "epoch": 0.8205750422825208,
+      "grad_norm": 0.4994440972805023,
+      "learning_rate": 1.642796020381343e-05,
+      "loss": 0.3522,
+      "step": 11159
+    },
+    {
+      "epoch": 0.8206485771012575,
+      "grad_norm": 0.47278767824172974,
+      "learning_rate": 1.6414883793911064e-05,
+      "loss": 0.3255,
+      "step": 11160
+    },
+    {
+      "epoch": 0.8207221119199941,
+      "grad_norm": 0.44726336002349854,
+      "learning_rate": 1.6401812125044725e-05,
+      "loss": 0.3877,
+      "step": 11161
+    },
+    {
+      "epoch": 0.8207956467387308,
+      "grad_norm": 0.5330522060394287,
+      "learning_rate": 1.6388745197955792e-05,
+      "loss": 0.3651,
+      "step": 11162
+    },
+    {
+      "epoch": 0.8208691815574675,
+      "grad_norm": 0.7215948104858398,
+      "learning_rate": 1.637568301338547e-05,
+      "loss": 0.3774,
+      "step": 11163
+    },
+    {
+      "epoch": 0.8209427163762041,
+      "grad_norm": 0.40432703495025635,
+      "learning_rate": 1.636262557207472e-05,
+      "loss": 0.3178,
+      "step": 11164
+    },
+    {
+      "epoch": 0.8210162511949408,
+      "grad_norm": 0.5855151414871216,
+      "learning_rate": 1.634957287476405e-05,
+      "loss": 0.3767,
+      "step": 11165
+    },
+    {
+      "epoch": 0.8210897860136774,
+      "grad_norm": 0.43734443187713623,
+      "learning_rate": 1.6336524922193895e-05,
+      "loss": 0.371,
+      "step": 11166
+    },
+    {
+      "epoch": 0.8211633208324142,
+      "grad_norm": 0.543808102607727,
+      "learning_rate": 1.632348171510434e-05,
+      "loss": 0.3448,
+      "step": 11167
+    },
+    {
+      "epoch": 0.8212368556511508,
+      "grad_norm": 0.48968198895454407,
+      "learning_rate": 1.631044325423523e-05,
+      "loss": 0.3475,
+      "step": 11168
+    },
+    {
+      "epoch": 0.8213103904698875,
+      "grad_norm": 0.5164883136749268,
+      "learning_rate": 1.629740954032608e-05,
+      "loss": 0.3532,
+      "step": 11169
+    },
+    {
+      "epoch": 0.8213839252886241,
+      "grad_norm": 0.3985673487186432,
+      "learning_rate": 1.62843805741162e-05,
+      "loss": 0.3486,
+      "step": 11170
+    },
+    {
+      "epoch": 0.8214574601073609,
+      "grad_norm": 0.41379326581954956,
+      "learning_rate": 1.6271356356344614e-05,
+      "loss": 0.3376,
+      "step": 11171
+    },
+    {
+      "epoch": 0.8215309949260975,
+      "grad_norm": 0.40191566944122314,
+      "learning_rate": 1.6258336887750046e-05,
+      "loss": 0.3278,
+      "step": 11172
+    },
+    {
+      "epoch": 0.8216045297448342,
+      "grad_norm": 0.42428484559059143,
+      "learning_rate": 1.6245322169070977e-05,
+      "loss": 0.3494,
+      "step": 11173
+    },
+    {
+      "epoch": 0.8216780645635708,
+      "grad_norm": 0.46732351183891296,
+      "learning_rate": 1.6232312201045652e-05,
+      "loss": 0.3499,
+      "step": 11174
+    },
+    {
+      "epoch": 0.8217515993823076,
+      "grad_norm": 0.42614027857780457,
+      "learning_rate": 1.6219306984411963e-05,
+      "loss": 0.3381,
+      "step": 11175
+    },
+    {
+      "epoch": 0.8218251342010442,
+      "grad_norm": 0.3984580636024475,
+      "learning_rate": 1.620630651990761e-05,
+      "loss": 0.3332,
+      "step": 11176
+    },
+    {
+      "epoch": 0.8218986690197808,
+      "grad_norm": 0.5103112459182739,
+      "learning_rate": 1.619331080826999e-05,
+      "loss": 0.3624,
+      "step": 11177
+    },
+    {
+      "epoch": 0.8219722038385175,
+      "grad_norm": 0.3845076858997345,
+      "learning_rate": 1.6180319850236202e-05,
+      "loss": 0.3412,
+      "step": 11178
+    },
+    {
+      "epoch": 0.8220457386572542,
+      "grad_norm": 0.4047830104827881,
+      "learning_rate": 1.6167333646543136e-05,
+      "loss": 0.3664,
+      "step": 11179
+    },
+    {
+      "epoch": 0.8221192734759909,
+      "grad_norm": 0.5079177618026733,
+      "learning_rate": 1.615435219792737e-05,
+      "loss": 0.3616,
+      "step": 11180
+    },
+    {
+      "epoch": 0.8221928082947275,
+      "grad_norm": 0.4153731167316437,
+      "learning_rate": 1.6141375505125268e-05,
+      "loss": 0.3532,
+      "step": 11181
+    },
+    {
+      "epoch": 0.8222663431134642,
+      "grad_norm": 0.44639909267425537,
+      "learning_rate": 1.6128403568872807e-05,
+      "loss": 0.3472,
+      "step": 11182
+    },
+    {
+      "epoch": 0.8223398779322009,
+      "grad_norm": 0.4247015416622162,
+      "learning_rate": 1.6115436389905813e-05,
+      "loss": 0.3304,
+      "step": 11183
+    },
+    {
+      "epoch": 0.8224134127509376,
+      "grad_norm": 0.4835190176963806,
+      "learning_rate": 1.610247396895983e-05,
+      "loss": 0.3309,
+      "step": 11184
+    },
+    {
+      "epoch": 0.8224869475696742,
+      "grad_norm": 0.4178142845630646,
+      "learning_rate": 1.6089516306770024e-05,
+      "loss": 0.3848,
+      "step": 11185
+    },
+    {
+      "epoch": 0.822560482388411,
+      "grad_norm": 0.3731592297554016,
+      "learning_rate": 1.607656340407142e-05,
+      "loss": 0.3641,
+      "step": 11186
+    },
+    {
+      "epoch": 0.8226340172071476,
+      "grad_norm": 0.43289801478385925,
+      "learning_rate": 1.606361526159872e-05,
+      "loss": 0.3438,
+      "step": 11187
+    },
+    {
+      "epoch": 0.8227075520258843,
+      "grad_norm": 0.43719905614852905,
+      "learning_rate": 1.6050671880086354e-05,
+      "loss": 0.3562,
+      "step": 11188
+    },
+    {
+      "epoch": 0.8227810868446209,
+      "grad_norm": 0.4374128580093384,
+      "learning_rate": 1.603773326026845e-05,
+      "loss": 0.3382,
+      "step": 11189
+    },
+    {
+      "epoch": 0.8228546216633577,
+      "grad_norm": 0.3884068727493286,
+      "learning_rate": 1.602479940287892e-05,
+      "loss": 0.3199,
+      "step": 11190
+    },
+    {
+      "epoch": 0.8229281564820943,
+      "grad_norm": 0.48704448342323303,
+      "learning_rate": 1.6011870308651413e-05,
+      "loss": 0.3667,
+      "step": 11191
+    },
+    {
+      "epoch": 0.8230016913008309,
+      "grad_norm": 0.5189053416252136,
+      "learning_rate": 1.5998945978319236e-05,
+      "loss": 0.3488,
+      "step": 11192
+    },
+    {
+      "epoch": 0.8230752261195676,
+      "grad_norm": 0.5274531841278076,
+      "learning_rate": 1.5986026412615495e-05,
+      "loss": 0.3262,
+      "step": 11193
+    },
+    {
+      "epoch": 0.8231487609383042,
+      "grad_norm": 0.3958130478858948,
+      "learning_rate": 1.5973111612273017e-05,
+      "loss": 0.3471,
+      "step": 11194
+    },
+    {
+      "epoch": 0.823222295757041,
+      "grad_norm": 0.46836647391319275,
+      "learning_rate": 1.5960201578024302e-05,
+      "loss": 0.3523,
+      "step": 11195
+    },
+    {
+      "epoch": 0.8232958305757776,
+      "grad_norm": 0.6595235466957092,
+      "learning_rate": 1.5947296310601657e-05,
+      "loss": 0.3493,
+      "step": 11196
+    },
+    {
+      "epoch": 0.8233693653945143,
+      "grad_norm": 0.4571598172187805,
+      "learning_rate": 1.5934395810737078e-05,
+      "loss": 0.384,
+      "step": 11197
+    },
+    {
+      "epoch": 0.823442900213251,
+      "grad_norm": 0.5648274421691895,
+      "learning_rate": 1.5921500079162276e-05,
+      "loss": 0.3616,
+      "step": 11198
+    },
+    {
+      "epoch": 0.8235164350319877,
+      "grad_norm": 0.45968955755233765,
+      "learning_rate": 1.590860911660873e-05,
+      "loss": 0.3613,
+      "step": 11199
+    },
+    {
+      "epoch": 0.8235899698507243,
+      "grad_norm": 0.6382176280021667,
+      "learning_rate": 1.5895722923807632e-05,
+      "loss": 0.3568,
+      "step": 11200
+    },
+    {
+      "epoch": 0.823663504669461,
+      "grad_norm": 0.6318148970603943,
+      "learning_rate": 1.5882841501489863e-05,
+      "loss": 0.3683,
+      "step": 11201
+    },
+    {
+      "epoch": 0.8237370394881977,
+      "grad_norm": 0.4579249322414398,
+      "learning_rate": 1.58699648503861e-05,
+      "loss": 0.3548,
+      "step": 11202
+    },
+    {
+      "epoch": 0.8238105743069344,
+      "grad_norm": 0.7595580816268921,
+      "learning_rate": 1.5857092971226706e-05,
+      "loss": 0.3638,
+      "step": 11203
+    },
+    {
+      "epoch": 0.823884109125671,
+      "grad_norm": 0.4363320767879486,
+      "learning_rate": 1.5844225864741823e-05,
+      "loss": 0.3278,
+      "step": 11204
+    },
+    {
+      "epoch": 0.8239576439444076,
+      "grad_norm": 0.5734108686447144,
+      "learning_rate": 1.5831363531661247e-05,
+      "loss": 0.3413,
+      "step": 11205
+    },
+    {
+      "epoch": 0.8240311787631444,
+      "grad_norm": 0.48553481698036194,
+      "learning_rate": 1.581850597271456e-05,
+      "loss": 0.3572,
+      "step": 11206
+    },
+    {
+      "epoch": 0.824104713581881,
+      "grad_norm": 0.475483775138855,
+      "learning_rate": 1.5805653188631064e-05,
+      "loss": 0.3507,
+      "step": 11207
+    },
+    {
+      "epoch": 0.8241782484006177,
+      "grad_norm": 0.3955496847629547,
+      "learning_rate": 1.5792805180139757e-05,
+      "loss": 0.3419,
+      "step": 11208
+    },
+    {
+      "epoch": 0.8242517832193543,
+      "grad_norm": 0.507948637008667,
+      "learning_rate": 1.5779961947969413e-05,
+      "loss": 0.3535,
+      "step": 11209
+    },
+    {
+      "epoch": 0.8243253180380911,
+      "grad_norm": 0.4806073009967804,
+      "learning_rate": 1.5767123492848522e-05,
+      "loss": 0.3389,
+      "step": 11210
+    },
+    {
+      "epoch": 0.8243988528568277,
+      "grad_norm": 0.4583953619003296,
+      "learning_rate": 1.5754289815505286e-05,
+      "loss": 0.3415,
+      "step": 11211
+    },
+    {
+      "epoch": 0.8244723876755644,
+      "grad_norm": 0.4011795222759247,
+      "learning_rate": 1.5741460916667616e-05,
+      "loss": 0.3656,
+      "step": 11212
+    },
+    {
+      "epoch": 0.824545922494301,
+      "grad_norm": 0.39676257967948914,
+      "learning_rate": 1.57286367970632e-05,
+      "loss": 0.3465,
+      "step": 11213
+    },
+    {
+      "epoch": 0.8246194573130378,
+      "grad_norm": 0.4160221517086029,
+      "learning_rate": 1.5715817457419468e-05,
+      "loss": 0.3306,
+      "step": 11214
+    },
+    {
+      "epoch": 0.8246929921317744,
+      "grad_norm": 0.4265352487564087,
+      "learning_rate": 1.5703002898463504e-05,
+      "loss": 0.3098,
+      "step": 11215
+    },
+    {
+      "epoch": 0.8247665269505111,
+      "grad_norm": 0.4620455205440521,
+      "learning_rate": 1.5690193120922158e-05,
+      "loss": 0.3377,
+      "step": 11216
+    },
+    {
+      "epoch": 0.8248400617692477,
+      "grad_norm": 0.4182717800140381,
+      "learning_rate": 1.5677388125522074e-05,
+      "loss": 0.3318,
+      "step": 11217
+    },
+    {
+      "epoch": 0.8249135965879845,
+      "grad_norm": 0.5944543480873108,
+      "learning_rate": 1.5664587912989492e-05,
+      "loss": 0.347,
+      "step": 11218
+    },
+    {
+      "epoch": 0.8249871314067211,
+      "grad_norm": 0.5007476210594177,
+      "learning_rate": 1.5651792484050486e-05,
+      "loss": 0.3805,
+      "step": 11219
+    },
+    {
+      "epoch": 0.8250606662254577,
+      "grad_norm": 0.535198450088501,
+      "learning_rate": 1.5639001839430857e-05,
+      "loss": 0.3836,
+      "step": 11220
+    },
+    {
+      "epoch": 0.8251342010441944,
+      "grad_norm": 0.41611406207084656,
+      "learning_rate": 1.5626215979856052e-05,
+      "loss": 0.3757,
+      "step": 11221
+    },
+    {
+      "epoch": 0.8252077358629311,
+      "grad_norm": 0.3651052713394165,
+      "learning_rate": 1.5613434906051304e-05,
+      "loss": 0.3182,
+      "step": 11222
+    },
+    {
+      "epoch": 0.8252812706816678,
+      "grad_norm": 0.5316131114959717,
+      "learning_rate": 1.5600658618741647e-05,
+      "loss": 0.3414,
+      "step": 11223
+    },
+    {
+      "epoch": 0.8253548055004044,
+      "grad_norm": 0.38723182678222656,
+      "learning_rate": 1.5587887118651655e-05,
+      "loss": 0.3471,
+      "step": 11224
+    },
+    {
+      "epoch": 0.8254283403191411,
+      "grad_norm": 0.4088706374168396,
+      "learning_rate": 1.5575120406505783e-05,
+      "loss": 0.338,
+      "step": 11225
+    },
+    {
+      "epoch": 0.8255018751378778,
+      "grad_norm": 0.4233280420303345,
+      "learning_rate": 1.5562358483028182e-05,
+      "loss": 0.3281,
+      "step": 11226
+    },
+    {
+      "epoch": 0.8255754099566145,
+      "grad_norm": 0.47682467103004456,
+      "learning_rate": 1.5549601348942745e-05,
+      "loss": 0.3485,
+      "step": 11227
+    },
+    {
+      "epoch": 0.8256489447753511,
+      "grad_norm": 0.4124368131160736,
+      "learning_rate": 1.553684900497301e-05,
+      "loss": 0.3231,
+      "step": 11228
+    },
+    {
+      "epoch": 0.8257224795940878,
+      "grad_norm": 0.44670188426971436,
+      "learning_rate": 1.552410145184234e-05,
+      "loss": 0.3413,
+      "step": 11229
+    },
+    {
+      "epoch": 0.8257960144128245,
+      "grad_norm": 0.3973243832588196,
+      "learning_rate": 1.5511358690273814e-05,
+      "loss": 0.3478,
+      "step": 11230
+    },
+    {
+      "epoch": 0.8258695492315612,
+      "grad_norm": 0.44096115231513977,
+      "learning_rate": 1.5498620720990164e-05,
+      "loss": 0.3253,
+      "step": 11231
+    },
+    {
+      "epoch": 0.8259430840502978,
+      "grad_norm": 0.41050228476524353,
+      "learning_rate": 1.5485887544713927e-05,
+      "loss": 0.3489,
+      "step": 11232
+    },
+    {
+      "epoch": 0.8260166188690344,
+      "grad_norm": 0.4547068476676941,
+      "learning_rate": 1.5473159162167362e-05,
+      "loss": 0.3484,
+      "step": 11233
+    },
+    {
+      "epoch": 0.8260901536877712,
+      "grad_norm": 0.37847429513931274,
+      "learning_rate": 1.5460435574072386e-05,
+      "loss": 0.3206,
+      "step": 11234
+    },
+    {
+      "epoch": 0.8261636885065078,
+      "grad_norm": 0.5008982419967651,
+      "learning_rate": 1.5447716781150747e-05,
+      "loss": 0.3497,
+      "step": 11235
+    },
+    {
+      "epoch": 0.8262372233252445,
+      "grad_norm": 0.443997323513031,
+      "learning_rate": 1.543500278412382e-05,
+      "loss": 0.3569,
+      "step": 11236
+    },
+    {
+      "epoch": 0.8263107581439811,
+      "grad_norm": 0.37622368335723877,
+      "learning_rate": 1.542229358371281e-05,
+      "loss": 0.3246,
+      "step": 11237
+    },
+    {
+      "epoch": 0.8263842929627179,
+      "grad_norm": 0.3809570074081421,
+      "learning_rate": 1.540958918063853e-05,
+      "loss": 0.3383,
+      "step": 11238
+    },
+    {
+      "epoch": 0.8264578277814545,
+      "grad_norm": 0.4448092579841614,
+      "learning_rate": 1.539688957562163e-05,
+      "loss": 0.342,
+      "step": 11239
+    },
+    {
+      "epoch": 0.8265313626001912,
+      "grad_norm": 0.3914278447628021,
+      "learning_rate": 1.5384194769382467e-05,
+      "loss": 0.3256,
+      "step": 11240
+    },
+    {
+      "epoch": 0.8266048974189278,
+      "grad_norm": 0.38537219166755676,
+      "learning_rate": 1.5371504762641053e-05,
+      "loss": 0.3336,
+      "step": 11241
+    },
+    {
+      "epoch": 0.8266784322376646,
+      "grad_norm": 0.43223586678504944,
+      "learning_rate": 1.535881955611719e-05,
+      "loss": 0.3445,
+      "step": 11242
+    },
+    {
+      "epoch": 0.8267519670564012,
+      "grad_norm": 0.38925573229789734,
+      "learning_rate": 1.5346139150530435e-05,
+      "loss": 0.3291,
+      "step": 11243
+    },
+    {
+      "epoch": 0.8268255018751379,
+      "grad_norm": 0.45099881291389465,
+      "learning_rate": 1.533346354659999e-05,
+      "loss": 0.3342,
+      "step": 11244
+    },
+    {
+      "epoch": 0.8268990366938745,
+      "grad_norm": 0.39092808961868286,
+      "learning_rate": 1.5320792745044833e-05,
+      "loss": 0.347,
+      "step": 11245
+    },
+    {
+      "epoch": 0.8269725715126113,
+      "grad_norm": 0.41925525665283203,
+      "learning_rate": 1.5308126746583728e-05,
+      "loss": 0.359,
+      "step": 11246
+    },
+    {
+      "epoch": 0.8270461063313479,
+      "grad_norm": 0.5125086307525635,
+      "learning_rate": 1.5295465551935017e-05,
+      "loss": 0.3259,
+      "step": 11247
+    },
+    {
+      "epoch": 0.8271196411500845,
+      "grad_norm": 0.417959600687027,
+      "learning_rate": 1.528280916181688e-05,
+      "loss": 0.323,
+      "step": 11248
+    },
+    {
+      "epoch": 0.8271931759688212,
+      "grad_norm": 0.3809225559234619,
+      "learning_rate": 1.527015757694722e-05,
+      "loss": 0.3333,
+      "step": 11249
+    },
+    {
+      "epoch": 0.8272667107875579,
+      "grad_norm": 0.46491795778274536,
+      "learning_rate": 1.5257510798043662e-05,
+      "loss": 0.3355,
+      "step": 11250
+    },
+    {
+      "epoch": 0.8273402456062946,
+      "grad_norm": 0.44957661628723145,
+      "learning_rate": 1.5244868825823489e-05,
+      "loss": 0.3328,
+      "step": 11251
+    },
+    {
+      "epoch": 0.8274137804250312,
+      "grad_norm": 0.38519519567489624,
+      "learning_rate": 1.5232231661003815e-05,
+      "loss": 0.3542,
+      "step": 11252
+    },
+    {
+      "epoch": 0.827487315243768,
+      "grad_norm": 0.5290947556495667,
+      "learning_rate": 1.5219599304301446e-05,
+      "loss": 0.3339,
+      "step": 11253
+    },
+    {
+      "epoch": 0.8275608500625046,
+      "grad_norm": 0.5762312412261963,
+      "learning_rate": 1.5206971756432842e-05,
+      "loss": 0.3355,
+      "step": 11254
+    },
+    {
+      "epoch": 0.8276343848812413,
+      "grad_norm": 0.41766270995140076,
+      "learning_rate": 1.5194349018114296e-05,
+      "loss": 0.3274,
+      "step": 11255
+    },
+    {
+      "epoch": 0.8277079196999779,
+      "grad_norm": 0.5126779675483704,
+      "learning_rate": 1.518173109006179e-05,
+      "loss": 0.3547,
+      "step": 11256
+    },
+    {
+      "epoch": 0.8277814545187147,
+      "grad_norm": 0.5379419922828674,
+      "learning_rate": 1.5169117972990987e-05,
+      "loss": 0.3335,
+      "step": 11257
+    },
+    {
+      "epoch": 0.8278549893374513,
+      "grad_norm": 0.4051291346549988,
+      "learning_rate": 1.5156509667617358e-05,
+      "loss": 0.3137,
+      "step": 11258
+    },
+    {
+      "epoch": 0.827928524156188,
+      "grad_norm": 0.49245399236679077,
+      "learning_rate": 1.5143906174656019e-05,
+      "loss": 0.3202,
+      "step": 11259
+    },
+    {
+      "epoch": 0.8280020589749246,
+      "grad_norm": 0.3781643807888031,
+      "learning_rate": 1.51313074948219e-05,
+      "loss": 0.3277,
+      "step": 11260
+    },
+    {
+      "epoch": 0.8280755937936612,
+      "grad_norm": 0.4745306372642517,
+      "learning_rate": 1.5118713628829561e-05,
+      "loss": 0.3263,
+      "step": 11261
+    },
+    {
+      "epoch": 0.828149128612398,
+      "grad_norm": 0.45390358567237854,
+      "learning_rate": 1.5106124577393366e-05,
+      "loss": 0.3323,
+      "step": 11262
+    },
+    {
+      "epoch": 0.8282226634311346,
+      "grad_norm": 0.3976733684539795,
+      "learning_rate": 1.5093540341227397e-05,
+      "loss": 0.3318,
+      "step": 11263
+    },
+    {
+      "epoch": 0.8282961982498713,
+      "grad_norm": 0.5660276412963867,
+      "learning_rate": 1.5080960921045394e-05,
+      "loss": 0.338,
+      "step": 11264
+    },
+    {
+      "epoch": 0.828369733068608,
+      "grad_norm": 0.6043781638145447,
+      "learning_rate": 1.5068386317560924e-05,
+      "loss": 0.3481,
+      "step": 11265
+    },
+    {
+      "epoch": 0.8284432678873447,
+      "grad_norm": 0.42854222655296326,
+      "learning_rate": 1.505581653148722e-05,
+      "loss": 0.3447,
+      "step": 11266
+    },
+    {
+      "epoch": 0.8285168027060813,
+      "grad_norm": 0.7710435390472412,
+      "learning_rate": 1.504325156353723e-05,
+      "loss": 0.3415,
+      "step": 11267
+    },
+    {
+      "epoch": 0.828590337524818,
+      "grad_norm": 0.515830397605896,
+      "learning_rate": 1.503069141442368e-05,
+      "loss": 0.3271,
+      "step": 11268
+    },
+    {
+      "epoch": 0.8286638723435547,
+      "grad_norm": 0.4669438302516937,
+      "learning_rate": 1.501813608485899e-05,
+      "loss": 0.3212,
+      "step": 11269
+    },
+    {
+      "epoch": 0.8287374071622914,
+      "grad_norm": 0.7955787777900696,
+      "learning_rate": 1.5005585575555325e-05,
+      "loss": 0.3355,
+      "step": 11270
+    },
+    {
+      "epoch": 0.828810941981028,
+      "grad_norm": 0.4751366078853607,
+      "learning_rate": 1.4993039887224503e-05,
+      "loss": 0.328,
+      "step": 11271
+    },
+    {
+      "epoch": 0.8288844767997647,
+      "grad_norm": 0.48179954290390015,
+      "learning_rate": 1.4980499020578187e-05,
+      "loss": 0.3467,
+      "step": 11272
+    },
+    {
+      "epoch": 0.8289580116185014,
+      "grad_norm": 0.7440222501754761,
+      "learning_rate": 1.49679629763277e-05,
+      "loss": 0.3112,
+      "step": 11273
+    },
+    {
+      "epoch": 0.8290315464372381,
+      "grad_norm": 0.577435314655304,
+      "learning_rate": 1.495543175518408e-05,
+      "loss": 0.3134,
+      "step": 11274
+    },
+    {
+      "epoch": 0.8291050812559747,
+      "grad_norm": 0.7969591617584229,
+      "learning_rate": 1.4942905357858118e-05,
+      "loss": 0.3213,
+      "step": 11275
+    },
+    {
+      "epoch": 0.8291786160747113,
+      "grad_norm": 0.8564674854278564,
+      "learning_rate": 1.4930383785060343e-05,
+      "loss": 0.3431,
+      "step": 11276
+    },
+    {
+      "epoch": 0.8292521508934481,
+      "grad_norm": 0.40485256910324097,
+      "learning_rate": 1.4917867037500965e-05,
+      "loss": 0.349,
+      "step": 11277
+    },
+    {
+      "epoch": 0.8293256857121847,
+      "grad_norm": 1.0047528743743896,
+      "learning_rate": 1.4905355115889952e-05,
+      "loss": 0.329,
+      "step": 11278
+    },
+    {
+      "epoch": 0.8293992205309214,
+      "grad_norm": 0.5431396961212158,
+      "learning_rate": 1.4892848020937022e-05,
+      "loss": 0.3151,
+      "step": 11279
+    },
+    {
+      "epoch": 0.829472755349658,
+      "grad_norm": 0.4404856860637665,
+      "learning_rate": 1.4880345753351554e-05,
+      "loss": 0.3414,
+      "step": 11280
+    },
+    {
+      "epoch": 0.8295462901683948,
+      "grad_norm": 0.8626044392585754,
+      "learning_rate": 1.4867848313842725e-05,
+      "loss": 0.3425,
+      "step": 11281
+    },
+    {
+      "epoch": 0.8296198249871314,
+      "grad_norm": 0.5814406275749207,
+      "learning_rate": 1.4855355703119356e-05,
+      "loss": 0.3325,
+      "step": 11282
+    },
+    {
+      "epoch": 0.8296933598058681,
+      "grad_norm": 0.6886680126190186,
+      "learning_rate": 1.4842867921890102e-05,
+      "loss": 0.3497,
+      "step": 11283
+    },
+    {
+      "epoch": 0.8297668946246047,
+      "grad_norm": 0.7043771743774414,
+      "learning_rate": 1.483038497086322e-05,
+      "loss": 0.3492,
+      "step": 11284
+    },
+    {
+      "epoch": 0.8298404294433415,
+      "grad_norm": 0.4020889699459076,
+      "learning_rate": 1.4817906850746788e-05,
+      "loss": 0.3365,
+      "step": 11285
+    },
+    {
+      "epoch": 0.8299139642620781,
+      "grad_norm": 0.5285816788673401,
+      "learning_rate": 1.4805433562248594e-05,
+      "loss": 0.3475,
+      "step": 11286
+    },
+    {
+      "epoch": 0.8299874990808148,
+      "grad_norm": 0.4258135259151459,
+      "learning_rate": 1.4792965106076096e-05,
+      "loss": 0.3451,
+      "step": 11287
+    },
+    {
+      "epoch": 0.8300610338995514,
+      "grad_norm": 0.42733487486839294,
+      "learning_rate": 1.4780501482936538e-05,
+      "loss": 0.3151,
+      "step": 11288
+    },
+    {
+      "epoch": 0.8301345687182881,
+      "grad_norm": 0.4382658898830414,
+      "learning_rate": 1.4768042693536899e-05,
+      "loss": 0.3371,
+      "step": 11289
+    },
+    {
+      "epoch": 0.8302081035370248,
+      "grad_norm": 0.3992336690425873,
+      "learning_rate": 1.4755588738583802e-05,
+      "loss": 0.3223,
+      "step": 11290
+    },
+    {
+      "epoch": 0.8302816383557614,
+      "grad_norm": 0.40064507722854614,
+      "learning_rate": 1.4743139618783675e-05,
+      "loss": 0.3268,
+      "step": 11291
+    },
+    {
+      "epoch": 0.8303551731744981,
+      "grad_norm": 0.40676236152648926,
+      "learning_rate": 1.473069533484267e-05,
+      "loss": 0.3573,
+      "step": 11292
+    },
+    {
+      "epoch": 0.8304287079932348,
+      "grad_norm": 0.48495399951934814,
+      "learning_rate": 1.4718255887466603e-05,
+      "loss": 0.3404,
+      "step": 11293
+    },
+    {
+      "epoch": 0.8305022428119715,
+      "grad_norm": 0.47825533151626587,
+      "learning_rate": 1.4705821277361054e-05,
+      "loss": 0.3286,
+      "step": 11294
+    },
+    {
+      "epoch": 0.8305757776307081,
+      "grad_norm": 0.5683743953704834,
+      "learning_rate": 1.4693391505231324e-05,
+      "loss": 0.337,
+      "step": 11295
+    },
+    {
+      "epoch": 0.8306493124494448,
+      "grad_norm": 0.6055570244789124,
+      "learning_rate": 1.4680966571782484e-05,
+      "loss": 0.3576,
+      "step": 11296
+    },
+    {
+      "epoch": 0.8307228472681815,
+      "grad_norm": 0.5807060599327087,
+      "learning_rate": 1.4668546477719246e-05,
+      "loss": 0.3454,
+      "step": 11297
+    },
+    {
+      "epoch": 0.8307963820869182,
+      "grad_norm": 0.4311404526233673,
+      "learning_rate": 1.4656131223746094e-05,
+      "loss": 0.3252,
+      "step": 11298
+    },
+    {
+      "epoch": 0.8308699169056548,
+      "grad_norm": 0.44875115156173706,
+      "learning_rate": 1.4643720810567274e-05,
+      "loss": 0.3339,
+      "step": 11299
+    },
+    {
+      "epoch": 0.8309434517243915,
+      "grad_norm": 0.45564529299736023,
+      "learning_rate": 1.4631315238886667e-05,
+      "loss": 0.3465,
+      "step": 11300
+    },
+    {
+      "epoch": 0.8310169865431282,
+      "grad_norm": 0.37694376707077026,
+      "learning_rate": 1.4618914509407956e-05,
+      "loss": 0.3718,
+      "step": 11301
+    },
+    {
+      "epoch": 0.8310905213618649,
+      "grad_norm": 0.4383949637413025,
+      "learning_rate": 1.4606518622834531e-05,
+      "loss": 0.345,
+      "step": 11302
+    },
+    {
+      "epoch": 0.8311640561806015,
+      "grad_norm": 0.4895568788051605,
+      "learning_rate": 1.4594127579869476e-05,
+      "loss": 0.3469,
+      "step": 11303
+    },
+    {
+      "epoch": 0.8312375909993381,
+      "grad_norm": 0.3474035859107971,
+      "learning_rate": 1.458174138121563e-05,
+      "loss": 0.3579,
+      "step": 11304
+    },
+    {
+      "epoch": 0.8313111258180749,
+      "grad_norm": 0.41523224115371704,
+      "learning_rate": 1.4569360027575597e-05,
+      "loss": 0.3705,
+      "step": 11305
+    },
+    {
+      "epoch": 0.8313846606368115,
+      "grad_norm": 0.41653114557266235,
+      "learning_rate": 1.4556983519651601e-05,
+      "loss": 0.3473,
+      "step": 11306
+    },
+    {
+      "epoch": 0.8314581954555482,
+      "grad_norm": 0.3962634801864624,
+      "learning_rate": 1.4544611858145662e-05,
+      "loss": 0.3242,
+      "step": 11307
+    },
+    {
+      "epoch": 0.8315317302742848,
+      "grad_norm": 0.4730844497680664,
+      "learning_rate": 1.4532245043759519e-05,
+      "loss": 0.3603,
+      "step": 11308
+    },
+    {
+      "epoch": 0.8316052650930216,
+      "grad_norm": 0.4086093008518219,
+      "learning_rate": 1.4519883077194663e-05,
+      "loss": 0.356,
+      "step": 11309
+    },
+    {
+      "epoch": 0.8316787999117582,
+      "grad_norm": 0.3935164511203766,
+      "learning_rate": 1.4507525959152223e-05,
+      "loss": 0.3258,
+      "step": 11310
+    },
+    {
+      "epoch": 0.8317523347304949,
+      "grad_norm": 0.40279313921928406,
+      "learning_rate": 1.449517369033313e-05,
+      "loss": 0.3382,
+      "step": 11311
+    },
+    {
+      "epoch": 0.8318258695492315,
+      "grad_norm": 0.5891317129135132,
+      "learning_rate": 1.4482826271438043e-05,
+      "loss": 0.3483,
+      "step": 11312
+    },
+    {
+      "epoch": 0.8318994043679683,
+      "grad_norm": 0.40651074051856995,
+      "learning_rate": 1.4470483703167281e-05,
+      "loss": 0.3549,
+      "step": 11313
+    },
+    {
+      "epoch": 0.8319729391867049,
+      "grad_norm": 0.523239016532898,
+      "learning_rate": 1.445814598622095e-05,
+      "loss": 0.3431,
+      "step": 11314
+    },
+    {
+      "epoch": 0.8320464740054416,
+      "grad_norm": 0.7009807825088501,
+      "learning_rate": 1.444581312129888e-05,
+      "loss": 0.3503,
+      "step": 11315
+    },
+    {
+      "epoch": 0.8321200088241782,
+      "grad_norm": 0.451742947101593,
+      "learning_rate": 1.4433485109100575e-05,
+      "loss": 0.328,
+      "step": 11316
+    },
+    {
+      "epoch": 0.8321935436429149,
+      "grad_norm": 0.47632279992103577,
+      "learning_rate": 1.4421161950325268e-05,
+      "loss": 0.3514,
+      "step": 11317
+    },
+    {
+      "epoch": 0.8322670784616516,
+      "grad_norm": 0.5562422871589661,
+      "learning_rate": 1.4408843645671976e-05,
+      "loss": 0.3277,
+      "step": 11318
+    },
+    {
+      "epoch": 0.8323406132803882,
+      "grad_norm": 0.498769611120224,
+      "learning_rate": 1.4396530195839419e-05,
+      "loss": 0.3433,
+      "step": 11319
+    },
+    {
+      "epoch": 0.832414148099125,
+      "grad_norm": 0.4318583011627197,
+      "learning_rate": 1.4384221601525993e-05,
+      "loss": 0.3338,
+      "step": 11320
+    },
+    {
+      "epoch": 0.8324876829178616,
+      "grad_norm": 0.6582991480827332,
+      "learning_rate": 1.4371917863429862e-05,
+      "loss": 0.3432,
+      "step": 11321
+    },
+    {
+      "epoch": 0.8325612177365983,
+      "grad_norm": 0.4597577750682831,
+      "learning_rate": 1.4359618982248957e-05,
+      "loss": 0.346,
+      "step": 11322
+    },
+    {
+      "epoch": 0.8326347525553349,
+      "grad_norm": 0.47852644324302673,
+      "learning_rate": 1.4347324958680809e-05,
+      "loss": 0.3446,
+      "step": 11323
+    },
+    {
+      "epoch": 0.8327082873740717,
+      "grad_norm": 0.6307781338691711,
+      "learning_rate": 1.433503579342278e-05,
+      "loss": 0.3567,
+      "step": 11324
+    },
+    {
+      "epoch": 0.8327818221928083,
+      "grad_norm": 0.4353782534599304,
+      "learning_rate": 1.4322751487171949e-05,
+      "loss": 0.3358,
+      "step": 11325
+    },
+    {
+      "epoch": 0.832855357011545,
+      "grad_norm": 0.6790350079536438,
+      "learning_rate": 1.4310472040625055e-05,
+      "loss": 0.3675,
+      "step": 11326
+    },
+    {
+      "epoch": 0.8329288918302816,
+      "grad_norm": 0.4322991967201233,
+      "learning_rate": 1.4298197454478613e-05,
+      "loss": 0.3646,
+      "step": 11327
+    },
+    {
+      "epoch": 0.8330024266490184,
+      "grad_norm": 0.4863453209400177,
+      "learning_rate": 1.4285927729428904e-05,
+      "loss": 0.3429,
+      "step": 11328
+    },
+    {
+      "epoch": 0.833075961467755,
+      "grad_norm": 0.6958715915679932,
+      "learning_rate": 1.4273662866171788e-05,
+      "loss": 0.3398,
+      "step": 11329
+    },
+    {
+      "epoch": 0.8331494962864917,
+      "grad_norm": 0.4329492151737213,
+      "learning_rate": 1.4261402865402984e-05,
+      "loss": 0.3655,
+      "step": 11330
+    },
+    {
+      "epoch": 0.8332230311052283,
+      "grad_norm": 0.7924992442131042,
+      "learning_rate": 1.4249147727817902e-05,
+      "loss": 0.3368,
+      "step": 11331
+    },
+    {
+      "epoch": 0.833296565923965,
+      "grad_norm": 0.542705774307251,
+      "learning_rate": 1.4236897454111664e-05,
+      "loss": 0.3616,
+      "step": 11332
+    },
+    {
+      "epoch": 0.8333701007427017,
+      "grad_norm": 0.5784003734588623,
+      "learning_rate": 1.4224652044979103e-05,
+      "loss": 0.3454,
+      "step": 11333
+    },
+    {
+      "epoch": 0.8334436355614383,
+      "grad_norm": 0.8168320059776306,
+      "learning_rate": 1.4212411501114786e-05,
+      "loss": 0.3871,
+      "step": 11334
+    },
+    {
+      "epoch": 0.833517170380175,
+      "grad_norm": 0.43909329175949097,
+      "learning_rate": 1.4200175823213058e-05,
+      "loss": 0.353,
+      "step": 11335
+    },
+    {
+      "epoch": 0.8335907051989117,
+      "grad_norm": 0.8312194347381592,
+      "learning_rate": 1.4187945011967884e-05,
+      "loss": 0.3555,
+      "step": 11336
+    },
+    {
+      "epoch": 0.8336642400176484,
+      "grad_norm": 0.45733267068862915,
+      "learning_rate": 1.4175719068073023e-05,
+      "loss": 0.3372,
+      "step": 11337
+    },
+    {
+      "epoch": 0.833737774836385,
+      "grad_norm": 0.6373110413551331,
+      "learning_rate": 1.4163497992221975e-05,
+      "loss": 0.3365,
+      "step": 11338
+    },
+    {
+      "epoch": 0.8338113096551217,
+      "grad_norm": 0.8302019834518433,
+      "learning_rate": 1.4151281785107884e-05,
+      "loss": 0.3601,
+      "step": 11339
+    },
+    {
+      "epoch": 0.8338848444738584,
+      "grad_norm": 0.426247775554657,
+      "learning_rate": 1.4139070447423709e-05,
+      "loss": 0.3345,
+      "step": 11340
+    },
+    {
+      "epoch": 0.8339583792925951,
+      "grad_norm": 0.8455081582069397,
+      "learning_rate": 1.412686397986206e-05,
+      "loss": 0.3646,
+      "step": 11341
+    },
+    {
+      "epoch": 0.8340319141113317,
+      "grad_norm": 0.6394001841545105,
+      "learning_rate": 1.411466238311533e-05,
+      "loss": 0.3481,
+      "step": 11342
+    },
+    {
+      "epoch": 0.8341054489300684,
+      "grad_norm": 0.5389865040779114,
+      "learning_rate": 1.4102465657875552e-05,
+      "loss": 0.3614,
+      "step": 11343
+    },
+    {
+      "epoch": 0.8341789837488051,
+      "grad_norm": 0.7310149073600769,
+      "learning_rate": 1.4090273804834586e-05,
+      "loss": 0.3785,
+      "step": 11344
+    },
+    {
+      "epoch": 0.8342525185675417,
+      "grad_norm": 0.4514445960521698,
+      "learning_rate": 1.4078086824683978e-05,
+      "loss": 0.3398,
+      "step": 11345
+    },
+    {
+      "epoch": 0.8343260533862784,
+      "grad_norm": 0.5177509784698486,
+      "learning_rate": 1.406590471811493e-05,
+      "loss": 0.3766,
+      "step": 11346
+    },
+    {
+      "epoch": 0.834399588205015,
+      "grad_norm": 0.544063150882721,
+      "learning_rate": 1.4053727485818469e-05,
+      "loss": 0.3695,
+      "step": 11347
+    },
+    {
+      "epoch": 0.8344731230237518,
+      "grad_norm": 0.4730186462402344,
+      "learning_rate": 1.4041555128485306e-05,
+      "loss": 0.3592,
+      "step": 11348
+    },
+    {
+      "epoch": 0.8345466578424884,
+      "grad_norm": 0.4494360089302063,
+      "learning_rate": 1.4029387646805825e-05,
+      "loss": 0.3527,
+      "step": 11349
+    },
+    {
+      "epoch": 0.8346201926612251,
+      "grad_norm": 0.6067209839820862,
+      "learning_rate": 1.4017225041470205e-05,
+      "loss": 0.3382,
+      "step": 11350
+    },
+    {
+      "epoch": 0.8346937274799617,
+      "grad_norm": 0.3822944462299347,
+      "learning_rate": 1.4005067313168374e-05,
+      "loss": 0.3662,
+      "step": 11351
+    },
+    {
+      "epoch": 0.8347672622986985,
+      "grad_norm": 0.70224529504776,
+      "learning_rate": 1.3992914462589823e-05,
+      "loss": 0.3352,
+      "step": 11352
+    },
+    {
+      "epoch": 0.8348407971174351,
+      "grad_norm": 0.4763491153717041,
+      "learning_rate": 1.3980766490423935e-05,
+      "loss": 0.355,
+      "step": 11353
+    },
+    {
+      "epoch": 0.8349143319361718,
+      "grad_norm": 0.5666704177856445,
+      "learning_rate": 1.3968623397359759e-05,
+      "loss": 0.3263,
+      "step": 11354
+    },
+    {
+      "epoch": 0.8349878667549084,
+      "grad_norm": 0.5635077953338623,
+      "learning_rate": 1.3956485184086065e-05,
+      "loss": 0.3471,
+      "step": 11355
+    },
+    {
+      "epoch": 0.8350614015736452,
+      "grad_norm": 0.43700483441352844,
+      "learning_rate": 1.3944351851291327e-05,
+      "loss": 0.3446,
+      "step": 11356
+    },
+    {
+      "epoch": 0.8351349363923818,
+      "grad_norm": 0.41940876841545105,
+      "learning_rate": 1.3932223399663768e-05,
+      "loss": 0.3468,
+      "step": 11357
+    },
+    {
+      "epoch": 0.8352084712111185,
+      "grad_norm": 0.4760427176952362,
+      "learning_rate": 1.3920099829891341e-05,
+      "loss": 0.3338,
+      "step": 11358
+    },
+    {
+      "epoch": 0.8352820060298551,
+      "grad_norm": 0.5478196740150452,
+      "learning_rate": 1.3907981142661686e-05,
+      "loss": 0.3505,
+      "step": 11359
+    },
+    {
+      "epoch": 0.8353555408485918,
+      "grad_norm": 0.4392942488193512,
+      "learning_rate": 1.3895867338662183e-05,
+      "loss": 0.3348,
+      "step": 11360
+    },
+    {
+      "epoch": 0.8354290756673285,
+      "grad_norm": 0.39291590452194214,
+      "learning_rate": 1.3883758418579995e-05,
+      "loss": 0.3687,
+      "step": 11361
+    },
+    {
+      "epoch": 0.8355026104860651,
+      "grad_norm": 0.5490813851356506,
+      "learning_rate": 1.3871654383101873e-05,
+      "loss": 0.3695,
+      "step": 11362
+    },
+    {
+      "epoch": 0.8355761453048018,
+      "grad_norm": 0.4712924659252167,
+      "learning_rate": 1.3859555232914434e-05,
+      "loss": 0.3426,
+      "step": 11363
+    },
+    {
+      "epoch": 0.8356496801235385,
+      "grad_norm": 0.40685737133026123,
+      "learning_rate": 1.3847460968703917e-05,
+      "loss": 0.388,
+      "step": 11364
+    },
+    {
+      "epoch": 0.8357232149422752,
+      "grad_norm": 0.40407994389533997,
+      "learning_rate": 1.3835371591156342e-05,
+      "loss": 0.3458,
+      "step": 11365
+    },
+    {
+      "epoch": 0.8357967497610118,
+      "grad_norm": 0.4715094268321991,
+      "learning_rate": 1.3823287100957416e-05,
+      "loss": 0.3639,
+      "step": 11366
+    },
+    {
+      "epoch": 0.8358702845797485,
+      "grad_norm": 0.38039758801460266,
+      "learning_rate": 1.3811207498792588e-05,
+      "loss": 0.3694,
+      "step": 11367
+    },
+    {
+      "epoch": 0.8359438193984852,
+      "grad_norm": 0.42863887548446655,
+      "learning_rate": 1.3799132785347047e-05,
+      "loss": 0.3564,
+      "step": 11368
+    },
+    {
+      "epoch": 0.8360173542172219,
+      "grad_norm": 0.5748140215873718,
+      "learning_rate": 1.3787062961305642e-05,
+      "loss": 0.3624,
+      "step": 11369
+    },
+    {
+      "epoch": 0.8360908890359585,
+      "grad_norm": 0.4116644561290741,
+      "learning_rate": 1.3774998027353014e-05,
+      "loss": 0.349,
+      "step": 11370
+    },
+    {
+      "epoch": 0.8361644238546952,
+      "grad_norm": 0.3944898247718811,
+      "learning_rate": 1.3762937984173519e-05,
+      "loss": 0.3781,
+      "step": 11371
+    },
+    {
+      "epoch": 0.8362379586734319,
+      "grad_norm": 0.394394189119339,
+      "learning_rate": 1.3750882832451161e-05,
+      "loss": 0.3359,
+      "step": 11372
+    },
+    {
+      "epoch": 0.8363114934921685,
+      "grad_norm": 0.5282060503959656,
+      "learning_rate": 1.3738832572869753e-05,
+      "loss": 0.4056,
+      "step": 11373
+    },
+    {
+      "epoch": 0.8363850283109052,
+      "grad_norm": 0.43237873911857605,
+      "learning_rate": 1.3726787206112823e-05,
+      "loss": 0.3552,
+      "step": 11374
+    },
+    {
+      "epoch": 0.8364585631296418,
+      "grad_norm": 0.4012094736099243,
+      "learning_rate": 1.371474673286357e-05,
+      "loss": 0.3595,
+      "step": 11375
+    },
+    {
+      "epoch": 0.8365320979483786,
+      "grad_norm": 0.4389244019985199,
+      "learning_rate": 1.3702711153804925e-05,
+      "loss": 0.3763,
+      "step": 11376
+    },
+    {
+      "epoch": 0.8366056327671152,
+      "grad_norm": 0.405712753534317,
+      "learning_rate": 1.3690680469619565e-05,
+      "loss": 0.3506,
+      "step": 11377
+    },
+    {
+      "epoch": 0.8366791675858519,
+      "grad_norm": 0.431721955537796,
+      "learning_rate": 1.3678654680989921e-05,
+      "loss": 0.3884,
+      "step": 11378
+    },
+    {
+      "epoch": 0.8367527024045885,
+      "grad_norm": 0.4393473267555237,
+      "learning_rate": 1.3666633788598071e-05,
+      "loss": 0.3828,
+      "step": 11379
+    },
+    {
+      "epoch": 0.8368262372233253,
+      "grad_norm": 0.7678342461585999,
+      "learning_rate": 1.3654617793125857e-05,
+      "loss": 0.3984,
+      "step": 11380
+    },
+    {
+      "epoch": 0.8368997720420619,
+      "grad_norm": 0.592255711555481,
+      "learning_rate": 1.3642606695254855e-05,
+      "loss": 0.343,
+      "step": 11381
+    },
+    {
+      "epoch": 0.8369733068607986,
+      "grad_norm": 0.6515989899635315,
+      "learning_rate": 1.3630600495666335e-05,
+      "loss": 0.3496,
+      "step": 11382
+    },
+    {
+      "epoch": 0.8370468416795352,
+      "grad_norm": 0.4287654459476471,
+      "learning_rate": 1.3618599195041292e-05,
+      "loss": 0.3518,
+      "step": 11383
+    },
+    {
+      "epoch": 0.837120376498272,
+      "grad_norm": 0.3615548610687256,
+      "learning_rate": 1.3606602794060486e-05,
+      "loss": 0.3498,
+      "step": 11384
+    },
+    {
+      "epoch": 0.8371939113170086,
+      "grad_norm": 0.6124247908592224,
+      "learning_rate": 1.3594611293404324e-05,
+      "loss": 0.3704,
+      "step": 11385
+    },
+    {
+      "epoch": 0.8372674461357453,
+      "grad_norm": 0.5729259252548218,
+      "learning_rate": 1.3582624693753021e-05,
+      "loss": 0.3501,
+      "step": 11386
+    },
+    {
+      "epoch": 0.837340980954482,
+      "grad_norm": 0.3964386284351349,
+      "learning_rate": 1.357064299578643e-05,
+      "loss": 0.3685,
+      "step": 11387
+    },
+    {
+      "epoch": 0.8374145157732186,
+      "grad_norm": 0.5016558766365051,
+      "learning_rate": 1.3558666200184189e-05,
+      "loss": 0.3703,
+      "step": 11388
+    },
+    {
+      "epoch": 0.8374880505919553,
+      "grad_norm": 0.4891703724861145,
+      "learning_rate": 1.3546694307625618e-05,
+      "loss": 0.3512,
+      "step": 11389
+    },
+    {
+      "epoch": 0.8375615854106919,
+      "grad_norm": 0.4360412061214447,
+      "learning_rate": 1.3534727318789776e-05,
+      "loss": 0.3693,
+      "step": 11390
+    },
+    {
+      "epoch": 0.8376351202294287,
+      "grad_norm": 0.5019968748092651,
+      "learning_rate": 1.3522765234355473e-05,
+      "loss": 0.376,
+      "step": 11391
+    },
+    {
+      "epoch": 0.8377086550481653,
+      "grad_norm": 0.5177532434463501,
+      "learning_rate": 1.3510808055001168e-05,
+      "loss": 0.353,
+      "step": 11392
+    },
+    {
+      "epoch": 0.837782189866902,
+      "grad_norm": 0.3588496446609497,
+      "learning_rate": 1.3498855781405106e-05,
+      "loss": 0.3682,
+      "step": 11393
+    },
+    {
+      "epoch": 0.8378557246856386,
+      "grad_norm": 0.4614690840244293,
+      "learning_rate": 1.3486908414245247e-05,
+      "loss": 0.3711,
+      "step": 11394
+    },
+    {
+      "epoch": 0.8379292595043754,
+      "grad_norm": 0.5014209747314453,
+      "learning_rate": 1.3474965954199237e-05,
+      "loss": 0.368,
+      "step": 11395
+    },
+    {
+      "epoch": 0.838002794323112,
+      "grad_norm": 0.3956346809864044,
+      "learning_rate": 1.3463028401944455e-05,
+      "loss": 0.3654,
+      "step": 11396
+    },
+    {
+      "epoch": 0.8380763291418487,
+      "grad_norm": 0.3884434998035431,
+      "learning_rate": 1.3451095758158061e-05,
+      "loss": 0.3375,
+      "step": 11397
+    },
+    {
+      "epoch": 0.8381498639605853,
+      "grad_norm": 0.5551765561103821,
+      "learning_rate": 1.3439168023516858e-05,
+      "loss": 0.3783,
+      "step": 11398
+    },
+    {
+      "epoch": 0.8382233987793221,
+      "grad_norm": 0.4232933223247528,
+      "learning_rate": 1.3427245198697359e-05,
+      "loss": 0.368,
+      "step": 11399
+    },
+    {
+      "epoch": 0.8382969335980587,
+      "grad_norm": 0.39370936155319214,
+      "learning_rate": 1.341532728437589e-05,
+      "loss": 0.3743,
+      "step": 11400
+    },
+    {
+      "epoch": 0.8383704684167953,
+      "grad_norm": 0.46810340881347656,
+      "learning_rate": 1.3403414281228443e-05,
+      "loss": 0.3694,
+      "step": 11401
+    },
+    {
+      "epoch": 0.838444003235532,
+      "grad_norm": 0.42353329062461853,
+      "learning_rate": 1.3391506189930713e-05,
+      "loss": 0.3915,
+      "step": 11402
+    },
+    {
+      "epoch": 0.8385175380542687,
+      "grad_norm": 0.44399264454841614,
+      "learning_rate": 1.337960301115816e-05,
+      "loss": 0.3778,
+      "step": 11403
+    },
+    {
+      "epoch": 0.8385910728730054,
+      "grad_norm": 0.454972505569458,
+      "learning_rate": 1.336770474558595e-05,
+      "loss": 0.3685,
+      "step": 11404
+    },
+    {
+      "epoch": 0.838664607691742,
+      "grad_norm": 0.4580768942832947,
+      "learning_rate": 1.335581139388894e-05,
+      "loss": 0.3572,
+      "step": 11405
+    },
+    {
+      "epoch": 0.8387381425104787,
+      "grad_norm": 0.4869502782821655,
+      "learning_rate": 1.3343922956741738e-05,
+      "loss": 0.3925,
+      "step": 11406
+    },
+    {
+      "epoch": 0.8388116773292154,
+      "grad_norm": 0.47774019837379456,
+      "learning_rate": 1.3332039434818699e-05,
+      "loss": 0.3507,
+      "step": 11407
+    },
+    {
+      "epoch": 0.8388852121479521,
+      "grad_norm": 0.35038620233535767,
+      "learning_rate": 1.3320160828793836e-05,
+      "loss": 0.3535,
+      "step": 11408
+    },
+    {
+      "epoch": 0.8389587469666887,
+      "grad_norm": 0.5177643895149231,
+      "learning_rate": 1.3308287139340913e-05,
+      "loss": 0.37,
+      "step": 11409
+    },
+    {
+      "epoch": 0.8390322817854254,
+      "grad_norm": 0.5502418279647827,
+      "learning_rate": 1.3296418367133468e-05,
+      "loss": 0.361,
+      "step": 11410
+    },
+    {
+      "epoch": 0.8391058166041621,
+      "grad_norm": 0.4156407415866852,
+      "learning_rate": 1.328455451284467e-05,
+      "loss": 0.3581,
+      "step": 11411
+    },
+    {
+      "epoch": 0.8391793514228988,
+      "grad_norm": 0.38400503993034363,
+      "learning_rate": 1.327269557714743e-05,
+      "loss": 0.3397,
+      "step": 11412
+    },
+    {
+      "epoch": 0.8392528862416354,
+      "grad_norm": 0.49977338314056396,
+      "learning_rate": 1.3260841560714433e-05,
+      "loss": 0.3651,
+      "step": 11413
+    },
+    {
+      "epoch": 0.8393264210603721,
+      "grad_norm": 0.5332417488098145,
+      "learning_rate": 1.3248992464218058e-05,
+      "loss": 0.356,
+      "step": 11414
+    },
+    {
+      "epoch": 0.8393999558791088,
+      "grad_norm": 0.5692765116691589,
+      "learning_rate": 1.323714828833036e-05,
+      "loss": 0.3565,
+      "step": 11415
+    },
+    {
+      "epoch": 0.8394734906978454,
+      "grad_norm": 0.4621616005897522,
+      "learning_rate": 1.3225309033723177e-05,
+      "loss": 0.3447,
+      "step": 11416
+    },
+    {
+      "epoch": 0.8395470255165821,
+      "grad_norm": 0.5455951690673828,
+      "learning_rate": 1.3213474701068063e-05,
+      "loss": 0.38,
+      "step": 11417
+    },
+    {
+      "epoch": 0.8396205603353187,
+      "grad_norm": 0.5633748173713684,
+      "learning_rate": 1.3201645291036235e-05,
+      "loss": 0.3763,
+      "step": 11418
+    },
+    {
+      "epoch": 0.8396940951540555,
+      "grad_norm": 0.5245444774627686,
+      "learning_rate": 1.318982080429868e-05,
+      "loss": 0.3733,
+      "step": 11419
+    },
+    {
+      "epoch": 0.8397676299727921,
+      "grad_norm": 0.42772749066352844,
+      "learning_rate": 1.3178001241526128e-05,
+      "loss": 0.3335,
+      "step": 11420
+    },
+    {
+      "epoch": 0.8398411647915288,
+      "grad_norm": 0.5024347305297852,
+      "learning_rate": 1.3166186603388975e-05,
+      "loss": 0.3565,
+      "step": 11421
+    },
+    {
+      "epoch": 0.8399146996102654,
+      "grad_norm": 0.5521784424781799,
+      "learning_rate": 1.315437689055734e-05,
+      "loss": 0.3646,
+      "step": 11422
+    },
+    {
+      "epoch": 0.8399882344290022,
+      "grad_norm": 0.38182613253593445,
+      "learning_rate": 1.314257210370109e-05,
+      "loss": 0.3641,
+      "step": 11423
+    },
+    {
+      "epoch": 0.8400617692477388,
+      "grad_norm": 0.4264533519744873,
+      "learning_rate": 1.3130772243489842e-05,
+      "loss": 0.3609,
+      "step": 11424
+    },
+    {
+      "epoch": 0.8401353040664755,
+      "grad_norm": 0.44017016887664795,
+      "learning_rate": 1.3118977310592851e-05,
+      "loss": 0.3987,
+      "step": 11425
+    },
+    {
+      "epoch": 0.8402088388852121,
+      "grad_norm": 0.5465112328529358,
+      "learning_rate": 1.3107187305679147e-05,
+      "loss": 0.3505,
+      "step": 11426
+    },
+    {
+      "epoch": 0.8402823737039489,
+      "grad_norm": 0.500700831413269,
+      "learning_rate": 1.3095402229417509e-05,
+      "loss": 0.3512,
+      "step": 11427
+    },
+    {
+      "epoch": 0.8403559085226855,
+      "grad_norm": 0.43772897124290466,
+      "learning_rate": 1.3083622082476355e-05,
+      "loss": 0.3615,
+      "step": 11428
+    },
+    {
+      "epoch": 0.8404294433414221,
+      "grad_norm": 0.5167012810707092,
+      "learning_rate": 1.3071846865523873e-05,
+      "loss": 0.3494,
+      "step": 11429
+    },
+    {
+      "epoch": 0.8405029781601588,
+      "grad_norm": 0.5590175986289978,
+      "learning_rate": 1.3060076579227997e-05,
+      "loss": 0.3697,
+      "step": 11430
+    },
+    {
+      "epoch": 0.8405765129788955,
+      "grad_norm": 0.40555042028427124,
+      "learning_rate": 1.3048311224256305e-05,
+      "loss": 0.3297,
+      "step": 11431
+    },
+    {
+      "epoch": 0.8406500477976322,
+      "grad_norm": 0.44621819257736206,
+      "learning_rate": 1.3036550801276171e-05,
+      "loss": 0.368,
+      "step": 11432
+    },
+    {
+      "epoch": 0.8407235826163688,
+      "grad_norm": 0.5295449495315552,
+      "learning_rate": 1.3024795310954674e-05,
+      "loss": 0.39,
+      "step": 11433
+    },
+    {
+      "epoch": 0.8407971174351055,
+      "grad_norm": 0.5085777044296265,
+      "learning_rate": 1.301304475395857e-05,
+      "loss": 0.4105,
+      "step": 11434
+    },
+    {
+      "epoch": 0.8408706522538422,
+      "grad_norm": 0.4290333390235901,
+      "learning_rate": 1.3001299130954347e-05,
+      "loss": 0.3767,
+      "step": 11435
+    },
+    {
+      "epoch": 0.8409441870725789,
+      "grad_norm": 0.7151695489883423,
+      "learning_rate": 1.2989558442608251e-05,
+      "loss": 0.3696,
+      "step": 11436
+    },
+    {
+      "epoch": 0.8410177218913155,
+      "grad_norm": 0.5377374887466431,
+      "learning_rate": 1.297782268958625e-05,
+      "loss": 0.3616,
+      "step": 11437
+    },
+    {
+      "epoch": 0.8410912567100522,
+      "grad_norm": 0.4346553385257721,
+      "learning_rate": 1.2966091872553965e-05,
+      "loss": 0.3466,
+      "step": 11438
+    },
+    {
+      "epoch": 0.8411647915287889,
+      "grad_norm": 0.6926043629646301,
+      "learning_rate": 1.2954365992176798e-05,
+      "loss": 0.3841,
+      "step": 11439
+    },
+    {
+      "epoch": 0.8412383263475256,
+      "grad_norm": 0.48623228073120117,
+      "learning_rate": 1.2942645049119873e-05,
+      "loss": 0.3716,
+      "step": 11440
+    },
+    {
+      "epoch": 0.8413118611662622,
+      "grad_norm": 0.5654765963554382,
+      "learning_rate": 1.293092904404798e-05,
+      "loss": 0.3497,
+      "step": 11441
+    },
+    {
+      "epoch": 0.841385395984999,
+      "grad_norm": 0.5933129787445068,
+      "learning_rate": 1.2919217977625675e-05,
+      "loss": 0.3763,
+      "step": 11442
+    },
+    {
+      "epoch": 0.8414589308037356,
+      "grad_norm": 0.40599891543388367,
+      "learning_rate": 1.2907511850517251e-05,
+      "loss": 0.3405,
+      "step": 11443
+    },
+    {
+      "epoch": 0.8415324656224722,
+      "grad_norm": 0.38505151867866516,
+      "learning_rate": 1.2895810663386643e-05,
+      "loss": 0.3528,
+      "step": 11444
+    },
+    {
+      "epoch": 0.8416060004412089,
+      "grad_norm": 0.5559920072555542,
+      "learning_rate": 1.2884114416897596e-05,
+      "loss": 0.4053,
+      "step": 11445
+    },
+    {
+      "epoch": 0.8416795352599455,
+      "grad_norm": 0.5203341841697693,
+      "learning_rate": 1.2872423111713505e-05,
+      "loss": 0.385,
+      "step": 11446
+    },
+    {
+      "epoch": 0.8417530700786823,
+      "grad_norm": 0.4083956480026245,
+      "learning_rate": 1.2860736748497537e-05,
+      "loss": 0.3684,
+      "step": 11447
+    },
+    {
+      "epoch": 0.8418266048974189,
+      "grad_norm": 0.5614303350448608,
+      "learning_rate": 1.284905532791253e-05,
+      "loss": 0.3634,
+      "step": 11448
+    },
+    {
+      "epoch": 0.8419001397161556,
+      "grad_norm": 0.4672212600708008,
+      "learning_rate": 1.2837378850621074e-05,
+      "loss": 0.3576,
+      "step": 11449
+    },
+    {
+      "epoch": 0.8419736745348922,
+      "grad_norm": 0.43969401717185974,
+      "learning_rate": 1.2825707317285495e-05,
+      "loss": 0.3675,
+      "step": 11450
+    },
+    {
+      "epoch": 0.842047209353629,
+      "grad_norm": 0.6169220805168152,
+      "learning_rate": 1.2814040728567777e-05,
+      "loss": 0.3485,
+      "step": 11451
+    },
+    {
+      "epoch": 0.8421207441723656,
+      "grad_norm": 0.46799641847610474,
+      "learning_rate": 1.2802379085129678e-05,
+      "loss": 0.3536,
+      "step": 11452
+    },
+    {
+      "epoch": 0.8421942789911023,
+      "grad_norm": 0.4617173969745636,
+      "learning_rate": 1.2790722387632681e-05,
+      "loss": 0.3525,
+      "step": 11453
+    },
+    {
+      "epoch": 0.842267813809839,
+      "grad_norm": 0.4154883623123169,
+      "learning_rate": 1.2779070636737922e-05,
+      "loss": 0.3627,
+      "step": 11454
+    },
+    {
+      "epoch": 0.8423413486285757,
+      "grad_norm": 0.4086017310619354,
+      "learning_rate": 1.2767423833106328e-05,
+      "loss": 0.373,
+      "step": 11455
+    },
+    {
+      "epoch": 0.8424148834473123,
+      "grad_norm": 0.41067075729370117,
+      "learning_rate": 1.275578197739854e-05,
+      "loss": 0.3536,
+      "step": 11456
+    },
+    {
+      "epoch": 0.8424884182660489,
+      "grad_norm": 0.462220698595047,
+      "learning_rate": 1.2744145070274871e-05,
+      "loss": 0.3537,
+      "step": 11457
+    },
+    {
+      "epoch": 0.8425619530847857,
+      "grad_norm": 0.4580128788948059,
+      "learning_rate": 1.273251311239535e-05,
+      "loss": 0.3484,
+      "step": 11458
+    },
+    {
+      "epoch": 0.8426354879035223,
+      "grad_norm": 0.37101036310195923,
+      "learning_rate": 1.2720886104419783e-05,
+      "loss": 0.355,
+      "step": 11459
+    },
+    {
+      "epoch": 0.842709022722259,
+      "grad_norm": 0.4188695549964905,
+      "learning_rate": 1.2709264047007696e-05,
+      "loss": 0.3579,
+      "step": 11460
+    },
+    {
+      "epoch": 0.8427825575409956,
+      "grad_norm": 0.4291454255580902,
+      "learning_rate": 1.2697646940818242e-05,
+      "loss": 0.3598,
+      "step": 11461
+    },
+    {
+      "epoch": 0.8428560923597324,
+      "grad_norm": 0.5926879048347473,
+      "learning_rate": 1.2686034786510392e-05,
+      "loss": 0.3891,
+      "step": 11462
+    },
+    {
+      "epoch": 0.842929627178469,
+      "grad_norm": 0.38753941655158997,
+      "learning_rate": 1.267442758474281e-05,
+      "loss": 0.3494,
+      "step": 11463
+    },
+    {
+      "epoch": 0.8430031619972057,
+      "grad_norm": 0.5304745435714722,
+      "learning_rate": 1.2662825336173845e-05,
+      "loss": 0.3454,
+      "step": 11464
+    },
+    {
+      "epoch": 0.8430766968159423,
+      "grad_norm": 0.4618358314037323,
+      "learning_rate": 1.2651228041461582e-05,
+      "loss": 0.3202,
+      "step": 11465
+    },
+    {
+      "epoch": 0.8431502316346791,
+      "grad_norm": 0.43896400928497314,
+      "learning_rate": 1.2639635701263885e-05,
+      "loss": 0.3253,
+      "step": 11466
+    },
+    {
+      "epoch": 0.8432237664534157,
+      "grad_norm": 0.4653896987438202,
+      "learning_rate": 1.2628048316238217e-05,
+      "loss": 0.3603,
+      "step": 11467
+    },
+    {
+      "epoch": 0.8432973012721524,
+      "grad_norm": 0.469888836145401,
+      "learning_rate": 1.2616465887041872e-05,
+      "loss": 0.3313,
+      "step": 11468
+    },
+    {
+      "epoch": 0.843370836090889,
+      "grad_norm": 0.6034677028656006,
+      "learning_rate": 1.2604888414331784e-05,
+      "loss": 0.3797,
+      "step": 11469
+    },
+    {
+      "epoch": 0.8434443709096258,
+      "grad_norm": 0.5872542262077332,
+      "learning_rate": 1.259331589876468e-05,
+      "loss": 0.34,
+      "step": 11470
+    },
+    {
+      "epoch": 0.8435179057283624,
+      "grad_norm": 0.39557939767837524,
+      "learning_rate": 1.2581748340996925e-05,
+      "loss": 0.3325,
+      "step": 11471
+    },
+    {
+      "epoch": 0.843591440547099,
+      "grad_norm": 0.4733774662017822,
+      "learning_rate": 1.2570185741684648e-05,
+      "loss": 0.3538,
+      "step": 11472
+    },
+    {
+      "epoch": 0.8436649753658357,
+      "grad_norm": 0.38245266675949097,
+      "learning_rate": 1.2558628101483738e-05,
+      "loss": 0.3579,
+      "step": 11473
+    },
+    {
+      "epoch": 0.8437385101845724,
+      "grad_norm": 0.7058207392692566,
+      "learning_rate": 1.2547075421049692e-05,
+      "loss": 0.3592,
+      "step": 11474
+    },
+    {
+      "epoch": 0.8438120450033091,
+      "grad_norm": 0.3750567138195038,
+      "learning_rate": 1.253552770103783e-05,
+      "loss": 0.343,
+      "step": 11475
+    },
+    {
+      "epoch": 0.8438855798220457,
+      "grad_norm": 0.5900620818138123,
+      "learning_rate": 1.2523984942103151e-05,
+      "loss": 0.3412,
+      "step": 11476
+    },
+    {
+      "epoch": 0.8439591146407824,
+      "grad_norm": 0.42745864391326904,
+      "learning_rate": 1.2512447144900352e-05,
+      "loss": 0.3402,
+      "step": 11477
+    },
+    {
+      "epoch": 0.8440326494595191,
+      "grad_norm": 0.4820197820663452,
+      "learning_rate": 1.2500914310083877e-05,
+      "loss": 0.3667,
+      "step": 11478
+    },
+    {
+      "epoch": 0.8441061842782558,
+      "grad_norm": 0.38567137718200684,
+      "learning_rate": 1.2489386438307904e-05,
+      "loss": 0.3391,
+      "step": 11479
+    },
+    {
+      "epoch": 0.8441797190969924,
+      "grad_norm": 0.395542174577713,
+      "learning_rate": 1.2477863530226286e-05,
+      "loss": 0.3467,
+      "step": 11480
+    },
+    {
+      "epoch": 0.8442532539157291,
+      "grad_norm": 0.3630913197994232,
+      "learning_rate": 1.24663455864926e-05,
+      "loss": 0.3512,
+      "step": 11481
+    },
+    {
+      "epoch": 0.8443267887344658,
+      "grad_norm": 0.43232521414756775,
+      "learning_rate": 1.2454832607760159e-05,
+      "loss": 0.3586,
+      "step": 11482
+    },
+    {
+      "epoch": 0.8444003235532025,
+      "grad_norm": 0.38727959990501404,
+      "learning_rate": 1.2443324594682038e-05,
+      "loss": 0.3523,
+      "step": 11483
+    },
+    {
+      "epoch": 0.8444738583719391,
+      "grad_norm": 0.43064847588539124,
+      "learning_rate": 1.2431821547910915e-05,
+      "loss": 0.3607,
+      "step": 11484
+    },
+    {
+      "epoch": 0.8445473931906757,
+      "grad_norm": 0.4508702754974365,
+      "learning_rate": 1.2420323468099294e-05,
+      "loss": 0.3476,
+      "step": 11485
+    },
+    {
+      "epoch": 0.8446209280094125,
+      "grad_norm": 0.432282418012619,
+      "learning_rate": 1.2408830355899371e-05,
+      "loss": 0.3649,
+      "step": 11486
+    },
+    {
+      "epoch": 0.8446944628281491,
+      "grad_norm": 0.4851264953613281,
+      "learning_rate": 1.2397342211963003e-05,
+      "loss": 0.3664,
+      "step": 11487
+    },
+    {
+      "epoch": 0.8447679976468858,
+      "grad_norm": 0.38598906993865967,
+      "learning_rate": 1.2385859036941838e-05,
+      "loss": 0.3448,
+      "step": 11488
+    },
+    {
+      "epoch": 0.8448415324656224,
+      "grad_norm": 0.46442726254463196,
+      "learning_rate": 1.2374380831487231e-05,
+      "loss": 0.3581,
+      "step": 11489
+    },
+    {
+      "epoch": 0.8449150672843592,
+      "grad_norm": 0.4024268090724945,
+      "learning_rate": 1.2362907596250206e-05,
+      "loss": 0.329,
+      "step": 11490
+    },
+    {
+      "epoch": 0.8449886021030958,
+      "grad_norm": 0.5120164752006531,
+      "learning_rate": 1.2351439331881542e-05,
+      "loss": 0.3504,
+      "step": 11491
+    },
+    {
+      "epoch": 0.8450621369218325,
+      "grad_norm": 0.3791288435459137,
+      "learning_rate": 1.2339976039031764e-05,
+      "loss": 0.3145,
+      "step": 11492
+    },
+    {
+      "epoch": 0.8451356717405691,
+      "grad_norm": 0.4198800325393677,
+      "learning_rate": 1.2328517718351052e-05,
+      "loss": 0.3385,
+      "step": 11493
+    },
+    {
+      "epoch": 0.8452092065593059,
+      "grad_norm": 0.39260661602020264,
+      "learning_rate": 1.2317064370489317e-05,
+      "loss": 0.3678,
+      "step": 11494
+    },
+    {
+      "epoch": 0.8452827413780425,
+      "grad_norm": 0.5167377591133118,
+      "learning_rate": 1.2305615996096232e-05,
+      "loss": 0.3636,
+      "step": 11495
+    },
+    {
+      "epoch": 0.8453562761967792,
+      "grad_norm": 0.40363988280296326,
+      "learning_rate": 1.2294172595821174e-05,
+      "loss": 0.3263,
+      "step": 11496
+    },
+    {
+      "epoch": 0.8454298110155158,
+      "grad_norm": 0.44222453236579895,
+      "learning_rate": 1.2282734170313182e-05,
+      "loss": 0.3668,
+      "step": 11497
+    },
+    {
+      "epoch": 0.8455033458342526,
+      "grad_norm": 0.479416161775589,
+      "learning_rate": 1.2271300720221079e-05,
+      "loss": 0.3267,
+      "step": 11498
+    },
+    {
+      "epoch": 0.8455768806529892,
+      "grad_norm": 0.4798383116722107,
+      "learning_rate": 1.2259872246193405e-05,
+      "loss": 0.3534,
+      "step": 11499
+    },
+    {
+      "epoch": 0.8456504154717258,
+      "grad_norm": 0.37064656615257263,
+      "learning_rate": 1.224844874887835e-05,
+      "loss": 0.3405,
+      "step": 11500
+    },
+    {
+      "epoch": 0.8457239502904625,
+      "grad_norm": 0.46244189143180847,
+      "learning_rate": 1.2237030228923895e-05,
+      "loss": 0.3494,
+      "step": 11501
+    },
+    {
+      "epoch": 0.8457974851091992,
+      "grad_norm": 0.471134752035141,
+      "learning_rate": 1.2225616686977726e-05,
+      "loss": 0.3438,
+      "step": 11502
+    },
+    {
+      "epoch": 0.8458710199279359,
+      "grad_norm": 0.4575377106666565,
+      "learning_rate": 1.221420812368721e-05,
+      "loss": 0.3428,
+      "step": 11503
+    },
+    {
+      "epoch": 0.8459445547466725,
+      "grad_norm": 0.4627022445201874,
+      "learning_rate": 1.2202804539699431e-05,
+      "loss": 0.3477,
+      "step": 11504
+    },
+    {
+      "epoch": 0.8460180895654092,
+      "grad_norm": 0.40497085452079773,
+      "learning_rate": 1.2191405935661237e-05,
+      "loss": 0.3613,
+      "step": 11505
+    },
+    {
+      "epoch": 0.8460916243841459,
+      "grad_norm": 0.45131799578666687,
+      "learning_rate": 1.218001231221918e-05,
+      "loss": 0.3544,
+      "step": 11506
+    },
+    {
+      "epoch": 0.8461651592028826,
+      "grad_norm": 0.5552036166191101,
+      "learning_rate": 1.2168623670019486e-05,
+      "loss": 0.3645,
+      "step": 11507
+    },
+    {
+      "epoch": 0.8462386940216192,
+      "grad_norm": 0.40757256746292114,
+      "learning_rate": 1.215724000970816e-05,
+      "loss": 0.3692,
+      "step": 11508
+    },
+    {
+      "epoch": 0.846312228840356,
+      "grad_norm": 0.39623066782951355,
+      "learning_rate": 1.2145861331930897e-05,
+      "loss": 0.3465,
+      "step": 11509
+    },
+    {
+      "epoch": 0.8463857636590926,
+      "grad_norm": 0.4475249946117401,
+      "learning_rate": 1.213448763733307e-05,
+      "loss": 0.3405,
+      "step": 11510
+    },
+    {
+      "epoch": 0.8464592984778293,
+      "grad_norm": 0.47182416915893555,
+      "learning_rate": 1.212311892655984e-05,
+      "loss": 0.3638,
+      "step": 11511
+    },
+    {
+      "epoch": 0.8465328332965659,
+      "grad_norm": 0.38574305176734924,
+      "learning_rate": 1.211175520025607e-05,
+      "loss": 0.3592,
+      "step": 11512
+    },
+    {
+      "epoch": 0.8466063681153025,
+      "grad_norm": 0.49712100625038147,
+      "learning_rate": 1.2100396459066266e-05,
+      "loss": 0.3554,
+      "step": 11513
+    },
+    {
+      "epoch": 0.8466799029340393,
+      "grad_norm": 0.4946998953819275,
+      "learning_rate": 1.2089042703634746e-05,
+      "loss": 0.3607,
+      "step": 11514
+    },
+    {
+      "epoch": 0.8467534377527759,
+      "grad_norm": 0.3787076473236084,
+      "learning_rate": 1.2077693934605516e-05,
+      "loss": 0.3473,
+      "step": 11515
+    },
+    {
+      "epoch": 0.8468269725715126,
+      "grad_norm": 0.44359275698661804,
+      "learning_rate": 1.2066350152622274e-05,
+      "loss": 0.3667,
+      "step": 11516
+    },
+    {
+      "epoch": 0.8469005073902492,
+      "grad_norm": 0.380738228559494,
+      "learning_rate": 1.2055011358328427e-05,
+      "loss": 0.3461,
+      "step": 11517
+    },
+    {
+      "epoch": 0.846974042208986,
+      "grad_norm": 0.396933376789093,
+      "learning_rate": 1.2043677552367161e-05,
+      "loss": 0.345,
+      "step": 11518
+    },
+    {
+      "epoch": 0.8470475770277226,
+      "grad_norm": 0.4448789060115814,
+      "learning_rate": 1.2032348735381338e-05,
+      "loss": 0.3561,
+      "step": 11519
+    },
+    {
+      "epoch": 0.8471211118464593,
+      "grad_norm": 0.4672824442386627,
+      "learning_rate": 1.2021024908013512e-05,
+      "loss": 0.3823,
+      "step": 11520
+    },
+    {
+      "epoch": 0.847194646665196,
+      "grad_norm": 0.4189094007015228,
+      "learning_rate": 1.2009706070906001e-05,
+      "loss": 0.3796,
+      "step": 11521
+    },
+    {
+      "epoch": 0.8472681814839327,
+      "grad_norm": 0.37844985723495483,
+      "learning_rate": 1.199839222470085e-05,
+      "loss": 0.3546,
+      "step": 11522
+    },
+    {
+      "epoch": 0.8473417163026693,
+      "grad_norm": 0.4001138508319855,
+      "learning_rate": 1.198708337003973e-05,
+      "loss": 0.3553,
+      "step": 11523
+    },
+    {
+      "epoch": 0.847415251121406,
+      "grad_norm": 0.38108378648757935,
+      "learning_rate": 1.1975779507564133e-05,
+      "loss": 0.3562,
+      "step": 11524
+    },
+    {
+      "epoch": 0.8474887859401427,
+      "grad_norm": 0.36021119356155396,
+      "learning_rate": 1.1964480637915243e-05,
+      "loss": 0.3457,
+      "step": 11525
+    },
+    {
+      "epoch": 0.8475623207588794,
+      "grad_norm": 0.4061339795589447,
+      "learning_rate": 1.1953186761733893e-05,
+      "loss": 0.3519,
+      "step": 11526
+    },
+    {
+      "epoch": 0.847635855577616,
+      "grad_norm": 0.39723965525627136,
+      "learning_rate": 1.1941897879660724e-05,
+      "loss": 0.34,
+      "step": 11527
+    },
+    {
+      "epoch": 0.8477093903963526,
+      "grad_norm": 0.38767337799072266,
+      "learning_rate": 1.1930613992336026e-05,
+      "loss": 0.3747,
+      "step": 11528
+    },
+    {
+      "epoch": 0.8477829252150894,
+      "grad_norm": 0.4394706189632416,
+      "learning_rate": 1.1919335100399864e-05,
+      "loss": 0.3439,
+      "step": 11529
+    },
+    {
+      "epoch": 0.847856460033826,
+      "grad_norm": 0.4915807247161865,
+      "learning_rate": 1.190806120449195e-05,
+      "loss": 0.3366,
+      "step": 11530
+    },
+    {
+      "epoch": 0.8479299948525627,
+      "grad_norm": 0.5293079018592834,
+      "learning_rate": 1.189679230525177e-05,
+      "loss": 0.3374,
+      "step": 11531
+    },
+    {
+      "epoch": 0.8480035296712993,
+      "grad_norm": 0.4686998426914215,
+      "learning_rate": 1.1885528403318524e-05,
+      "loss": 0.3351,
+      "step": 11532
+    },
+    {
+      "epoch": 0.8480770644900361,
+      "grad_norm": 0.43612393736839294,
+      "learning_rate": 1.1874269499331092e-05,
+      "loss": 0.3702,
+      "step": 11533
+    },
+    {
+      "epoch": 0.8481505993087727,
+      "grad_norm": 0.3937479257583618,
+      "learning_rate": 1.1863015593928084e-05,
+      "loss": 0.3594,
+      "step": 11534
+    },
+    {
+      "epoch": 0.8482241341275094,
+      "grad_norm": 0.6293612122535706,
+      "learning_rate": 1.1851766687747879e-05,
+      "loss": 0.3696,
+      "step": 11535
+    },
+    {
+      "epoch": 0.848297668946246,
+      "grad_norm": 0.6978211998939514,
+      "learning_rate": 1.1840522781428465e-05,
+      "loss": 0.3914,
+      "step": 11536
+    },
+    {
+      "epoch": 0.8483712037649828,
+      "grad_norm": 0.41263291239738464,
+      "learning_rate": 1.1829283875607633e-05,
+      "loss": 0.3398,
+      "step": 11537
+    },
+    {
+      "epoch": 0.8484447385837194,
+      "grad_norm": 0.7091941237449646,
+      "learning_rate": 1.1818049970922895e-05,
+      "loss": 0.365,
+      "step": 11538
+    },
+    {
+      "epoch": 0.8485182734024561,
+      "grad_norm": 0.5689542889595032,
+      "learning_rate": 1.180682106801143e-05,
+      "loss": 0.3541,
+      "step": 11539
+    },
+    {
+      "epoch": 0.8485918082211927,
+      "grad_norm": 0.4456079602241516,
+      "learning_rate": 1.179559716751012e-05,
+      "loss": 0.3508,
+      "step": 11540
+    },
+    {
+      "epoch": 0.8486653430399294,
+      "grad_norm": 0.4403965175151825,
+      "learning_rate": 1.178437827005563e-05,
+      "loss": 0.3505,
+      "step": 11541
+    },
+    {
+      "epoch": 0.8487388778586661,
+      "grad_norm": 0.6300821900367737,
+      "learning_rate": 1.1773164376284318e-05,
+      "loss": 0.3669,
+      "step": 11542
+    },
+    {
+      "epoch": 0.8488124126774027,
+      "grad_norm": 0.45781952142715454,
+      "learning_rate": 1.176195548683221e-05,
+      "loss": 0.3433,
+      "step": 11543
+    },
+    {
+      "epoch": 0.8488859474961394,
+      "grad_norm": 0.47859352827072144,
+      "learning_rate": 1.1750751602335119e-05,
+      "loss": 0.358,
+      "step": 11544
+    },
+    {
+      "epoch": 0.8489594823148761,
+      "grad_norm": 0.5753034353256226,
+      "learning_rate": 1.1739552723428537e-05,
+      "loss": 0.347,
+      "step": 11545
+    },
+    {
+      "epoch": 0.8490330171336128,
+      "grad_norm": 0.5289343595504761,
+      "learning_rate": 1.1728358850747655e-05,
+      "loss": 0.3697,
+      "step": 11546
+    },
+    {
+      "epoch": 0.8491065519523494,
+      "grad_norm": 0.5825220346450806,
+      "learning_rate": 1.1717169984927423e-05,
+      "loss": 0.3779,
+      "step": 11547
+    },
+    {
+      "epoch": 0.8491800867710861,
+      "grad_norm": 0.6168119311332703,
+      "learning_rate": 1.1705986126602487e-05,
+      "loss": 0.3487,
+      "step": 11548
+    },
+    {
+      "epoch": 0.8492536215898228,
+      "grad_norm": 0.41818851232528687,
+      "learning_rate": 1.1694807276407182e-05,
+      "loss": 0.3744,
+      "step": 11549
+    },
+    {
+      "epoch": 0.8493271564085595,
+      "grad_norm": 0.48595136404037476,
+      "learning_rate": 1.1683633434975626e-05,
+      "loss": 0.3487,
+      "step": 11550
+    },
+    {
+      "epoch": 0.8494006912272961,
+      "grad_norm": 0.6742846965789795,
+      "learning_rate": 1.1672464602941557e-05,
+      "loss": 0.3441,
+      "step": 11551
+    },
+    {
+      "epoch": 0.8494742260460328,
+      "grad_norm": 0.4393071234226227,
+      "learning_rate": 1.1661300780938544e-05,
+      "loss": 0.3882,
+      "step": 11552
+    },
+    {
+      "epoch": 0.8495477608647695,
+      "grad_norm": 0.4139839708805084,
+      "learning_rate": 1.1650141969599749e-05,
+      "loss": 0.3736,
+      "step": 11553
+    },
+    {
+      "epoch": 0.8496212956835062,
+      "grad_norm": 0.4690726101398468,
+      "learning_rate": 1.1638988169558141e-05,
+      "loss": 0.3363,
+      "step": 11554
+    },
+    {
+      "epoch": 0.8496948305022428,
+      "grad_norm": 0.5826889872550964,
+      "learning_rate": 1.1627839381446393e-05,
+      "loss": 0.3415,
+      "step": 11555
+    },
+    {
+      "epoch": 0.8497683653209794,
+      "grad_norm": 0.42355987429618835,
+      "learning_rate": 1.1616695605896843e-05,
+      "loss": 0.3423,
+      "step": 11556
+    },
+    {
+      "epoch": 0.8498419001397162,
+      "grad_norm": 0.3820217549800873,
+      "learning_rate": 1.1605556843541587e-05,
+      "loss": 0.3659,
+      "step": 11557
+    },
+    {
+      "epoch": 0.8499154349584528,
+      "grad_norm": 0.45160549879074097,
+      "learning_rate": 1.1594423095012453e-05,
+      "loss": 0.3489,
+      "step": 11558
+    },
+    {
+      "epoch": 0.8499889697771895,
+      "grad_norm": 0.4249917268753052,
+      "learning_rate": 1.1583294360940922e-05,
+      "loss": 0.3506,
+      "step": 11559
+    },
+    {
+      "epoch": 0.8500625045959261,
+      "grad_norm": 0.37108737230300903,
+      "learning_rate": 1.1572170641958246e-05,
+      "loss": 0.3353,
+      "step": 11560
+    },
+    {
+      "epoch": 0.8501360394146629,
+      "grad_norm": 0.43046656250953674,
+      "learning_rate": 1.15610519386954e-05,
+      "loss": 0.3484,
+      "step": 11561
+    },
+    {
+      "epoch": 0.8502095742333995,
+      "grad_norm": 0.380727082490921,
+      "learning_rate": 1.154993825178301e-05,
+      "loss": 0.3629,
+      "step": 11562
+    },
+    {
+      "epoch": 0.8502831090521362,
+      "grad_norm": 0.39276865124702454,
+      "learning_rate": 1.1538829581851451e-05,
+      "loss": 0.3448,
+      "step": 11563
+    },
+    {
+      "epoch": 0.8503566438708728,
+      "grad_norm": 0.4399825632572174,
+      "learning_rate": 1.152772592953083e-05,
+      "loss": 0.39,
+      "step": 11564
+    },
+    {
+      "epoch": 0.8504301786896096,
+      "grad_norm": 0.42473480105400085,
+      "learning_rate": 1.1516627295450989e-05,
+      "loss": 0.3549,
+      "step": 11565
+    },
+    {
+      "epoch": 0.8505037135083462,
+      "grad_norm": 0.5401582717895508,
+      "learning_rate": 1.150553368024141e-05,
+      "loss": 0.3594,
+      "step": 11566
+    },
+    {
+      "epoch": 0.8505772483270829,
+      "grad_norm": 0.4195626974105835,
+      "learning_rate": 1.1494445084531346e-05,
+      "loss": 0.3489,
+      "step": 11567
+    },
+    {
+      "epoch": 0.8506507831458195,
+      "grad_norm": 0.3673669099807739,
+      "learning_rate": 1.1483361508949786e-05,
+      "loss": 0.3349,
+      "step": 11568
+    },
+    {
+      "epoch": 0.8507243179645562,
+      "grad_norm": 0.4172593653202057,
+      "learning_rate": 1.1472282954125357e-05,
+      "loss": 0.333,
+      "step": 11569
+    },
+    {
+      "epoch": 0.8507978527832929,
+      "grad_norm": 0.49098363518714905,
+      "learning_rate": 1.1461209420686458e-05,
+      "loss": 0.3566,
+      "step": 11570
+    },
+    {
+      "epoch": 0.8508713876020295,
+      "grad_norm": 0.42035526037216187,
+      "learning_rate": 1.145014090926123e-05,
+      "loss": 0.3561,
+      "step": 11571
+    },
+    {
+      "epoch": 0.8509449224207662,
+      "grad_norm": 0.42111077904701233,
+      "learning_rate": 1.1439077420477429e-05,
+      "loss": 0.3447,
+      "step": 11572
+    },
+    {
+      "epoch": 0.8510184572395029,
+      "grad_norm": 0.5728985071182251,
+      "learning_rate": 1.142801895496265e-05,
+      "loss": 0.3517,
+      "step": 11573
+    },
+    {
+      "epoch": 0.8510919920582396,
+      "grad_norm": 0.5004011392593384,
+      "learning_rate": 1.141696551334408e-05,
+      "loss": 0.3634,
+      "step": 11574
+    },
+    {
+      "epoch": 0.8511655268769762,
+      "grad_norm": 0.37841811776161194,
+      "learning_rate": 1.1405917096248742e-05,
+      "loss": 0.3483,
+      "step": 11575
+    },
+    {
+      "epoch": 0.851239061695713,
+      "grad_norm": 0.6373973488807678,
+      "learning_rate": 1.1394873704303255e-05,
+      "loss": 0.3602,
+      "step": 11576
+    },
+    {
+      "epoch": 0.8513125965144496,
+      "grad_norm": 0.48754698038101196,
+      "learning_rate": 1.1383835338134042e-05,
+      "loss": 0.3691,
+      "step": 11577
+    },
+    {
+      "epoch": 0.8513861313331863,
+      "grad_norm": 0.4557613134384155,
+      "learning_rate": 1.1372801998367221e-05,
+      "loss": 0.3575,
+      "step": 11578
+    },
+    {
+      "epoch": 0.8514596661519229,
+      "grad_norm": 0.4115234613418579,
+      "learning_rate": 1.1361773685628596e-05,
+      "loss": 0.3594,
+      "step": 11579
+    },
+    {
+      "epoch": 0.8515332009706597,
+      "grad_norm": 0.40068143606185913,
+      "learning_rate": 1.1350750400543708e-05,
+      "loss": 0.3845,
+      "step": 11580
+    },
+    {
+      "epoch": 0.8516067357893963,
+      "grad_norm": 0.3801517188549042,
+      "learning_rate": 1.1339732143737825e-05,
+      "loss": 0.3481,
+      "step": 11581
+    },
+    {
+      "epoch": 0.851680270608133,
+      "grad_norm": 0.41065356135368347,
+      "learning_rate": 1.132871891583589e-05,
+      "loss": 0.3752,
+      "step": 11582
+    },
+    {
+      "epoch": 0.8517538054268696,
+      "grad_norm": 0.4241009056568146,
+      "learning_rate": 1.1317710717462593e-05,
+      "loss": 0.3687,
+      "step": 11583
+    },
+    {
+      "epoch": 0.8518273402456062,
+      "grad_norm": 0.44471898674964905,
+      "learning_rate": 1.1306707549242357e-05,
+      "loss": 0.3496,
+      "step": 11584
+    },
+    {
+      "epoch": 0.851900875064343,
+      "grad_norm": 0.4077790081501007,
+      "learning_rate": 1.1295709411799271e-05,
+      "loss": 0.3578,
+      "step": 11585
+    },
+    {
+      "epoch": 0.8519744098830796,
+      "grad_norm": 0.40483298897743225,
+      "learning_rate": 1.1284716305757136e-05,
+      "loss": 0.353,
+      "step": 11586
+    },
+    {
+      "epoch": 0.8520479447018163,
+      "grad_norm": 0.37025168538093567,
+      "learning_rate": 1.1273728231739522e-05,
+      "loss": 0.3695,
+      "step": 11587
+    },
+    {
+      "epoch": 0.852121479520553,
+      "grad_norm": 0.41210320591926575,
+      "learning_rate": 1.1262745190369694e-05,
+      "loss": 0.337,
+      "step": 11588
+    },
+    {
+      "epoch": 0.8521950143392897,
+      "grad_norm": 0.3791486918926239,
+      "learning_rate": 1.125176718227059e-05,
+      "loss": 0.3352,
+      "step": 11589
+    },
+    {
+      "epoch": 0.8522685491580263,
+      "grad_norm": 0.4499974846839905,
+      "learning_rate": 1.1240794208064908e-05,
+      "loss": 0.3509,
+      "step": 11590
+    },
+    {
+      "epoch": 0.852342083976763,
+      "grad_norm": 0.4097249507904053,
+      "learning_rate": 1.1229826268375076e-05,
+      "loss": 0.3443,
+      "step": 11591
+    },
+    {
+      "epoch": 0.8524156187954997,
+      "grad_norm": 0.38215723633766174,
+      "learning_rate": 1.121886336382315e-05,
+      "loss": 0.3296,
+      "step": 11592
+    },
+    {
+      "epoch": 0.8524891536142364,
+      "grad_norm": 0.3698875904083252,
+      "learning_rate": 1.1207905495031001e-05,
+      "loss": 0.3531,
+      "step": 11593
+    },
+    {
+      "epoch": 0.852562688432973,
+      "grad_norm": 0.4515865445137024,
+      "learning_rate": 1.1196952662620175e-05,
+      "loss": 0.3264,
+      "step": 11594
+    },
+    {
+      "epoch": 0.8526362232517097,
+      "grad_norm": 0.44103455543518066,
+      "learning_rate": 1.1186004867211886e-05,
+      "loss": 0.3304,
+      "step": 11595
+    },
+    {
+      "epoch": 0.8527097580704464,
+      "grad_norm": 0.5037939548492432,
+      "learning_rate": 1.1175062109427136e-05,
+      "loss": 0.3558,
+      "step": 11596
+    },
+    {
+      "epoch": 0.852783292889183,
+      "grad_norm": 0.43421462178230286,
+      "learning_rate": 1.116412438988662e-05,
+      "loss": 0.346,
+      "step": 11597
+    },
+    {
+      "epoch": 0.8528568277079197,
+      "grad_norm": 0.5290495753288269,
+      "learning_rate": 1.115319170921072e-05,
+      "loss": 0.3679,
+      "step": 11598
+    },
+    {
+      "epoch": 0.8529303625266563,
+      "grad_norm": 0.47010406851768494,
+      "learning_rate": 1.1142264068019537e-05,
+      "loss": 0.3447,
+      "step": 11599
+    },
+    {
+      "epoch": 0.8530038973453931,
+      "grad_norm": 0.49428075551986694,
+      "learning_rate": 1.1131341466932899e-05,
+      "loss": 0.3435,
+      "step": 11600
+    },
+    {
+      "epoch": 0.8530774321641297,
+      "grad_norm": 0.37565985321998596,
+      "learning_rate": 1.1120423906570387e-05,
+      "loss": 0.3769,
+      "step": 11601
+    },
+    {
+      "epoch": 0.8531509669828664,
+      "grad_norm": 0.5314556956291199,
+      "learning_rate": 1.1109511387551219e-05,
+      "loss": 0.3478,
+      "step": 11602
+    },
+    {
+      "epoch": 0.853224501801603,
+      "grad_norm": 0.4752013087272644,
+      "learning_rate": 1.1098603910494366e-05,
+      "loss": 0.3667,
+      "step": 11603
+    },
+    {
+      "epoch": 0.8532980366203398,
+      "grad_norm": 0.4498734772205353,
+      "learning_rate": 1.1087701476018542e-05,
+      "loss": 0.3521,
+      "step": 11604
+    },
+    {
+      "epoch": 0.8533715714390764,
+      "grad_norm": 0.3972336947917938,
+      "learning_rate": 1.1076804084742099e-05,
+      "loss": 0.3528,
+      "step": 11605
+    },
+    {
+      "epoch": 0.8534451062578131,
+      "grad_norm": 0.41322600841522217,
+      "learning_rate": 1.1065911737283175e-05,
+      "loss": 0.3355,
+      "step": 11606
+    },
+    {
+      "epoch": 0.8535186410765497,
+      "grad_norm": 0.3773898482322693,
+      "learning_rate": 1.1055024434259608e-05,
+      "loss": 0.3538,
+      "step": 11607
+    },
+    {
+      "epoch": 0.8535921758952865,
+      "grad_norm": 0.41221320629119873,
+      "learning_rate": 1.104414217628893e-05,
+      "loss": 0.3453,
+      "step": 11608
+    },
+    {
+      "epoch": 0.8536657107140231,
+      "grad_norm": 0.5639612674713135,
+      "learning_rate": 1.1033264963988366e-05,
+      "loss": 0.3567,
+      "step": 11609
+    },
+    {
+      "epoch": 0.8537392455327598,
+      "grad_norm": 0.5311180949211121,
+      "learning_rate": 1.10223927979749e-05,
+      "loss": 0.3309,
+      "step": 11610
+    },
+    {
+      "epoch": 0.8538127803514964,
+      "grad_norm": 0.36537760496139526,
+      "learning_rate": 1.1011525678865241e-05,
+      "loss": 0.3593,
+      "step": 11611
+    },
+    {
+      "epoch": 0.8538863151702331,
+      "grad_norm": 0.4034891128540039,
+      "learning_rate": 1.100066360727573e-05,
+      "loss": 0.3553,
+      "step": 11612
+    },
+    {
+      "epoch": 0.8539598499889698,
+      "grad_norm": 0.36516115069389343,
+      "learning_rate": 1.0989806583822515e-05,
+      "loss": 0.3331,
+      "step": 11613
+    },
+    {
+      "epoch": 0.8540333848077064,
+      "grad_norm": 0.4049392342567444,
+      "learning_rate": 1.0978954609121416e-05,
+      "loss": 0.3484,
+      "step": 11614
+    },
+    {
+      "epoch": 0.8541069196264431,
+      "grad_norm": 0.44317150115966797,
+      "learning_rate": 1.0968107683787953e-05,
+      "loss": 0.3548,
+      "step": 11615
+    },
+    {
+      "epoch": 0.8541804544451798,
+      "grad_norm": 0.36274492740631104,
+      "learning_rate": 1.0957265808437368e-05,
+      "loss": 0.3463,
+      "step": 11616
+    },
+    {
+      "epoch": 0.8542539892639165,
+      "grad_norm": 0.685131847858429,
+      "learning_rate": 1.0946428983684665e-05,
+      "loss": 0.3389,
+      "step": 11617
+    },
+    {
+      "epoch": 0.8543275240826531,
+      "grad_norm": 0.3767233192920685,
+      "learning_rate": 1.0935597210144465e-05,
+      "loss": 0.3536,
+      "step": 11618
+    },
+    {
+      "epoch": 0.8544010589013898,
+      "grad_norm": 0.48163893818855286,
+      "learning_rate": 1.092477048843119e-05,
+      "loss": 0.3341,
+      "step": 11619
+    },
+    {
+      "epoch": 0.8544745937201265,
+      "grad_norm": 0.4454469680786133,
+      "learning_rate": 1.0913948819158958e-05,
+      "loss": 0.3285,
+      "step": 11620
+    },
+    {
+      "epoch": 0.8545481285388632,
+      "grad_norm": 0.46909669041633606,
+      "learning_rate": 1.0903132202941569e-05,
+      "loss": 0.3869,
+      "step": 11621
+    },
+    {
+      "epoch": 0.8546216633575998,
+      "grad_norm": 0.4149765968322754,
+      "learning_rate": 1.0892320640392518e-05,
+      "loss": 0.3196,
+      "step": 11622
+    },
+    {
+      "epoch": 0.8546951981763365,
+      "grad_norm": 0.5931438207626343,
+      "learning_rate": 1.0881514132125092e-05,
+      "loss": 0.331,
+      "step": 11623
+    },
+    {
+      "epoch": 0.8547687329950732,
+      "grad_norm": 0.49381223320961,
+      "learning_rate": 1.0870712678752248e-05,
+      "loss": 0.3819,
+      "step": 11624
+    },
+    {
+      "epoch": 0.8548422678138098,
+      "grad_norm": 0.37682044506073,
+      "learning_rate": 1.0859916280886618e-05,
+      "loss": 0.3356,
+      "step": 11625
+    },
+    {
+      "epoch": 0.8549158026325465,
+      "grad_norm": 0.5056223273277283,
+      "learning_rate": 1.0849124939140621e-05,
+      "loss": 0.349,
+      "step": 11626
+    },
+    {
+      "epoch": 0.8549893374512831,
+      "grad_norm": 0.5597566366195679,
+      "learning_rate": 1.0838338654126345e-05,
+      "loss": 0.3371,
+      "step": 11627
+    },
+    {
+      "epoch": 0.8550628722700199,
+      "grad_norm": 0.38469767570495605,
+      "learning_rate": 1.0827557426455592e-05,
+      "loss": 0.3483,
+      "step": 11628
+    },
+    {
+      "epoch": 0.8551364070887565,
+      "grad_norm": 0.5038096904754639,
+      "learning_rate": 1.0816781256739883e-05,
+      "loss": 0.3753,
+      "step": 11629
+    },
+    {
+      "epoch": 0.8552099419074932,
+      "grad_norm": 0.5111382007598877,
+      "learning_rate": 1.0806010145590485e-05,
+      "loss": 0.3423,
+      "step": 11630
+    },
+    {
+      "epoch": 0.8552834767262298,
+      "grad_norm": 0.4376243054866791,
+      "learning_rate": 1.0795244093618294e-05,
+      "loss": 0.347,
+      "step": 11631
+    },
+    {
+      "epoch": 0.8553570115449666,
+      "grad_norm": 0.40918582677841187,
+      "learning_rate": 1.0784483101434028e-05,
+      "loss": 0.3236,
+      "step": 11632
+    },
+    {
+      "epoch": 0.8554305463637032,
+      "grad_norm": 0.39768803119659424,
+      "learning_rate": 1.0773727169648006e-05,
+      "loss": 0.3661,
+      "step": 11633
+    },
+    {
+      "epoch": 0.8555040811824399,
+      "grad_norm": 0.5865820050239563,
+      "learning_rate": 1.0762976298870364e-05,
+      "loss": 0.345,
+      "step": 11634
+    },
+    {
+      "epoch": 0.8555776160011765,
+      "grad_norm": 0.41905680298805237,
+      "learning_rate": 1.0752230489710869e-05,
+      "loss": 0.3371,
+      "step": 11635
+    },
+    {
+      "epoch": 0.8556511508199133,
+      "grad_norm": 0.522143542766571,
+      "learning_rate": 1.0741489742779043e-05,
+      "loss": 0.3666,
+      "step": 11636
+    },
+    {
+      "epoch": 0.8557246856386499,
+      "grad_norm": 0.45703694224357605,
+      "learning_rate": 1.0730754058684133e-05,
+      "loss": 0.3737,
+      "step": 11637
+    },
+    {
+      "epoch": 0.8557982204573866,
+      "grad_norm": 0.3971352279186249,
+      "learning_rate": 1.0720023438035042e-05,
+      "loss": 0.3274,
+      "step": 11638
+    },
+    {
+      "epoch": 0.8558717552761232,
+      "grad_norm": 0.3686651587486267,
+      "learning_rate": 1.070929788144045e-05,
+      "loss": 0.3502,
+      "step": 11639
+    },
+    {
+      "epoch": 0.8559452900948599,
+      "grad_norm": 0.39521610736846924,
+      "learning_rate": 1.0698577389508734e-05,
+      "loss": 0.3257,
+      "step": 11640
+    },
+    {
+      "epoch": 0.8560188249135966,
+      "grad_norm": 0.37875404953956604,
+      "learning_rate": 1.0687861962847933e-05,
+      "loss": 0.3225,
+      "step": 11641
+    },
+    {
+      "epoch": 0.8560923597323332,
+      "grad_norm": 0.467369019985199,
+      "learning_rate": 1.067715160206585e-05,
+      "loss": 0.3387,
+      "step": 11642
+    },
+    {
+      "epoch": 0.85616589455107,
+      "grad_norm": 0.5235664248466492,
+      "learning_rate": 1.0666446307770017e-05,
+      "loss": 0.3511,
+      "step": 11643
+    },
+    {
+      "epoch": 0.8562394293698066,
+      "grad_norm": 0.39200401306152344,
+      "learning_rate": 1.0655746080567619e-05,
+      "loss": 0.3596,
+      "step": 11644
+    },
+    {
+      "epoch": 0.8563129641885433,
+      "grad_norm": 0.47743842005729675,
+      "learning_rate": 1.064505092106558e-05,
+      "loss": 0.3428,
+      "step": 11645
+    },
+    {
+      "epoch": 0.8563864990072799,
+      "grad_norm": 0.46662652492523193,
+      "learning_rate": 1.063436082987056e-05,
+      "loss": 0.3637,
+      "step": 11646
+    },
+    {
+      "epoch": 0.8564600338260167,
+      "grad_norm": 0.4952358305454254,
+      "learning_rate": 1.0623675807588918e-05,
+      "loss": 0.3752,
+      "step": 11647
+    },
+    {
+      "epoch": 0.8565335686447533,
+      "grad_norm": 0.47353696823120117,
+      "learning_rate": 1.061299585482668e-05,
+      "loss": 0.3388,
+      "step": 11648
+    },
+    {
+      "epoch": 0.85660710346349,
+      "grad_norm": 0.46039465069770813,
+      "learning_rate": 1.0602320972189661e-05,
+      "loss": 0.3494,
+      "step": 11649
+    },
+    {
+      "epoch": 0.8566806382822266,
+      "grad_norm": 0.46498510241508484,
+      "learning_rate": 1.0591651160283355e-05,
+      "loss": 0.327,
+      "step": 11650
+    },
+    {
+      "epoch": 0.8567541731009634,
+      "grad_norm": 0.45175713300704956,
+      "learning_rate": 1.0580986419712936e-05,
+      "loss": 0.3652,
+      "step": 11651
+    },
+    {
+      "epoch": 0.8568277079197,
+      "grad_norm": 0.5680481195449829,
+      "learning_rate": 1.0570326751083337e-05,
+      "loss": 0.3427,
+      "step": 11652
+    },
+    {
+      "epoch": 0.8569012427384366,
+      "grad_norm": 0.4857522249221802,
+      "learning_rate": 1.055967215499919e-05,
+      "loss": 0.3505,
+      "step": 11653
+    },
+    {
+      "epoch": 0.8569747775571733,
+      "grad_norm": 0.40183407068252563,
+      "learning_rate": 1.054902263206482e-05,
+      "loss": 0.3587,
+      "step": 11654
+    },
+    {
+      "epoch": 0.85704831237591,
+      "grad_norm": 0.4579949676990509,
+      "learning_rate": 1.0538378182884312e-05,
+      "loss": 0.3386,
+      "step": 11655
+    },
+    {
+      "epoch": 0.8571218471946467,
+      "grad_norm": 0.5121473073959351,
+      "learning_rate": 1.0527738808061383e-05,
+      "loss": 0.3491,
+      "step": 11656
+    },
+    {
+      "epoch": 0.8571953820133833,
+      "grad_norm": 0.4476338326931,
+      "learning_rate": 1.0517104508199549e-05,
+      "loss": 0.347,
+      "step": 11657
+    },
+    {
+      "epoch": 0.85726891683212,
+      "grad_norm": 0.44555434584617615,
+      "learning_rate": 1.0506475283901963e-05,
+      "loss": 0.3316,
+      "step": 11658
+    },
+    {
+      "epoch": 0.8573424516508567,
+      "grad_norm": 0.40342822670936584,
+      "learning_rate": 1.049585113577154e-05,
+      "loss": 0.3284,
+      "step": 11659
+    },
+    {
+      "epoch": 0.8574159864695934,
+      "grad_norm": 0.4233834147453308,
+      "learning_rate": 1.0485232064410921e-05,
+      "loss": 0.3704,
+      "step": 11660
+    },
+    {
+      "epoch": 0.85748952128833,
+      "grad_norm": 0.3941044211387634,
+      "learning_rate": 1.0474618070422393e-05,
+      "loss": 0.3532,
+      "step": 11661
+    },
+    {
+      "epoch": 0.8575630561070667,
+      "grad_norm": 0.4901599586009979,
+      "learning_rate": 1.0464009154408006e-05,
+      "loss": 0.3471,
+      "step": 11662
+    },
+    {
+      "epoch": 0.8576365909258034,
+      "grad_norm": 0.45093581080436707,
+      "learning_rate": 1.0453405316969533e-05,
+      "loss": 0.3505,
+      "step": 11663
+    },
+    {
+      "epoch": 0.8577101257445401,
+      "grad_norm": 0.46090462803840637,
+      "learning_rate": 1.0442806558708384e-05,
+      "loss": 0.3303,
+      "step": 11664
+    },
+    {
+      "epoch": 0.8577836605632767,
+      "grad_norm": 0.4842780530452728,
+      "learning_rate": 1.0432212880225767e-05,
+      "loss": 0.361,
+      "step": 11665
+    },
+    {
+      "epoch": 0.8578571953820134,
+      "grad_norm": 0.42158016562461853,
+      "learning_rate": 1.0421624282122578e-05,
+      "loss": 0.3604,
+      "step": 11666
+    },
+    {
+      "epoch": 0.8579307302007501,
+      "grad_norm": 0.39036017656326294,
+      "learning_rate": 1.0411040764999402e-05,
+      "loss": 0.3459,
+      "step": 11667
+    },
+    {
+      "epoch": 0.8580042650194867,
+      "grad_norm": 0.5472516417503357,
+      "learning_rate": 1.0400462329456507e-05,
+      "loss": 0.4009,
+      "step": 11668
+    },
+    {
+      "epoch": 0.8580777998382234,
+      "grad_norm": 0.40169626474380493,
+      "learning_rate": 1.0389888976093964e-05,
+      "loss": 0.3732,
+      "step": 11669
+    },
+    {
+      "epoch": 0.85815133465696,
+      "grad_norm": 0.3736766576766968,
+      "learning_rate": 1.0379320705511498e-05,
+      "loss": 0.3518,
+      "step": 11670
+    },
+    {
+      "epoch": 0.8582248694756968,
+      "grad_norm": 0.5916483402252197,
+      "learning_rate": 1.0368757518308514e-05,
+      "loss": 0.3411,
+      "step": 11671
+    },
+    {
+      "epoch": 0.8582984042944334,
+      "grad_norm": 0.5027695894241333,
+      "learning_rate": 1.0358199415084214e-05,
+      "loss": 0.346,
+      "step": 11672
+    },
+    {
+      "epoch": 0.8583719391131701,
+      "grad_norm": 0.4481693208217621,
+      "learning_rate": 1.0347646396437448e-05,
+      "loss": 0.3671,
+      "step": 11673
+    },
+    {
+      "epoch": 0.8584454739319067,
+      "grad_norm": 0.39724957942962646,
+      "learning_rate": 1.0337098462966788e-05,
+      "loss": 0.3615,
+      "step": 11674
+    },
+    {
+      "epoch": 0.8585190087506435,
+      "grad_norm": 0.46073174476623535,
+      "learning_rate": 1.0326555615270517e-05,
+      "loss": 0.328,
+      "step": 11675
+    },
+    {
+      "epoch": 0.8585925435693801,
+      "grad_norm": 0.44855713844299316,
+      "learning_rate": 1.031601785394667e-05,
+      "loss": 0.3789,
+      "step": 11676
+    },
+    {
+      "epoch": 0.8586660783881168,
+      "grad_norm": 0.3956020176410675,
+      "learning_rate": 1.0305485179592922e-05,
+      "loss": 0.3474,
+      "step": 11677
+    },
+    {
+      "epoch": 0.8587396132068534,
+      "grad_norm": 0.41368699073791504,
+      "learning_rate": 1.0294957592806709e-05,
+      "loss": 0.3243,
+      "step": 11678
+    },
+    {
+      "epoch": 0.8588131480255902,
+      "grad_norm": 0.42898568511009216,
+      "learning_rate": 1.0284435094185185e-05,
+      "loss": 0.3647,
+      "step": 11679
+    },
+    {
+      "epoch": 0.8588866828443268,
+      "grad_norm": 0.3859142065048218,
+      "learning_rate": 1.0273917684325196e-05,
+      "loss": 0.3421,
+      "step": 11680
+    },
+    {
+      "epoch": 0.8589602176630634,
+      "grad_norm": 0.3613939583301544,
+      "learning_rate": 1.0263405363823264e-05,
+      "loss": 0.3428,
+      "step": 11681
+    },
+    {
+      "epoch": 0.8590337524818001,
+      "grad_norm": 0.42200857400894165,
+      "learning_rate": 1.0252898133275678e-05,
+      "loss": 0.3765,
+      "step": 11682
+    },
+    {
+      "epoch": 0.8591072873005368,
+      "grad_norm": 0.370128333568573,
+      "learning_rate": 1.0242395993278442e-05,
+      "loss": 0.3633,
+      "step": 11683
+    },
+    {
+      "epoch": 0.8591808221192735,
+      "grad_norm": 0.42299801111221313,
+      "learning_rate": 1.023189894442722e-05,
+      "loss": 0.3272,
+      "step": 11684
+    },
+    {
+      "epoch": 0.8592543569380101,
+      "grad_norm": 0.387885719537735,
+      "learning_rate": 1.0221406987317417e-05,
+      "loss": 0.3739,
+      "step": 11685
+    },
+    {
+      "epoch": 0.8593278917567468,
+      "grad_norm": 0.4151538610458374,
+      "learning_rate": 1.0210920122544177e-05,
+      "loss": 0.3416,
+      "step": 11686
+    },
+    {
+      "epoch": 0.8594014265754835,
+      "grad_norm": 0.5238819718360901,
+      "learning_rate": 1.0200438350702291e-05,
+      "loss": 0.3305,
+      "step": 11687
+    },
+    {
+      "epoch": 0.8594749613942202,
+      "grad_norm": 0.41887661814689636,
+      "learning_rate": 1.0189961672386306e-05,
+      "loss": 0.3298,
+      "step": 11688
+    },
+    {
+      "epoch": 0.8595484962129568,
+      "grad_norm": 0.42439112067222595,
+      "learning_rate": 1.0179490088190501e-05,
+      "loss": 0.3314,
+      "step": 11689
+    },
+    {
+      "epoch": 0.8596220310316935,
+      "grad_norm": 0.47986599802970886,
+      "learning_rate": 1.0169023598708805e-05,
+      "loss": 0.3409,
+      "step": 11690
+    },
+    {
+      "epoch": 0.8596955658504302,
+      "grad_norm": 0.38195475935935974,
+      "learning_rate": 1.0158562204534871e-05,
+      "loss": 0.3478,
+      "step": 11691
+    },
+    {
+      "epoch": 0.8597691006691669,
+      "grad_norm": 0.5010635256767273,
+      "learning_rate": 1.0148105906262106e-05,
+      "loss": 0.3352,
+      "step": 11692
+    },
+    {
+      "epoch": 0.8598426354879035,
+      "grad_norm": 0.4051930606365204,
+      "learning_rate": 1.0137654704483623e-05,
+      "loss": 0.3391,
+      "step": 11693
+    },
+    {
+      "epoch": 0.8599161703066402,
+      "grad_norm": 0.4077640175819397,
+      "learning_rate": 1.012720859979217e-05,
+      "loss": 0.3369,
+      "step": 11694
+    },
+    {
+      "epoch": 0.8599897051253769,
+      "grad_norm": 0.3938378095626831,
+      "learning_rate": 1.0116767592780297e-05,
+      "loss": 0.3784,
+      "step": 11695
+    },
+    {
+      "epoch": 0.8600632399441135,
+      "grad_norm": 0.46870291233062744,
+      "learning_rate": 1.0106331684040249e-05,
+      "loss": 0.3343,
+      "step": 11696
+    },
+    {
+      "epoch": 0.8601367747628502,
+      "grad_norm": 0.3688088357448578,
+      "learning_rate": 1.009590087416391e-05,
+      "loss": 0.3433,
+      "step": 11697
+    },
+    {
+      "epoch": 0.8602103095815868,
+      "grad_norm": 0.4561610817909241,
+      "learning_rate": 1.0085475163742953e-05,
+      "loss": 0.3524,
+      "step": 11698
+    },
+    {
+      "epoch": 0.8602838444003236,
+      "grad_norm": 0.4394064247608185,
+      "learning_rate": 1.0075054553368758e-05,
+      "loss": 0.3223,
+      "step": 11699
+    },
+    {
+      "epoch": 0.8603573792190602,
+      "grad_norm": 0.48939818143844604,
+      "learning_rate": 1.0064639043632351e-05,
+      "loss": 0.3423,
+      "step": 11700
+    },
+    {
+      "epoch": 0.8604309140377969,
+      "grad_norm": 0.46342769265174866,
+      "learning_rate": 1.0054228635124519e-05,
+      "loss": 0.3484,
+      "step": 11701
+    },
+    {
+      "epoch": 0.8605044488565335,
+      "grad_norm": 0.4091889262199402,
+      "learning_rate": 1.0043823328435797e-05,
+      "loss": 0.3426,
+      "step": 11702
+    },
+    {
+      "epoch": 0.8605779836752703,
+      "grad_norm": 0.4906935393810272,
+      "learning_rate": 1.0033423124156339e-05,
+      "loss": 0.3592,
+      "step": 11703
+    },
+    {
+      "epoch": 0.8606515184940069,
+      "grad_norm": 0.4681829810142517,
+      "learning_rate": 1.0023028022876058e-05,
+      "loss": 0.3583,
+      "step": 11704
+    },
+    {
+      "epoch": 0.8607250533127436,
+      "grad_norm": 0.44467297196388245,
+      "learning_rate": 1.0012638025184573e-05,
+      "loss": 0.3662,
+      "step": 11705
+    },
+    {
+      "epoch": 0.8607985881314802,
+      "grad_norm": 0.4561808705329895,
+      "learning_rate": 1.0002253131671257e-05,
+      "loss": 0.3242,
+      "step": 11706
+    },
+    {
+      "epoch": 0.860872122950217,
+      "grad_norm": 0.5226560235023499,
+      "learning_rate": 9.991873342925095e-06,
+      "loss": 0.3903,
+      "step": 11707
+    },
+    {
+      "epoch": 0.8609456577689536,
+      "grad_norm": 0.3925004005432129,
+      "learning_rate": 9.981498659534871e-06,
+      "loss": 0.3058,
+      "step": 11708
+    },
+    {
+      "epoch": 0.8610191925876902,
+      "grad_norm": 0.43022024631500244,
+      "learning_rate": 9.971129082089059e-06,
+      "loss": 0.3555,
+      "step": 11709
+    },
+    {
+      "epoch": 0.861092727406427,
+      "grad_norm": 0.49261513352394104,
+      "learning_rate": 9.9607646111758e-06,
+      "loss": 0.3558,
+      "step": 11710
+    },
+    {
+      "epoch": 0.8611662622251636,
+      "grad_norm": 0.48378872871398926,
+      "learning_rate": 9.950405247382988e-06,
+      "loss": 0.3586,
+      "step": 11711
+    },
+    {
+      "epoch": 0.8612397970439003,
+      "grad_norm": 0.3702485263347626,
+      "learning_rate": 9.940050991298255e-06,
+      "loss": 0.3617,
+      "step": 11712
+    },
+    {
+      "epoch": 0.8613133318626369,
+      "grad_norm": 0.5144883394241333,
+      "learning_rate": 9.929701843508843e-06,
+      "loss": 0.3281,
+      "step": 11713
+    },
+    {
+      "epoch": 0.8613868666813737,
+      "grad_norm": 0.3657110333442688,
+      "learning_rate": 9.919357804601825e-06,
+      "loss": 0.3485,
+      "step": 11714
+    },
+    {
+      "epoch": 0.8614604015001103,
+      "grad_norm": 0.4361175000667572,
+      "learning_rate": 9.909018875163867e-06,
+      "loss": 0.3334,
+      "step": 11715
+    },
+    {
+      "epoch": 0.861533936318847,
+      "grad_norm": 0.4110100567340851,
+      "learning_rate": 9.898685055781464e-06,
+      "loss": 0.3448,
+      "step": 11716
+    },
+    {
+      "epoch": 0.8616074711375836,
+      "grad_norm": 0.40096452832221985,
+      "learning_rate": 9.888356347040716e-06,
+      "loss": 0.3851,
+      "step": 11717
+    },
+    {
+      "epoch": 0.8616810059563204,
+      "grad_norm": 0.4680735766887665,
+      "learning_rate": 9.878032749527488e-06,
+      "loss": 0.3594,
+      "step": 11718
+    },
+    {
+      "epoch": 0.861754540775057,
+      "grad_norm": 0.465150386095047,
+      "learning_rate": 9.867714263827377e-06,
+      "loss": 0.3395,
+      "step": 11719
+    },
+    {
+      "epoch": 0.8618280755937937,
+      "grad_norm": 0.3898922801017761,
+      "learning_rate": 9.857400890525604e-06,
+      "loss": 0.3431,
+      "step": 11720
+    },
+    {
+      "epoch": 0.8619016104125303,
+      "grad_norm": 0.5907020568847656,
+      "learning_rate": 9.84709263020719e-06,
+      "loss": 0.3566,
+      "step": 11721
+    },
+    {
+      "epoch": 0.8619751452312671,
+      "grad_norm": 0.41254740953445435,
+      "learning_rate": 9.836789483456843e-06,
+      "loss": 0.3721,
+      "step": 11722
+    },
+    {
+      "epoch": 0.8620486800500037,
+      "grad_norm": 0.5775502324104309,
+      "learning_rate": 9.826491450858933e-06,
+      "loss": 0.3334,
+      "step": 11723
+    },
+    {
+      "epoch": 0.8621222148687403,
+      "grad_norm": 0.5671221613883972,
+      "learning_rate": 9.81619853299759e-06,
+      "loss": 0.3488,
+      "step": 11724
+    },
+    {
+      "epoch": 0.862195749687477,
+      "grad_norm": 0.39245977997779846,
+      "learning_rate": 9.805910730456658e-06,
+      "loss": 0.321,
+      "step": 11725
+    },
+    {
+      "epoch": 0.8622692845062137,
+      "grad_norm": 0.4859367311000824,
+      "learning_rate": 9.79562804381966e-06,
+      "loss": 0.3596,
+      "step": 11726
+    },
+    {
+      "epoch": 0.8623428193249504,
+      "grad_norm": 0.43072959780693054,
+      "learning_rate": 9.785350473669807e-06,
+      "loss": 0.3251,
+      "step": 11727
+    },
+    {
+      "epoch": 0.862416354143687,
+      "grad_norm": 0.36382541060447693,
+      "learning_rate": 9.775078020590089e-06,
+      "loss": 0.3325,
+      "step": 11728
+    },
+    {
+      "epoch": 0.8624898889624237,
+      "grad_norm": 0.4470174014568329,
+      "learning_rate": 9.764810685163183e-06,
+      "loss": 0.3437,
+      "step": 11729
+    },
+    {
+      "epoch": 0.8625634237811604,
+      "grad_norm": 0.38572168350219727,
+      "learning_rate": 9.754548467971425e-06,
+      "loss": 0.3247,
+      "step": 11730
+    },
+    {
+      "epoch": 0.8626369585998971,
+      "grad_norm": 0.5047200918197632,
+      "learning_rate": 9.744291369596926e-06,
+      "loss": 0.3268,
+      "step": 11731
+    },
+    {
+      "epoch": 0.8627104934186337,
+      "grad_norm": 0.4307648539543152,
+      "learning_rate": 9.734039390621475e-06,
+      "loss": 0.3573,
+      "step": 11732
+    },
+    {
+      "epoch": 0.8627840282373704,
+      "grad_norm": 0.4050631523132324,
+      "learning_rate": 9.723792531626563e-06,
+      "loss": 0.3509,
+      "step": 11733
+    },
+    {
+      "epoch": 0.8628575630561071,
+      "grad_norm": 0.7167420983314514,
+      "learning_rate": 9.713550793193405e-06,
+      "loss": 0.3516,
+      "step": 11734
+    },
+    {
+      "epoch": 0.8629310978748438,
+      "grad_norm": 0.42636430263519287,
+      "learning_rate": 9.703314175902945e-06,
+      "loss": 0.3346,
+      "step": 11735
+    },
+    {
+      "epoch": 0.8630046326935804,
+      "grad_norm": 0.43308770656585693,
+      "learning_rate": 9.693082680335785e-06,
+      "loss": 0.3526,
+      "step": 11736
+    },
+    {
+      "epoch": 0.863078167512317,
+      "grad_norm": 0.6555598974227905,
+      "learning_rate": 9.682856307072296e-06,
+      "loss": 0.3258,
+      "step": 11737
+    },
+    {
+      "epoch": 0.8631517023310538,
+      "grad_norm": 0.6386967897415161,
+      "learning_rate": 9.672635056692503e-06,
+      "loss": 0.3684,
+      "step": 11738
+    },
+    {
+      "epoch": 0.8632252371497904,
+      "grad_norm": 0.3982323110103607,
+      "learning_rate": 9.662418929776184e-06,
+      "loss": 0.3622,
+      "step": 11739
+    },
+    {
+      "epoch": 0.8632987719685271,
+      "grad_norm": 0.6369119882583618,
+      "learning_rate": 9.65220792690279e-06,
+      "loss": 0.3567,
+      "step": 11740
+    },
+    {
+      "epoch": 0.8633723067872637,
+      "grad_norm": 0.6636961698532104,
+      "learning_rate": 9.6420020486515e-06,
+      "loss": 0.3379,
+      "step": 11741
+    },
+    {
+      "epoch": 0.8634458416060005,
+      "grad_norm": 0.4532431364059448,
+      "learning_rate": 9.63180129560124e-06,
+      "loss": 0.3927,
+      "step": 11742
+    },
+    {
+      "epoch": 0.8635193764247371,
+      "grad_norm": 0.6203767657279968,
+      "learning_rate": 9.621605668330557e-06,
+      "loss": 0.3645,
+      "step": 11743
+    },
+    {
+      "epoch": 0.8635929112434738,
+      "grad_norm": 0.6391769051551819,
+      "learning_rate": 9.611415167417782e-06,
+      "loss": 0.3343,
+      "step": 11744
+    },
+    {
+      "epoch": 0.8636664460622104,
+      "grad_norm": 0.3718448877334595,
+      "learning_rate": 9.601229793440947e-06,
+      "loss": 0.345,
+      "step": 11745
+    },
+    {
+      "epoch": 0.8637399808809472,
+      "grad_norm": 0.4889183044433594,
+      "learning_rate": 9.591049546977737e-06,
+      "loss": 0.3661,
+      "step": 11746
+    },
+    {
+      "epoch": 0.8638135156996838,
+      "grad_norm": 0.5762840509414673,
+      "learning_rate": 9.580874428605624e-06,
+      "loss": 0.3361,
+      "step": 11747
+    },
+    {
+      "epoch": 0.8638870505184205,
+      "grad_norm": 0.3471860885620117,
+      "learning_rate": 9.570704438901745e-06,
+      "loss": 0.3702,
+      "step": 11748
+    },
+    {
+      "epoch": 0.8639605853371571,
+      "grad_norm": 0.45939210057258606,
+      "learning_rate": 9.56053957844294e-06,
+      "loss": 0.3449,
+      "step": 11749
+    },
+    {
+      "epoch": 0.8640341201558939,
+      "grad_norm": 0.47470173239707947,
+      "learning_rate": 9.550379847805757e-06,
+      "loss": 0.3393,
+      "step": 11750
+    },
+    {
+      "epoch": 0.8641076549746305,
+      "grad_norm": 0.522782027721405,
+      "learning_rate": 9.540225247566493e-06,
+      "loss": 0.349,
+      "step": 11751
+    },
+    {
+      "epoch": 0.8641811897933671,
+      "grad_norm": 0.44894129037857056,
+      "learning_rate": 9.530075778301128e-06,
+      "loss": 0.3169,
+      "step": 11752
+    },
+    {
+      "epoch": 0.8642547246121038,
+      "grad_norm": 0.38770222663879395,
+      "learning_rate": 9.519931440585328e-06,
+      "loss": 0.3388,
+      "step": 11753
+    },
+    {
+      "epoch": 0.8643282594308405,
+      "grad_norm": 0.4255070984363556,
+      "learning_rate": 9.509792234994509e-06,
+      "loss": 0.3513,
+      "step": 11754
+    },
+    {
+      "epoch": 0.8644017942495772,
+      "grad_norm": 0.5249758958816528,
+      "learning_rate": 9.499658162103775e-06,
+      "loss": 0.3582,
+      "step": 11755
+    },
+    {
+      "epoch": 0.8644753290683138,
+      "grad_norm": 0.49533867835998535,
+      "learning_rate": 9.489529222487936e-06,
+      "loss": 0.3518,
+      "step": 11756
+    },
+    {
+      "epoch": 0.8645488638870505,
+      "grad_norm": 0.4829957187175751,
+      "learning_rate": 9.479405416721509e-06,
+      "loss": 0.363,
+      "step": 11757
+    },
+    {
+      "epoch": 0.8646223987057872,
+      "grad_norm": 0.3537195026874542,
+      "learning_rate": 9.46928674537877e-06,
+      "loss": 0.3269,
+      "step": 11758
+    },
+    {
+      "epoch": 0.8646959335245239,
+      "grad_norm": 0.39463716745376587,
+      "learning_rate": 9.45917320903359e-06,
+      "loss": 0.3665,
+      "step": 11759
+    },
+    {
+      "epoch": 0.8647694683432605,
+      "grad_norm": 0.5001710653305054,
+      "learning_rate": 9.44906480825969e-06,
+      "loss": 0.3601,
+      "step": 11760
+    },
+    {
+      "epoch": 0.8648430031619972,
+      "grad_norm": 0.38278672099113464,
+      "learning_rate": 9.438961543630376e-06,
+      "loss": 0.3492,
+      "step": 11761
+    },
+    {
+      "epoch": 0.8649165379807339,
+      "grad_norm": 0.4019836485385895,
+      "learning_rate": 9.428863415718747e-06,
+      "loss": 0.367,
+      "step": 11762
+    },
+    {
+      "epoch": 0.8649900727994706,
+      "grad_norm": 0.48346343636512756,
+      "learning_rate": 9.418770425097555e-06,
+      "loss": 0.3614,
+      "step": 11763
+    },
+    {
+      "epoch": 0.8650636076182072,
+      "grad_norm": 0.4525662362575531,
+      "learning_rate": 9.4086825723393e-06,
+      "loss": 0.3404,
+      "step": 11764
+    },
+    {
+      "epoch": 0.8651371424369438,
+      "grad_norm": 0.3713936507701874,
+      "learning_rate": 9.398599858016189e-06,
+      "loss": 0.3357,
+      "step": 11765
+    },
+    {
+      "epoch": 0.8652106772556806,
+      "grad_norm": 0.48597729206085205,
+      "learning_rate": 9.388522282700086e-06,
+      "loss": 0.3522,
+      "step": 11766
+    },
+    {
+      "epoch": 0.8652842120744172,
+      "grad_norm": 0.4090255796909332,
+      "learning_rate": 9.378449846962634e-06,
+      "loss": 0.3315,
+      "step": 11767
+    },
+    {
+      "epoch": 0.8653577468931539,
+      "grad_norm": 0.36077880859375,
+      "learning_rate": 9.368382551375166e-06,
+      "loss": 0.3471,
+      "step": 11768
+    },
+    {
+      "epoch": 0.8654312817118905,
+      "grad_norm": 0.44577690958976746,
+      "learning_rate": 9.358320396508658e-06,
+      "loss": 0.3803,
+      "step": 11769
+    },
+    {
+      "epoch": 0.8655048165306273,
+      "grad_norm": 0.4086272716522217,
+      "learning_rate": 9.348263382933875e-06,
+      "loss": 0.36,
+      "step": 11770
+    },
+    {
+      "epoch": 0.8655783513493639,
+      "grad_norm": 0.35697469115257263,
+      "learning_rate": 9.338211511221284e-06,
+      "loss": 0.3324,
+      "step": 11771
+    },
+    {
+      "epoch": 0.8656518861681006,
+      "grad_norm": 0.36980101466178894,
+      "learning_rate": 9.328164781941018e-06,
+      "loss": 0.3689,
+      "step": 11772
+    },
+    {
+      "epoch": 0.8657254209868372,
+      "grad_norm": 0.40634211897850037,
+      "learning_rate": 9.31812319566292e-06,
+      "loss": 0.3514,
+      "step": 11773
+    },
+    {
+      "epoch": 0.865798955805574,
+      "grad_norm": 0.44337520003318787,
+      "learning_rate": 9.308086752956569e-06,
+      "loss": 0.379,
+      "step": 11774
+    },
+    {
+      "epoch": 0.8658724906243106,
+      "grad_norm": 0.36633819341659546,
+      "learning_rate": 9.298055454391264e-06,
+      "loss": 0.3181,
+      "step": 11775
+    },
+    {
+      "epoch": 0.8659460254430473,
+      "grad_norm": 0.38993680477142334,
+      "learning_rate": 9.28802930053596e-06,
+      "loss": 0.3199,
+      "step": 11776
+    },
+    {
+      "epoch": 0.866019560261784,
+      "grad_norm": 0.3784288465976715,
+      "learning_rate": 9.278008291959362e-06,
+      "loss": 0.3707,
+      "step": 11777
+    },
+    {
+      "epoch": 0.8660930950805207,
+      "grad_norm": 0.4117378890514374,
+      "learning_rate": 9.267992429229899e-06,
+      "loss": 0.3429,
+      "step": 11778
+    },
+    {
+      "epoch": 0.8661666298992573,
+      "grad_norm": 0.49541017413139343,
+      "learning_rate": 9.257981712915642e-06,
+      "loss": 0.3378,
+      "step": 11779
+    },
+    {
+      "epoch": 0.8662401647179939,
+      "grad_norm": 0.4168902635574341,
+      "learning_rate": 9.247976143584425e-06,
+      "loss": 0.3782,
+      "step": 11780
+    },
+    {
+      "epoch": 0.8663136995367307,
+      "grad_norm": 0.38520753383636475,
+      "learning_rate": 9.237975721803793e-06,
+      "loss": 0.3406,
+      "step": 11781
+    },
+    {
+      "epoch": 0.8663872343554673,
+      "grad_norm": 0.4498429596424103,
+      "learning_rate": 9.227980448140949e-06,
+      "loss": 0.3358,
+      "step": 11782
+    },
+    {
+      "epoch": 0.866460769174204,
+      "grad_norm": 0.6520821452140808,
+      "learning_rate": 9.217990323162839e-06,
+      "loss": 0.3341,
+      "step": 11783
+    },
+    {
+      "epoch": 0.8665343039929406,
+      "grad_norm": 0.4503716230392456,
+      "learning_rate": 9.208005347436155e-06,
+      "loss": 0.351,
+      "step": 11784
+    },
+    {
+      "epoch": 0.8666078388116774,
+      "grad_norm": 0.698358416557312,
+      "learning_rate": 9.19802552152722e-06,
+      "loss": 0.3271,
+      "step": 11785
+    },
+    {
+      "epoch": 0.866681373630414,
+      "grad_norm": 0.5462785959243774,
+      "learning_rate": 9.188050846002084e-06,
+      "loss": 0.3711,
+      "step": 11786
+    },
+    {
+      "epoch": 0.8667549084491507,
+      "grad_norm": 0.45685136318206787,
+      "learning_rate": 9.178081321426534e-06,
+      "loss": 0.3458,
+      "step": 11787
+    },
+    {
+      "epoch": 0.8668284432678873,
+      "grad_norm": 0.5556853413581848,
+      "learning_rate": 9.16811694836609e-06,
+      "loss": 0.3555,
+      "step": 11788
+    },
+    {
+      "epoch": 0.8669019780866241,
+      "grad_norm": 0.8548439145088196,
+      "learning_rate": 9.158157727385885e-06,
+      "loss": 0.3616,
+      "step": 11789
+    },
+    {
+      "epoch": 0.8669755129053607,
+      "grad_norm": 0.39372754096984863,
+      "learning_rate": 9.148203659050847e-06,
+      "loss": 0.3461,
+      "step": 11790
+    },
+    {
+      "epoch": 0.8670490477240974,
+      "grad_norm": 0.5908892750740051,
+      "learning_rate": 9.138254743925601e-06,
+      "loss": 0.366,
+      "step": 11791
+    },
+    {
+      "epoch": 0.867122582542834,
+      "grad_norm": 0.6148998737335205,
+      "learning_rate": 9.128310982574407e-06,
+      "loss": 0.3388,
+      "step": 11792
+    },
+    {
+      "epoch": 0.8671961173615707,
+      "grad_norm": 0.39479053020477295,
+      "learning_rate": 9.118372375561325e-06,
+      "loss": 0.3472,
+      "step": 11793
+    },
+    {
+      "epoch": 0.8672696521803074,
+      "grad_norm": 0.4591132700443268,
+      "learning_rate": 9.10843892345008e-06,
+      "loss": 0.3447,
+      "step": 11794
+    },
+    {
+      "epoch": 0.867343186999044,
+      "grad_norm": 0.37825605273246765,
+      "learning_rate": 9.098510626804113e-06,
+      "loss": 0.3524,
+      "step": 11795
+    },
+    {
+      "epoch": 0.8674167218177807,
+      "grad_norm": 0.406710147857666,
+      "learning_rate": 9.088587486186528e-06,
+      "loss": 0.3404,
+      "step": 11796
+    },
+    {
+      "epoch": 0.8674902566365174,
+      "grad_norm": 0.45297643542289734,
+      "learning_rate": 9.078669502160208e-06,
+      "loss": 0.3567,
+      "step": 11797
+    },
+    {
+      "epoch": 0.8675637914552541,
+      "grad_norm": 0.37946754693984985,
+      "learning_rate": 9.068756675287727e-06,
+      "loss": 0.3362,
+      "step": 11798
+    },
+    {
+      "epoch": 0.8676373262739907,
+      "grad_norm": 0.578582227230072,
+      "learning_rate": 9.058849006131308e-06,
+      "loss": 0.3543,
+      "step": 11799
+    },
+    {
+      "epoch": 0.8677108610927274,
+      "grad_norm": 0.3989822566509247,
+      "learning_rate": 9.048946495252952e-06,
+      "loss": 0.3547,
+      "step": 11800
+    },
+    {
+      "epoch": 0.8677843959114641,
+      "grad_norm": 0.39064913988113403,
+      "learning_rate": 9.03904914321435e-06,
+      "loss": 0.3726,
+      "step": 11801
+    },
+    {
+      "epoch": 0.8678579307302008,
+      "grad_norm": 0.5049368739128113,
+      "learning_rate": 9.029156950576856e-06,
+      "loss": 0.3765,
+      "step": 11802
+    },
+    {
+      "epoch": 0.8679314655489374,
+      "grad_norm": 0.4878169000148773,
+      "learning_rate": 9.019269917901585e-06,
+      "loss": 0.3292,
+      "step": 11803
+    },
+    {
+      "epoch": 0.8680050003676741,
+      "grad_norm": 0.5359594821929932,
+      "learning_rate": 9.009388045749357e-06,
+      "loss": 0.3407,
+      "step": 11804
+    },
+    {
+      "epoch": 0.8680785351864108,
+      "grad_norm": 0.4582373797893524,
+      "learning_rate": 8.999511334680655e-06,
+      "loss": 0.3557,
+      "step": 11805
+    },
+    {
+      "epoch": 0.8681520700051475,
+      "grad_norm": 0.6053265929222107,
+      "learning_rate": 8.9896397852557e-06,
+      "loss": 0.3455,
+      "step": 11806
+    },
+    {
+      "epoch": 0.8682256048238841,
+      "grad_norm": 0.5540361404418945,
+      "learning_rate": 8.97977339803444e-06,
+      "loss": 0.334,
+      "step": 11807
+    },
+    {
+      "epoch": 0.8682991396426207,
+      "grad_norm": 0.3584483563899994,
+      "learning_rate": 8.9699121735765e-06,
+      "loss": 0.3261,
+      "step": 11808
+    },
+    {
+      "epoch": 0.8683726744613575,
+      "grad_norm": 0.556209146976471,
+      "learning_rate": 8.960056112441185e-06,
+      "loss": 0.3383,
+      "step": 11809
+    },
+    {
+      "epoch": 0.8684462092800941,
+      "grad_norm": 0.4977859854698181,
+      "learning_rate": 8.950205215187568e-06,
+      "loss": 0.3459,
+      "step": 11810
+    },
+    {
+      "epoch": 0.8685197440988308,
+      "grad_norm": 0.37616267800331116,
+      "learning_rate": 8.940359482374416e-06,
+      "loss": 0.3914,
+      "step": 11811
+    },
+    {
+      "epoch": 0.8685932789175674,
+      "grad_norm": 0.4658297598361969,
+      "learning_rate": 8.930518914560148e-06,
+      "loss": 0.3567,
+      "step": 11812
+    },
+    {
+      "epoch": 0.8686668137363042,
+      "grad_norm": 0.49521544575691223,
+      "learning_rate": 8.920683512302963e-06,
+      "loss": 0.3625,
+      "step": 11813
+    },
+    {
+      "epoch": 0.8687403485550408,
+      "grad_norm": 0.5149791836738586,
+      "learning_rate": 8.910853276160747e-06,
+      "loss": 0.3548,
+      "step": 11814
+    },
+    {
+      "epoch": 0.8688138833737775,
+      "grad_norm": 0.4154618978500366,
+      "learning_rate": 8.901028206691054e-06,
+      "loss": 0.3246,
+      "step": 11815
+    },
+    {
+      "epoch": 0.8688874181925141,
+      "grad_norm": 0.42796090245246887,
+      "learning_rate": 8.89120830445117e-06,
+      "loss": 0.3529,
+      "step": 11816
+    },
+    {
+      "epoch": 0.8689609530112509,
+      "grad_norm": 0.4776318073272705,
+      "learning_rate": 8.88139356999812e-06,
+      "loss": 0.3435,
+      "step": 11817
+    },
+    {
+      "epoch": 0.8690344878299875,
+      "grad_norm": 0.3742252588272095,
+      "learning_rate": 8.871584003888579e-06,
+      "loss": 0.3297,
+      "step": 11818
+    },
+    {
+      "epoch": 0.8691080226487242,
+      "grad_norm": 0.5106236934661865,
+      "learning_rate": 8.861779606678978e-06,
+      "loss": 0.3438,
+      "step": 11819
+    },
+    {
+      "epoch": 0.8691815574674608,
+      "grad_norm": 0.6051926016807556,
+      "learning_rate": 8.851980378925394e-06,
+      "loss": 0.3654,
+      "step": 11820
+    },
+    {
+      "epoch": 0.8692550922861975,
+      "grad_norm": 0.46891871094703674,
+      "learning_rate": 8.842186321183699e-06,
+      "loss": 0.337,
+      "step": 11821
+    },
+    {
+      "epoch": 0.8693286271049342,
+      "grad_norm": 0.3880816698074341,
+      "learning_rate": 8.832397434009388e-06,
+      "loss": 0.3672,
+      "step": 11822
+    },
+    {
+      "epoch": 0.8694021619236708,
+      "grad_norm": 0.4761076271533966,
+      "learning_rate": 8.822613717957696e-06,
+      "loss": 0.3308,
+      "step": 11823
+    },
+    {
+      "epoch": 0.8694756967424075,
+      "grad_norm": 0.5859849452972412,
+      "learning_rate": 8.8128351735836e-06,
+      "loss": 0.361,
+      "step": 11824
+    },
+    {
+      "epoch": 0.8695492315611442,
+      "grad_norm": 0.3709428012371063,
+      "learning_rate": 8.803061801441714e-06,
+      "loss": 0.3448,
+      "step": 11825
+    },
+    {
+      "epoch": 0.8696227663798809,
+      "grad_norm": 0.3715069890022278,
+      "learning_rate": 8.793293602086405e-06,
+      "loss": 0.3305,
+      "step": 11826
+    },
+    {
+      "epoch": 0.8696963011986175,
+      "grad_norm": 0.4986799955368042,
+      "learning_rate": 8.783530576071753e-06,
+      "loss": 0.3391,
+      "step": 11827
+    },
+    {
+      "epoch": 0.8697698360173542,
+      "grad_norm": 0.6204872131347656,
+      "learning_rate": 8.77377272395149e-06,
+      "loss": 0.3573,
+      "step": 11828
+    },
+    {
+      "epoch": 0.8698433708360909,
+      "grad_norm": 0.4025377333164215,
+      "learning_rate": 8.76402004627911e-06,
+      "loss": 0.3572,
+      "step": 11829
+    },
+    {
+      "epoch": 0.8699169056548276,
+      "grad_norm": 0.3793197274208069,
+      "learning_rate": 8.754272543607823e-06,
+      "loss": 0.3604,
+      "step": 11830
+    },
+    {
+      "epoch": 0.8699904404735642,
+      "grad_norm": 0.3879438042640686,
+      "learning_rate": 8.744530216490487e-06,
+      "loss": 0.3373,
+      "step": 11831
+    },
+    {
+      "epoch": 0.870063975292301,
+      "grad_norm": 0.37006622552871704,
+      "learning_rate": 8.734793065479696e-06,
+      "loss": 0.3459,
+      "step": 11832
+    },
+    {
+      "epoch": 0.8701375101110376,
+      "grad_norm": 0.4092656373977661,
+      "learning_rate": 8.72506109112775e-06,
+      "loss": 0.3458,
+      "step": 11833
+    },
+    {
+      "epoch": 0.8702110449297743,
+      "grad_norm": 0.5704792737960815,
+      "learning_rate": 8.715334293986676e-06,
+      "loss": 0.3658,
+      "step": 11834
+    },
+    {
+      "epoch": 0.8702845797485109,
+      "grad_norm": 0.5253576636314392,
+      "learning_rate": 8.705612674608177e-06,
+      "loss": 0.3364,
+      "step": 11835
+    },
+    {
+      "epoch": 0.8703581145672475,
+      "grad_norm": 0.39522427320480347,
+      "learning_rate": 8.695896233543665e-06,
+      "loss": 0.3351,
+      "step": 11836
+    },
+    {
+      "epoch": 0.8704316493859843,
+      "grad_norm": 0.44416776299476624,
+      "learning_rate": 8.686184971344292e-06,
+      "loss": 0.3364,
+      "step": 11837
+    },
+    {
+      "epoch": 0.8705051842047209,
+      "grad_norm": 0.45316657423973083,
+      "learning_rate": 8.676478888560857e-06,
+      "loss": 0.3588,
+      "step": 11838
+    },
+    {
+      "epoch": 0.8705787190234576,
+      "grad_norm": 0.4180791974067688,
+      "learning_rate": 8.666777985743912e-06,
+      "loss": 0.3304,
+      "step": 11839
+    },
+    {
+      "epoch": 0.8706522538421942,
+      "grad_norm": 0.374308317899704,
+      "learning_rate": 8.657082263443738e-06,
+      "loss": 0.3413,
+      "step": 11840
+    },
+    {
+      "epoch": 0.870725788660931,
+      "grad_norm": 0.43190786242485046,
+      "learning_rate": 8.64739172221023e-06,
+      "loss": 0.3235,
+      "step": 11841
+    },
+    {
+      "epoch": 0.8707993234796676,
+      "grad_norm": 0.4298003315925598,
+      "learning_rate": 8.63770636259309e-06,
+      "loss": 0.3335,
+      "step": 11842
+    },
+    {
+      "epoch": 0.8708728582984043,
+      "grad_norm": 0.38617366552352905,
+      "learning_rate": 8.628026185141647e-06,
+      "loss": 0.3471,
+      "step": 11843
+    },
+    {
+      "epoch": 0.870946393117141,
+      "grad_norm": 0.35723254084587097,
+      "learning_rate": 8.618351190405005e-06,
+      "loss": 0.3242,
+      "step": 11844
+    },
+    {
+      "epoch": 0.8710199279358777,
+      "grad_norm": 0.3927322328090668,
+      "learning_rate": 8.608681378931904e-06,
+      "loss": 0.3331,
+      "step": 11845
+    },
+    {
+      "epoch": 0.8710934627546143,
+      "grad_norm": 0.41386619210243225,
+      "learning_rate": 8.59901675127085e-06,
+      "loss": 0.3547,
+      "step": 11846
+    },
+    {
+      "epoch": 0.871166997573351,
+      "grad_norm": 0.37991175055503845,
+      "learning_rate": 8.589357307970036e-06,
+      "loss": 0.334,
+      "step": 11847
+    },
+    {
+      "epoch": 0.8712405323920877,
+      "grad_norm": 0.42558997869491577,
+      "learning_rate": 8.579703049577326e-06,
+      "loss": 0.345,
+      "step": 11848
+    },
+    {
+      "epoch": 0.8713140672108243,
+      "grad_norm": 0.38217824697494507,
+      "learning_rate": 8.570053976640346e-06,
+      "loss": 0.3532,
+      "step": 11849
+    },
+    {
+      "epoch": 0.871387602029561,
+      "grad_norm": 0.4287147521972656,
+      "learning_rate": 8.560410089706406e-06,
+      "loss": 0.3501,
+      "step": 11850
+    },
+    {
+      "epoch": 0.8714611368482976,
+      "grad_norm": 0.3816411793231964,
+      "learning_rate": 8.550771389322487e-06,
+      "loss": 0.3362,
+      "step": 11851
+    },
+    {
+      "epoch": 0.8715346716670344,
+      "grad_norm": 0.3819057047367096,
+      "learning_rate": 8.541137876035331e-06,
+      "loss": 0.3457,
+      "step": 11852
+    },
+    {
+      "epoch": 0.871608206485771,
+      "grad_norm": 0.45218631625175476,
+      "learning_rate": 8.53150955039137e-06,
+      "loss": 0.3389,
+      "step": 11853
+    },
+    {
+      "epoch": 0.8716817413045077,
+      "grad_norm": 0.45766985416412354,
+      "learning_rate": 8.52188641293672e-06,
+      "loss": 0.3438,
+      "step": 11854
+    },
+    {
+      "epoch": 0.8717552761232443,
+      "grad_norm": 0.5150686502456665,
+      "learning_rate": 8.512268464217188e-06,
+      "loss": 0.3331,
+      "step": 11855
+    },
+    {
+      "epoch": 0.8718288109419811,
+      "grad_norm": 0.3920876085758209,
+      "learning_rate": 8.50265570477835e-06,
+      "loss": 0.3388,
+      "step": 11856
+    },
+    {
+      "epoch": 0.8719023457607177,
+      "grad_norm": 0.36150914430618286,
+      "learning_rate": 8.49304813516545e-06,
+      "loss": 0.3431,
+      "step": 11857
+    },
+    {
+      "epoch": 0.8719758805794544,
+      "grad_norm": 0.47831466794013977,
+      "learning_rate": 8.483445755923414e-06,
+      "loss": 0.3558,
+      "step": 11858
+    },
+    {
+      "epoch": 0.872049415398191,
+      "grad_norm": 0.4538459777832031,
+      "learning_rate": 8.47384856759692e-06,
+      "loss": 0.3551,
+      "step": 11859
+    },
+    {
+      "epoch": 0.8721229502169278,
+      "grad_norm": 0.3849053978919983,
+      "learning_rate": 8.464256570730345e-06,
+      "loss": 0.3389,
+      "step": 11860
+    },
+    {
+      "epoch": 0.8721964850356644,
+      "grad_norm": 0.4236401915550232,
+      "learning_rate": 8.454669765867718e-06,
+      "loss": 0.3451,
+      "step": 11861
+    },
+    {
+      "epoch": 0.8722700198544011,
+      "grad_norm": 0.746968150138855,
+      "learning_rate": 8.445088153552838e-06,
+      "loss": 0.3684,
+      "step": 11862
+    },
+    {
+      "epoch": 0.8723435546731377,
+      "grad_norm": 0.352904349565506,
+      "learning_rate": 8.435511734329194e-06,
+      "loss": 0.3392,
+      "step": 11863
+    },
+    {
+      "epoch": 0.8724170894918744,
+      "grad_norm": 0.6682392954826355,
+      "learning_rate": 8.425940508739938e-06,
+      "loss": 0.3676,
+      "step": 11864
+    },
+    {
+      "epoch": 0.8724906243106111,
+      "grad_norm": 0.5539925694465637,
+      "learning_rate": 8.416374477327992e-06,
+      "loss": 0.3632,
+      "step": 11865
+    },
+    {
+      "epoch": 0.8725641591293477,
+      "grad_norm": 0.3987671434879303,
+      "learning_rate": 8.406813640635935e-06,
+      "loss": 0.3606,
+      "step": 11866
+    },
+    {
+      "epoch": 0.8726376939480844,
+      "grad_norm": 0.4651033282279968,
+      "learning_rate": 8.397257999206076e-06,
+      "loss": 0.3332,
+      "step": 11867
+    },
+    {
+      "epoch": 0.8727112287668211,
+      "grad_norm": 0.5878357291221619,
+      "learning_rate": 8.387707553580392e-06,
+      "loss": 0.3516,
+      "step": 11868
+    },
+    {
+      "epoch": 0.8727847635855578,
+      "grad_norm": 0.405377060174942,
+      "learning_rate": 8.37816230430063e-06,
+      "loss": 0.3457,
+      "step": 11869
+    },
+    {
+      "epoch": 0.8728582984042944,
+      "grad_norm": 0.4173547923564911,
+      "learning_rate": 8.368622251908208e-06,
+      "loss": 0.3592,
+      "step": 11870
+    },
+    {
+      "epoch": 0.8729318332230311,
+      "grad_norm": 0.5511661171913147,
+      "learning_rate": 8.35908739694422e-06,
+      "loss": 0.3658,
+      "step": 11871
+    },
+    {
+      "epoch": 0.8730053680417678,
+      "grad_norm": 0.5951762199401855,
+      "learning_rate": 8.349557739949498e-06,
+      "loss": 0.339,
+      "step": 11872
+    },
+    {
+      "epoch": 0.8730789028605045,
+      "grad_norm": 0.46378132700920105,
+      "learning_rate": 8.34003328146461e-06,
+      "loss": 0.3766,
+      "step": 11873
+    },
+    {
+      "epoch": 0.8731524376792411,
+      "grad_norm": 0.547805905342102,
+      "learning_rate": 8.330514022029746e-06,
+      "loss": 0.355,
+      "step": 11874
+    },
+    {
+      "epoch": 0.8732259724979778,
+      "grad_norm": 0.3726697862148285,
+      "learning_rate": 8.320999962184872e-06,
+      "loss": 0.4199,
+      "step": 11875
+    },
+    {
+      "epoch": 0.8732995073167145,
+      "grad_norm": 0.39825239777565,
+      "learning_rate": 8.311491102469649e-06,
+      "loss": 0.3323,
+      "step": 11876
+    },
+    {
+      "epoch": 0.8733730421354511,
+      "grad_norm": 0.5470010638237,
+      "learning_rate": 8.301987443423409e-06,
+      "loss": 0.3999,
+      "step": 11877
+    },
+    {
+      "epoch": 0.8734465769541878,
+      "grad_norm": 0.41378581523895264,
+      "learning_rate": 8.2924889855852e-06,
+      "loss": 0.3383,
+      "step": 11878
+    },
+    {
+      "epoch": 0.8735201117729244,
+      "grad_norm": 0.4149772822856903,
+      "learning_rate": 8.282995729493803e-06,
+      "loss": 0.3721,
+      "step": 11879
+    },
+    {
+      "epoch": 0.8735936465916612,
+      "grad_norm": 0.4971606731414795,
+      "learning_rate": 8.273507675687686e-06,
+      "loss": 0.3557,
+      "step": 11880
+    },
+    {
+      "epoch": 0.8736671814103978,
+      "grad_norm": 0.5187757015228271,
+      "learning_rate": 8.264024824705008e-06,
+      "loss": 0.3562,
+      "step": 11881
+    },
+    {
+      "epoch": 0.8737407162291345,
+      "grad_norm": 0.43794408440589905,
+      "learning_rate": 8.25454717708366e-06,
+      "loss": 0.3276,
+      "step": 11882
+    },
+    {
+      "epoch": 0.8738142510478711,
+      "grad_norm": 0.4683322608470917,
+      "learning_rate": 8.245074733361225e-06,
+      "loss": 0.3689,
+      "step": 11883
+    },
+    {
+      "epoch": 0.8738877858666079,
+      "grad_norm": 0.3892103433609009,
+      "learning_rate": 8.23560749407497e-06,
+      "loss": 0.3359,
+      "step": 11884
+    },
+    {
+      "epoch": 0.8739613206853445,
+      "grad_norm": 0.5490850210189819,
+      "learning_rate": 8.226145459761903e-06,
+      "loss": 0.3611,
+      "step": 11885
+    },
+    {
+      "epoch": 0.8740348555040812,
+      "grad_norm": 0.4388161599636078,
+      "learning_rate": 8.216688630958735e-06,
+      "loss": 0.3806,
+      "step": 11886
+    },
+    {
+      "epoch": 0.8741083903228178,
+      "grad_norm": 0.3731834888458252,
+      "learning_rate": 8.20723700820184e-06,
+      "loss": 0.3425,
+      "step": 11887
+    },
+    {
+      "epoch": 0.8741819251415546,
+      "grad_norm": 0.45444539189338684,
+      "learning_rate": 8.197790592027321e-06,
+      "loss": 0.342,
+      "step": 11888
+    },
+    {
+      "epoch": 0.8742554599602912,
+      "grad_norm": 0.39765027165412903,
+      "learning_rate": 8.188349382971027e-06,
+      "loss": 0.3418,
+      "step": 11889
+    },
+    {
+      "epoch": 0.8743289947790279,
+      "grad_norm": 0.5166404247283936,
+      "learning_rate": 8.178913381568454e-06,
+      "loss": 0.3551,
+      "step": 11890
+    },
+    {
+      "epoch": 0.8744025295977645,
+      "grad_norm": 0.39476168155670166,
+      "learning_rate": 8.169482588354794e-06,
+      "loss": 0.3418,
+      "step": 11891
+    },
+    {
+      "epoch": 0.8744760644165012,
+      "grad_norm": 0.4511842727661133,
+      "learning_rate": 8.160057003864985e-06,
+      "loss": 0.3384,
+      "step": 11892
+    },
+    {
+      "epoch": 0.8745495992352379,
+      "grad_norm": 0.6297795176506042,
+      "learning_rate": 8.1506366286337e-06,
+      "loss": 0.3591,
+      "step": 11893
+    },
+    {
+      "epoch": 0.8746231340539745,
+      "grad_norm": 0.4614957869052887,
+      "learning_rate": 8.141221463195204e-06,
+      "loss": 0.3357,
+      "step": 11894
+    },
+    {
+      "epoch": 0.8746966688727112,
+      "grad_norm": 0.3691481649875641,
+      "learning_rate": 8.131811508083587e-06,
+      "loss": 0.3378,
+      "step": 11895
+    },
+    {
+      "epoch": 0.8747702036914479,
+      "grad_norm": 0.42892783880233765,
+      "learning_rate": 8.122406763832579e-06,
+      "loss": 0.3734,
+      "step": 11896
+    },
+    {
+      "epoch": 0.8748437385101846,
+      "grad_norm": 0.3867032825946808,
+      "learning_rate": 8.1130072309756e-06,
+      "loss": 0.3788,
+      "step": 11897
+    },
+    {
+      "epoch": 0.8749172733289212,
+      "grad_norm": 0.5825932621955872,
+      "learning_rate": 8.103612910045833e-06,
+      "loss": 0.3643,
+      "step": 11898
+    },
+    {
+      "epoch": 0.874990808147658,
+      "grad_norm": 0.410986065864563,
+      "learning_rate": 8.094223801576139e-06,
+      "loss": 0.3165,
+      "step": 11899
+    },
+    {
+      "epoch": 0.8750643429663946,
+      "grad_norm": 0.4633553922176361,
+      "learning_rate": 8.084839906099039e-06,
+      "loss": 0.3535,
+      "step": 11900
+    },
+    {
+      "epoch": 0.8751378777851313,
+      "grad_norm": 0.45364144444465637,
+      "learning_rate": 8.075461224146852e-06,
+      "loss": 0.3584,
+      "step": 11901
+    },
+    {
+      "epoch": 0.8752114126038679,
+      "grad_norm": 0.4459924101829529,
+      "learning_rate": 8.066087756251483e-06,
+      "loss": 0.3599,
+      "step": 11902
+    },
+    {
+      "epoch": 0.8752849474226047,
+      "grad_norm": 0.5133794546127319,
+      "learning_rate": 8.056719502944665e-06,
+      "loss": 0.3685,
+      "step": 11903
+    },
+    {
+      "epoch": 0.8753584822413413,
+      "grad_norm": 0.391314834356308,
+      "learning_rate": 8.047356464757727e-06,
+      "loss": 0.3631,
+      "step": 11904
+    },
+    {
+      "epoch": 0.8754320170600779,
+      "grad_norm": 0.3789602816104889,
+      "learning_rate": 8.037998642221778e-06,
+      "loss": 0.3358,
+      "step": 11905
+    },
+    {
+      "epoch": 0.8755055518788146,
+      "grad_norm": 0.39660143852233887,
+      "learning_rate": 8.028646035867616e-06,
+      "loss": 0.3616,
+      "step": 11906
+    },
+    {
+      "epoch": 0.8755790866975512,
+      "grad_norm": 0.4018411934375763,
+      "learning_rate": 8.019298646225682e-06,
+      "loss": 0.332,
+      "step": 11907
+    },
+    {
+      "epoch": 0.875652621516288,
+      "grad_norm": 0.45628300309181213,
+      "learning_rate": 8.009956473826208e-06,
+      "loss": 0.3517,
+      "step": 11908
+    },
+    {
+      "epoch": 0.8757261563350246,
+      "grad_norm": 0.4412451386451721,
+      "learning_rate": 8.000619519199104e-06,
+      "loss": 0.3701,
+      "step": 11909
+    },
+    {
+      "epoch": 0.8757996911537613,
+      "grad_norm": 0.4255230128765106,
+      "learning_rate": 7.991287782873924e-06,
+      "loss": 0.3624,
+      "step": 11910
+    },
+    {
+      "epoch": 0.875873225972498,
+      "grad_norm": 0.4564383625984192,
+      "learning_rate": 7.98196126538001e-06,
+      "loss": 0.3726,
+      "step": 11911
+    },
+    {
+      "epoch": 0.8759467607912347,
+      "grad_norm": 0.44699472188949585,
+      "learning_rate": 7.972639967246386e-06,
+      "loss": 0.3688,
+      "step": 11912
+    },
+    {
+      "epoch": 0.8760202956099713,
+      "grad_norm": 0.4407024383544922,
+      "learning_rate": 7.96332388900174e-06,
+      "loss": 0.3594,
+      "step": 11913
+    },
+    {
+      "epoch": 0.876093830428708,
+      "grad_norm": 0.3979945778846741,
+      "learning_rate": 7.954013031174468e-06,
+      "loss": 0.343,
+      "step": 11914
+    },
+    {
+      "epoch": 0.8761673652474447,
+      "grad_norm": 0.5059375762939453,
+      "learning_rate": 7.944707394292717e-06,
+      "loss": 0.358,
+      "step": 11915
+    },
+    {
+      "epoch": 0.8762409000661814,
+      "grad_norm": 0.5040556192398071,
+      "learning_rate": 7.93540697888434e-06,
+      "loss": 0.3678,
+      "step": 11916
+    },
+    {
+      "epoch": 0.876314434884918,
+      "grad_norm": 0.4652264416217804,
+      "learning_rate": 7.926111785476808e-06,
+      "loss": 0.3704,
+      "step": 11917
+    },
+    {
+      "epoch": 0.8763879697036547,
+      "grad_norm": 0.3900601863861084,
+      "learning_rate": 7.916821814597397e-06,
+      "loss": 0.3335,
+      "step": 11918
+    },
+    {
+      "epoch": 0.8764615045223914,
+      "grad_norm": 0.42179566621780396,
+      "learning_rate": 7.90753706677304e-06,
+      "loss": 0.3427,
+      "step": 11919
+    },
+    {
+      "epoch": 0.876535039341128,
+      "grad_norm": 0.39669737219810486,
+      "learning_rate": 7.898257542530364e-06,
+      "loss": 0.3363,
+      "step": 11920
+    },
+    {
+      "epoch": 0.8766085741598647,
+      "grad_norm": 0.5271782279014587,
+      "learning_rate": 7.888983242395709e-06,
+      "loss": 0.3306,
+      "step": 11921
+    },
+    {
+      "epoch": 0.8766821089786013,
+      "grad_norm": 0.3765432834625244,
+      "learning_rate": 7.879714166895157e-06,
+      "loss": 0.3477,
+      "step": 11922
+    },
+    {
+      "epoch": 0.8767556437973381,
+      "grad_norm": 0.3797338306903839,
+      "learning_rate": 7.870450316554412e-06,
+      "loss": 0.3287,
+      "step": 11923
+    },
+    {
+      "epoch": 0.8768291786160747,
+      "grad_norm": 0.4343793988227844,
+      "learning_rate": 7.861191691898972e-06,
+      "loss": 0.3375,
+      "step": 11924
+    },
+    {
+      "epoch": 0.8769027134348114,
+      "grad_norm": 0.4398646056652069,
+      "learning_rate": 7.851938293453964e-06,
+      "loss": 0.3253,
+      "step": 11925
+    },
+    {
+      "epoch": 0.876976248253548,
+      "grad_norm": 0.3760644793510437,
+      "learning_rate": 7.842690121744279e-06,
+      "loss": 0.349,
+      "step": 11926
+    },
+    {
+      "epoch": 0.8770497830722848,
+      "grad_norm": 0.39946165680885315,
+      "learning_rate": 7.83344717729445e-06,
+      "loss": 0.3226,
+      "step": 11927
+    },
+    {
+      "epoch": 0.8771233178910214,
+      "grad_norm": 0.4100654721260071,
+      "learning_rate": 7.824209460628761e-06,
+      "loss": 0.3115,
+      "step": 11928
+    },
+    {
+      "epoch": 0.8771968527097581,
+      "grad_norm": 0.44861575961112976,
+      "learning_rate": 7.814976972271215e-06,
+      "loss": 0.3407,
+      "step": 11929
+    },
+    {
+      "epoch": 0.8772703875284947,
+      "grad_norm": 0.5293706655502319,
+      "learning_rate": 7.805749712745436e-06,
+      "loss": 0.3467,
+      "step": 11930
+    },
+    {
+      "epoch": 0.8773439223472315,
+      "grad_norm": 0.3615826368331909,
+      "learning_rate": 7.796527682574839e-06,
+      "loss": 0.3187,
+      "step": 11931
+    },
+    {
+      "epoch": 0.8774174571659681,
+      "grad_norm": 0.455220103263855,
+      "learning_rate": 7.787310882282506e-06,
+      "loss": 0.3353,
+      "step": 11932
+    },
+    {
+      "epoch": 0.8774909919847047,
+      "grad_norm": 0.497521311044693,
+      "learning_rate": 7.778099312391196e-06,
+      "loss": 0.339,
+      "step": 11933
+    },
+    {
+      "epoch": 0.8775645268034414,
+      "grad_norm": 0.39748406410217285,
+      "learning_rate": 7.768892973423425e-06,
+      "loss": 0.346,
+      "step": 11934
+    },
+    {
+      "epoch": 0.8776380616221781,
+      "grad_norm": 0.4943825900554657,
+      "learning_rate": 7.759691865901397e-06,
+      "loss": 0.3701,
+      "step": 11935
+    },
+    {
+      "epoch": 0.8777115964409148,
+      "grad_norm": 0.4357167184352875,
+      "learning_rate": 7.750495990346996e-06,
+      "loss": 0.3254,
+      "step": 11936
+    },
+    {
+      "epoch": 0.8777851312596514,
+      "grad_norm": 0.38730722665786743,
+      "learning_rate": 7.74130534728179e-06,
+      "loss": 0.3278,
+      "step": 11937
+    },
+    {
+      "epoch": 0.8778586660783881,
+      "grad_norm": 0.4688190817832947,
+      "learning_rate": 7.732119937227111e-06,
+      "loss": 0.3848,
+      "step": 11938
+    },
+    {
+      "epoch": 0.8779322008971248,
+      "grad_norm": 0.40982893109321594,
+      "learning_rate": 7.722939760703985e-06,
+      "loss": 0.3455,
+      "step": 11939
+    },
+    {
+      "epoch": 0.8780057357158615,
+      "grad_norm": 0.38695666193962097,
+      "learning_rate": 7.713764818233082e-06,
+      "loss": 0.3656,
+      "step": 11940
+    },
+    {
+      "epoch": 0.8780792705345981,
+      "grad_norm": 0.4137521982192993,
+      "learning_rate": 7.704595110334834e-06,
+      "loss": 0.3449,
+      "step": 11941
+    },
+    {
+      "epoch": 0.8781528053533348,
+      "grad_norm": 0.4021596312522888,
+      "learning_rate": 7.695430637529366e-06,
+      "loss": 0.3426,
+      "step": 11942
+    },
+    {
+      "epoch": 0.8782263401720715,
+      "grad_norm": 0.4135356545448303,
+      "learning_rate": 7.686271400336465e-06,
+      "loss": 0.3332,
+      "step": 11943
+    },
+    {
+      "epoch": 0.8782998749908082,
+      "grad_norm": 0.3864297568798065,
+      "learning_rate": 7.67711739927568e-06,
+      "loss": 0.3115,
+      "step": 11944
+    },
+    {
+      "epoch": 0.8783734098095448,
+      "grad_norm": 0.4438827335834503,
+      "learning_rate": 7.66796863486624e-06,
+      "loss": 0.3114,
+      "step": 11945
+    },
+    {
+      "epoch": 0.8784469446282815,
+      "grad_norm": 0.37835678458213806,
+      "learning_rate": 7.658825107627054e-06,
+      "loss": 0.3366,
+      "step": 11946
+    },
+    {
+      "epoch": 0.8785204794470182,
+      "grad_norm": 0.3597061336040497,
+      "learning_rate": 7.649686818076762e-06,
+      "loss": 0.3127,
+      "step": 11947
+    },
+    {
+      "epoch": 0.8785940142657548,
+      "grad_norm": 0.5391159653663635,
+      "learning_rate": 7.640553766733693e-06,
+      "loss": 0.3188,
+      "step": 11948
+    },
+    {
+      "epoch": 0.8786675490844915,
+      "grad_norm": 0.39080578088760376,
+      "learning_rate": 7.631425954115889e-06,
+      "loss": 0.3281,
+      "step": 11949
+    },
+    {
+      "epoch": 0.8787410839032281,
+      "grad_norm": 0.39101582765579224,
+      "learning_rate": 7.622303380741081e-06,
+      "loss": 0.3155,
+      "step": 11950
+    },
+    {
+      "epoch": 0.8788146187219649,
+      "grad_norm": 0.37373584508895874,
+      "learning_rate": 7.613186047126708e-06,
+      "loss": 0.3384,
+      "step": 11951
+    },
+    {
+      "epoch": 0.8788881535407015,
+      "grad_norm": 0.36467671394348145,
+      "learning_rate": 7.604073953789937e-06,
+      "loss": 0.3364,
+      "step": 11952
+    },
+    {
+      "epoch": 0.8789616883594382,
+      "grad_norm": 0.3992622196674347,
+      "learning_rate": 7.594967101247585e-06,
+      "loss": 0.3237,
+      "step": 11953
+    },
+    {
+      "epoch": 0.8790352231781748,
+      "grad_norm": 0.38031184673309326,
+      "learning_rate": 7.5858654900162195e-06,
+      "loss": 0.3334,
+      "step": 11954
+    },
+    {
+      "epoch": 0.8791087579969116,
+      "grad_norm": 0.5141035914421082,
+      "learning_rate": 7.576769120612104e-06,
+      "loss": 0.3237,
+      "step": 11955
+    },
+    {
+      "epoch": 0.8791822928156482,
+      "grad_norm": 0.38259953260421753,
+      "learning_rate": 7.56767799355117e-06,
+      "loss": 0.3311,
+      "step": 11956
+    },
+    {
+      "epoch": 0.8792558276343849,
+      "grad_norm": 0.41355785727500916,
+      "learning_rate": 7.5585921093490945e-06,
+      "loss": 0.3521,
+      "step": 11957
+    },
+    {
+      "epoch": 0.8793293624531215,
+      "grad_norm": 0.46931886672973633,
+      "learning_rate": 7.549511468521231e-06,
+      "loss": 0.3343,
+      "step": 11958
+    },
+    {
+      "epoch": 0.8794028972718583,
+      "grad_norm": 0.3826460540294647,
+      "learning_rate": 7.540436071582657e-06,
+      "loss": 0.3181,
+      "step": 11959
+    },
+    {
+      "epoch": 0.8794764320905949,
+      "grad_norm": 0.45294350385665894,
+      "learning_rate": 7.531365919048106e-06,
+      "loss": 0.3365,
+      "step": 11960
+    },
+    {
+      "epoch": 0.8795499669093315,
+      "grad_norm": 0.4032197892665863,
+      "learning_rate": 7.522301011432065e-06,
+      "loss": 0.3136,
+      "step": 11961
+    },
+    {
+      "epoch": 0.8796235017280682,
+      "grad_norm": 0.42257654666900635,
+      "learning_rate": 7.513241349248723e-06,
+      "loss": 0.3221,
+      "step": 11962
+    },
+    {
+      "epoch": 0.8796970365468049,
+      "grad_norm": 0.49633681774139404,
+      "learning_rate": 7.504186933011925e-06,
+      "loss": 0.3361,
+      "step": 11963
+    },
+    {
+      "epoch": 0.8797705713655416,
+      "grad_norm": 0.5137966275215149,
+      "learning_rate": 7.49513776323526e-06,
+      "loss": 0.3478,
+      "step": 11964
+    },
+    {
+      "epoch": 0.8798441061842782,
+      "grad_norm": 0.38185349106788635,
+      "learning_rate": 7.486093840432018e-06,
+      "loss": 0.3494,
+      "step": 11965
+    },
+    {
+      "epoch": 0.879917641003015,
+      "grad_norm": 0.4164541959762573,
+      "learning_rate": 7.477055165115143e-06,
+      "loss": 0.3001,
+      "step": 11966
+    },
+    {
+      "epoch": 0.8799911758217516,
+      "grad_norm": 0.46922504901885986,
+      "learning_rate": 7.468021737797348e-06,
+      "loss": 0.3365,
+      "step": 11967
+    },
+    {
+      "epoch": 0.8800647106404883,
+      "grad_norm": 0.4168310761451721,
+      "learning_rate": 7.458993558991034e-06,
+      "loss": 0.3196,
+      "step": 11968
+    },
+    {
+      "epoch": 0.8801382454592249,
+      "grad_norm": 0.43542298674583435,
+      "learning_rate": 7.449970629208247e-06,
+      "loss": 0.3039,
+      "step": 11969
+    },
+    {
+      "epoch": 0.8802117802779617,
+      "grad_norm": 0.4098599851131439,
+      "learning_rate": 7.4409529489608e-06,
+      "loss": 0.3257,
+      "step": 11970
+    },
+    {
+      "epoch": 0.8802853150966983,
+      "grad_norm": 0.426445335149765,
+      "learning_rate": 7.431940518760194e-06,
+      "loss": 0.3272,
+      "step": 11971
+    },
+    {
+      "epoch": 0.880358849915435,
+      "grad_norm": 0.4472188949584961,
+      "learning_rate": 7.422933339117621e-06,
+      "loss": 0.3514,
+      "step": 11972
+    },
+    {
+      "epoch": 0.8804323847341716,
+      "grad_norm": 0.3913955092430115,
+      "learning_rate": 7.413931410543951e-06,
+      "loss": 0.3527,
+      "step": 11973
+    },
+    {
+      "epoch": 0.8805059195529084,
+      "grad_norm": 0.39254945516586304,
+      "learning_rate": 7.404934733549795e-06,
+      "loss": 0.3424,
+      "step": 11974
+    },
+    {
+      "epoch": 0.880579454371645,
+      "grad_norm": 0.4167914092540741,
+      "learning_rate": 7.395943308645481e-06,
+      "loss": 0.3428,
+      "step": 11975
+    },
+    {
+      "epoch": 0.8806529891903816,
+      "grad_norm": 0.46875467896461487,
+      "learning_rate": 7.3869571363409775e-06,
+      "loss": 0.3164,
+      "step": 11976
+    },
+    {
+      "epoch": 0.8807265240091183,
+      "grad_norm": 0.4275713860988617,
+      "learning_rate": 7.377976217145999e-06,
+      "loss": 0.349,
+      "step": 11977
+    },
+    {
+      "epoch": 0.880800058827855,
+      "grad_norm": 0.43901315331459045,
+      "learning_rate": 7.369000551569982e-06,
+      "loss": 0.3272,
+      "step": 11978
+    },
+    {
+      "epoch": 0.8808735936465917,
+      "grad_norm": 0.5164607763290405,
+      "learning_rate": 7.3600301401219875e-06,
+      "loss": 0.3151,
+      "step": 11979
+    },
+    {
+      "epoch": 0.8809471284653283,
+      "grad_norm": 0.4153292179107666,
+      "learning_rate": 7.351064983310851e-06,
+      "loss": 0.3237,
+      "step": 11980
+    },
+    {
+      "epoch": 0.881020663284065,
+      "grad_norm": 0.41227641701698303,
+      "learning_rate": 7.342105081645101e-06,
+      "loss": 0.3547,
+      "step": 11981
+    },
+    {
+      "epoch": 0.8810941981028017,
+      "grad_norm": 0.4247814416885376,
+      "learning_rate": 7.3331504356329405e-06,
+      "loss": 0.3639,
+      "step": 11982
+    },
+    {
+      "epoch": 0.8811677329215384,
+      "grad_norm": 0.5062201023101807,
+      "learning_rate": 7.324201045782264e-06,
+      "loss": 0.3399,
+      "step": 11983
+    },
+    {
+      "epoch": 0.881241267740275,
+      "grad_norm": 0.43635404109954834,
+      "learning_rate": 7.3152569126006985e-06,
+      "loss": 0.3497,
+      "step": 11984
+    },
+    {
+      "epoch": 0.8813148025590117,
+      "grad_norm": 0.4164849519729614,
+      "learning_rate": 7.306318036595605e-06,
+      "loss": 0.3507,
+      "step": 11985
+    },
+    {
+      "epoch": 0.8813883373777484,
+      "grad_norm": 0.5592852234840393,
+      "learning_rate": 7.297384418273956e-06,
+      "loss": 0.3415,
+      "step": 11986
+    },
+    {
+      "epoch": 0.8814618721964851,
+      "grad_norm": 0.4824437201023102,
+      "learning_rate": 7.288456058142501e-06,
+      "loss": 0.3221,
+      "step": 11987
+    },
+    {
+      "epoch": 0.8815354070152217,
+      "grad_norm": 0.41129055619239807,
+      "learning_rate": 7.279532956707669e-06,
+      "loss": 0.343,
+      "step": 11988
+    },
+    {
+      "epoch": 0.8816089418339583,
+      "grad_norm": 0.4752427637577057,
+      "learning_rate": 7.270615114475565e-06,
+      "loss": 0.3706,
+      "step": 11989
+    },
+    {
+      "epoch": 0.8816824766526951,
+      "grad_norm": 0.4790986180305481,
+      "learning_rate": 7.2617025319520306e-06,
+      "loss": 0.3354,
+      "step": 11990
+    },
+    {
+      "epoch": 0.8817560114714317,
+      "grad_norm": 0.3932183086872101,
+      "learning_rate": 7.252795209642615e-06,
+      "loss": 0.3202,
+      "step": 11991
+    },
+    {
+      "epoch": 0.8818295462901684,
+      "grad_norm": 0.4033098816871643,
+      "learning_rate": 7.243893148052527e-06,
+      "loss": 0.3391,
+      "step": 11992
+    },
+    {
+      "epoch": 0.881903081108905,
+      "grad_norm": 0.39659810066223145,
+      "learning_rate": 7.234996347686706e-06,
+      "loss": 0.3441,
+      "step": 11993
+    },
+    {
+      "epoch": 0.8819766159276418,
+      "grad_norm": 0.47721508145332336,
+      "learning_rate": 7.226104809049805e-06,
+      "loss": 0.3495,
+      "step": 11994
+    },
+    {
+      "epoch": 0.8820501507463784,
+      "grad_norm": 0.53035569190979,
+      "learning_rate": 7.217218532646141e-06,
+      "loss": 0.3375,
+      "step": 11995
+    },
+    {
+      "epoch": 0.8821236855651151,
+      "grad_norm": 0.5276694297790527,
+      "learning_rate": 7.208337518979757e-06,
+      "loss": 0.3237,
+      "step": 11996
+    },
+    {
+      "epoch": 0.8821972203838517,
+      "grad_norm": 0.4250532388687134,
+      "learning_rate": 7.199461768554383e-06,
+      "loss": 0.3343,
+      "step": 11997
+    },
+    {
+      "epoch": 0.8822707552025885,
+      "grad_norm": 0.4192359447479248,
+      "learning_rate": 7.190591281873505e-06,
+      "loss": 0.3224,
+      "step": 11998
+    },
+    {
+      "epoch": 0.8823442900213251,
+      "grad_norm": 0.5085755586624146,
+      "learning_rate": 7.181726059440208e-06,
+      "loss": 0.3344,
+      "step": 11999
+    },
+    {
+      "epoch": 0.8824178248400618,
+      "grad_norm": 0.6104419231414795,
+      "learning_rate": 7.1728661017573586e-06,
+      "loss": 0.3386,
+      "step": 12000
+    },
+    {
+      "epoch": 0.8824913596587984,
+      "grad_norm": 0.3966170847415924,
+      "learning_rate": 7.164011409327531e-06,
+      "loss": 0.316,
+      "step": 12001
+    },
+    {
+      "epoch": 0.8825648944775352,
+      "grad_norm": 0.45335644483566284,
+      "learning_rate": 7.155161982652936e-06,
+      "loss": 0.3148,
+      "step": 12002
+    },
+    {
+      "epoch": 0.8826384292962718,
+      "grad_norm": 0.6760190725326538,
+      "learning_rate": 7.146317822235538e-06,
+      "loss": 0.3216,
+      "step": 12003
+    },
+    {
+      "epoch": 0.8827119641150084,
+      "grad_norm": 0.5280448198318481,
+      "learning_rate": 7.137478928577002e-06,
+      "loss": 0.3399,
+      "step": 12004
+    },
+    {
+      "epoch": 0.8827854989337451,
+      "grad_norm": 0.4391649067401886,
+      "learning_rate": 7.1286453021786384e-06,
+      "loss": 0.3346,
+      "step": 12005
+    },
+    {
+      "epoch": 0.8828590337524818,
+      "grad_norm": 0.486470103263855,
+      "learning_rate": 7.1198169435415464e-06,
+      "loss": 0.375,
+      "step": 12006
+    },
+    {
+      "epoch": 0.8829325685712185,
+      "grad_norm": 0.49235033988952637,
+      "learning_rate": 7.110993853166437e-06,
+      "loss": 0.3309,
+      "step": 12007
+    },
+    {
+      "epoch": 0.8830061033899551,
+      "grad_norm": 0.41948843002319336,
+      "learning_rate": 7.102176031553809e-06,
+      "loss": 0.3199,
+      "step": 12008
+    },
+    {
+      "epoch": 0.8830796382086918,
+      "grad_norm": 0.5047727227210999,
+      "learning_rate": 7.093363479203774e-06,
+      "loss": 0.3583,
+      "step": 12009
+    },
+    {
+      "epoch": 0.8831531730274285,
+      "grad_norm": 0.42658740282058716,
+      "learning_rate": 7.08455619661621e-06,
+      "loss": 0.3556,
+      "step": 12010
+    },
+    {
+      "epoch": 0.8832267078461652,
+      "grad_norm": 0.5572884678840637,
+      "learning_rate": 7.075754184290695e-06,
+      "loss": 0.3851,
+      "step": 12011
+    },
+    {
+      "epoch": 0.8833002426649018,
+      "grad_norm": 0.41190382838249207,
+      "learning_rate": 7.066957442726463e-06,
+      "loss": 0.3292,
+      "step": 12012
+    },
+    {
+      "epoch": 0.8833737774836385,
+      "grad_norm": 0.4654456377029419,
+      "learning_rate": 7.058165972422471e-06,
+      "loss": 0.3638,
+      "step": 12013
+    },
+    {
+      "epoch": 0.8834473123023752,
+      "grad_norm": 0.38303378224372864,
+      "learning_rate": 7.049379773877407e-06,
+      "loss": 0.3563,
+      "step": 12014
+    },
+    {
+      "epoch": 0.8835208471211119,
+      "grad_norm": 0.5252692103385925,
+      "learning_rate": 7.0405988475895965e-06,
+      "loss": 0.341,
+      "step": 12015
+    },
+    {
+      "epoch": 0.8835943819398485,
+      "grad_norm": 0.4002484083175659,
+      "learning_rate": 7.03182319405713e-06,
+      "loss": 0.357,
+      "step": 12016
+    },
+    {
+      "epoch": 0.8836679167585851,
+      "grad_norm": 0.5036580562591553,
+      "learning_rate": 7.023052813777775e-06,
+      "loss": 0.3792,
+      "step": 12017
+    },
+    {
+      "epoch": 0.8837414515773219,
+      "grad_norm": 0.4220319986343384,
+      "learning_rate": 7.01428770724899e-06,
+      "loss": 0.3414,
+      "step": 12018
+    },
+    {
+      "epoch": 0.8838149863960585,
+      "grad_norm": 0.41410133242607117,
+      "learning_rate": 7.005527874967921e-06,
+      "loss": 0.3428,
+      "step": 12019
+    },
+    {
+      "epoch": 0.8838885212147952,
+      "grad_norm": 0.765674889087677,
+      "learning_rate": 6.99677331743146e-06,
+      "loss": 0.3278,
+      "step": 12020
+    },
+    {
+      "epoch": 0.8839620560335318,
+      "grad_norm": 0.42060166597366333,
+      "learning_rate": 6.988024035136176e-06,
+      "loss": 0.3286,
+      "step": 12021
+    },
+    {
+      "epoch": 0.8840355908522686,
+      "grad_norm": 0.5415433645248413,
+      "learning_rate": 6.979280028578305e-06,
+      "loss": 0.3358,
+      "step": 12022
+    },
+    {
+      "epoch": 0.8841091256710052,
+      "grad_norm": 0.6564646363258362,
+      "learning_rate": 6.970541298253852e-06,
+      "loss": 0.3246,
+      "step": 12023
+    },
+    {
+      "epoch": 0.8841826604897419,
+      "grad_norm": 0.4923926293849945,
+      "learning_rate": 6.961807844658485e-06,
+      "loss": 0.3159,
+      "step": 12024
+    },
+    {
+      "epoch": 0.8842561953084785,
+      "grad_norm": 0.4130760431289673,
+      "learning_rate": 6.953079668287543e-06,
+      "loss": 0.3416,
+      "step": 12025
+    },
+    {
+      "epoch": 0.8843297301272153,
+      "grad_norm": 0.3838265836238861,
+      "learning_rate": 6.944356769636129e-06,
+      "loss": 0.3067,
+      "step": 12026
+    },
+    {
+      "epoch": 0.8844032649459519,
+      "grad_norm": 0.3868262469768524,
+      "learning_rate": 6.9356391491990135e-06,
+      "loss": 0.327,
+      "step": 12027
+    },
+    {
+      "epoch": 0.8844767997646886,
+      "grad_norm": 0.46829646825790405,
+      "learning_rate": 6.926926807470646e-06,
+      "loss": 0.3302,
+      "step": 12028
+    },
+    {
+      "epoch": 0.8845503345834252,
+      "grad_norm": 0.5239717960357666,
+      "learning_rate": 6.918219744945231e-06,
+      "loss": 0.328,
+      "step": 12029
+    },
+    {
+      "epoch": 0.884623869402162,
+      "grad_norm": 0.4413701593875885,
+      "learning_rate": 6.909517962116607e-06,
+      "loss": 0.3375,
+      "step": 12030
+    },
+    {
+      "epoch": 0.8846974042208986,
+      "grad_norm": 0.42454400658607483,
+      "learning_rate": 6.90082145947838e-06,
+      "loss": 0.3202,
+      "step": 12031
+    },
+    {
+      "epoch": 0.8847709390396352,
+      "grad_norm": 0.4949451982975006,
+      "learning_rate": 6.8921302375237975e-06,
+      "loss": 0.3328,
+      "step": 12032
+    },
+    {
+      "epoch": 0.884844473858372,
+      "grad_norm": 0.4506978988647461,
+      "learning_rate": 6.8834442967458555e-06,
+      "loss": 0.3481,
+      "step": 12033
+    },
+    {
+      "epoch": 0.8849180086771086,
+      "grad_norm": 0.41035255789756775,
+      "learning_rate": 6.874763637637227e-06,
+      "loss": 0.3326,
+      "step": 12034
+    },
+    {
+      "epoch": 0.8849915434958453,
+      "grad_norm": 0.36542418599128723,
+      "learning_rate": 6.866088260690273e-06,
+      "loss": 0.3059,
+      "step": 12035
+    },
+    {
+      "epoch": 0.8850650783145819,
+      "grad_norm": 0.4569353461265564,
+      "learning_rate": 6.85741816639709e-06,
+      "loss": 0.3222,
+      "step": 12036
+    },
+    {
+      "epoch": 0.8851386131333187,
+      "grad_norm": 0.45289117097854614,
+      "learning_rate": 6.8487533552494485e-06,
+      "loss": 0.3409,
+      "step": 12037
+    },
+    {
+      "epoch": 0.8852121479520553,
+      "grad_norm": 0.40406593680381775,
+      "learning_rate": 6.840093827738814e-06,
+      "loss": 0.3454,
+      "step": 12038
+    },
+    {
+      "epoch": 0.885285682770792,
+      "grad_norm": 0.4090568423271179,
+      "learning_rate": 6.83143958435637e-06,
+      "loss": 0.3343,
+      "step": 12039
+    },
+    {
+      "epoch": 0.8853592175895286,
+      "grad_norm": 0.4959743916988373,
+      "learning_rate": 6.822790625593023e-06,
+      "loss": 0.335,
+      "step": 12040
+    },
+    {
+      "epoch": 0.8854327524082654,
+      "grad_norm": 0.6682865619659424,
+      "learning_rate": 6.814146951939315e-06,
+      "loss": 0.3387,
+      "step": 12041
+    },
+    {
+      "epoch": 0.885506287227002,
+      "grad_norm": 0.4943106472492218,
+      "learning_rate": 6.8055085638855206e-06,
+      "loss": 0.3362,
+      "step": 12042
+    },
+    {
+      "epoch": 0.8855798220457387,
+      "grad_norm": 0.4861561059951782,
+      "learning_rate": 6.796875461921637e-06,
+      "loss": 0.3292,
+      "step": 12043
+    },
+    {
+      "epoch": 0.8856533568644753,
+      "grad_norm": 0.5039260983467102,
+      "learning_rate": 6.788247646537349e-06,
+      "loss": 0.3217,
+      "step": 12044
+    },
+    {
+      "epoch": 0.885726891683212,
+      "grad_norm": 0.45456767082214355,
+      "learning_rate": 6.779625118222011e-06,
+      "loss": 0.3235,
+      "step": 12045
+    },
+    {
+      "epoch": 0.8858004265019487,
+      "grad_norm": 0.3842422664165497,
+      "learning_rate": 6.771007877464708e-06,
+      "loss": 0.3149,
+      "step": 12046
+    },
+    {
+      "epoch": 0.8858739613206853,
+      "grad_norm": 0.3588144779205322,
+      "learning_rate": 6.7623959247542505e-06,
+      "loss": 0.3382,
+      "step": 12047
+    },
+    {
+      "epoch": 0.885947496139422,
+      "grad_norm": 0.40052539110183716,
+      "learning_rate": 6.75378926057908e-06,
+      "loss": 0.3405,
+      "step": 12048
+    },
+    {
+      "epoch": 0.8860210309581587,
+      "grad_norm": 0.4175138473510742,
+      "learning_rate": 6.745187885427384e-06,
+      "loss": 0.2976,
+      "step": 12049
+    },
+    {
+      "epoch": 0.8860945657768954,
+      "grad_norm": 0.46870550513267517,
+      "learning_rate": 6.736591799787073e-06,
+      "loss": 0.3377,
+      "step": 12050
+    },
+    {
+      "epoch": 0.886168100595632,
+      "grad_norm": 0.4669215977191925,
+      "learning_rate": 6.728001004145667e-06,
+      "loss": 0.3164,
+      "step": 12051
+    },
+    {
+      "epoch": 0.8862416354143687,
+      "grad_norm": 0.4453912079334259,
+      "learning_rate": 6.7194154989905115e-06,
+      "loss": 0.3315,
+      "step": 12052
+    },
+    {
+      "epoch": 0.8863151702331054,
+      "grad_norm": 0.43860122561454773,
+      "learning_rate": 6.710835284808525e-06,
+      "loss": 0.3325,
+      "step": 12053
+    },
+    {
+      "epoch": 0.8863887050518421,
+      "grad_norm": 0.544008195400238,
+      "learning_rate": 6.702260362086432e-06,
+      "loss": 0.3251,
+      "step": 12054
+    },
+    {
+      "epoch": 0.8864622398705787,
+      "grad_norm": 0.47985556721687317,
+      "learning_rate": 6.693690731310576e-06,
+      "loss": 0.3322,
+      "step": 12055
+    },
+    {
+      "epoch": 0.8865357746893154,
+      "grad_norm": 0.44952622056007385,
+      "learning_rate": 6.6851263929670446e-06,
+      "loss": 0.3461,
+      "step": 12056
+    },
+    {
+      "epoch": 0.8866093095080521,
+      "grad_norm": 0.45619863271713257,
+      "learning_rate": 6.676567347541651e-06,
+      "loss": 0.3264,
+      "step": 12057
+    },
+    {
+      "epoch": 0.8866828443267888,
+      "grad_norm": 0.4121030569076538,
+      "learning_rate": 6.668013595519829e-06,
+      "loss": 0.3356,
+      "step": 12058
+    },
+    {
+      "epoch": 0.8867563791455254,
+      "grad_norm": 0.3952648639678955,
+      "learning_rate": 6.659465137386767e-06,
+      "loss": 0.3808,
+      "step": 12059
+    },
+    {
+      "epoch": 0.886829913964262,
+      "grad_norm": 0.4059963524341583,
+      "learning_rate": 6.650921973627366e-06,
+      "loss": 0.3256,
+      "step": 12060
+    },
+    {
+      "epoch": 0.8869034487829988,
+      "grad_norm": 0.4439367651939392,
+      "learning_rate": 6.642384104726173e-06,
+      "loss": 0.3507,
+      "step": 12061
+    },
+    {
+      "epoch": 0.8869769836017354,
+      "grad_norm": 0.48976460099220276,
+      "learning_rate": 6.6338515311674675e-06,
+      "loss": 0.3172,
+      "step": 12062
+    },
+    {
+      "epoch": 0.8870505184204721,
+      "grad_norm": 0.8236969113349915,
+      "learning_rate": 6.625324253435261e-06,
+      "loss": 0.3114,
+      "step": 12063
+    },
+    {
+      "epoch": 0.8871240532392087,
+      "grad_norm": 0.4669771194458008,
+      "learning_rate": 6.616802272013201e-06,
+      "loss": 0.3221,
+      "step": 12064
+    },
+    {
+      "epoch": 0.8871975880579455,
+      "grad_norm": 0.3727920353412628,
+      "learning_rate": 6.608285587384655e-06,
+      "loss": 0.362,
+      "step": 12065
+    },
+    {
+      "epoch": 0.8872711228766821,
+      "grad_norm": 0.40282076597213745,
+      "learning_rate": 6.599774200032715e-06,
+      "loss": 0.3524,
+      "step": 12066
+    },
+    {
+      "epoch": 0.8873446576954188,
+      "grad_norm": 0.36843088269233704,
+      "learning_rate": 6.591268110440163e-06,
+      "loss": 0.3317,
+      "step": 12067
+    },
+    {
+      "epoch": 0.8874181925141554,
+      "grad_norm": 0.3981854319572449,
+      "learning_rate": 6.582767319089445e-06,
+      "loss": 0.3185,
+      "step": 12068
+    },
+    {
+      "epoch": 0.8874917273328922,
+      "grad_norm": 0.5312016606330872,
+      "learning_rate": 6.574271826462753e-06,
+      "loss": 0.3131,
+      "step": 12069
+    },
+    {
+      "epoch": 0.8875652621516288,
+      "grad_norm": 0.4556480944156647,
+      "learning_rate": 6.565781633041968e-06,
+      "loss": 0.3458,
+      "step": 12070
+    },
+    {
+      "epoch": 0.8876387969703655,
+      "grad_norm": 0.4729515612125397,
+      "learning_rate": 6.5572967393086495e-06,
+      "loss": 0.3236,
+      "step": 12071
+    },
+    {
+      "epoch": 0.8877123317891021,
+      "grad_norm": 0.3940280079841614,
+      "learning_rate": 6.548817145744068e-06,
+      "loss": 0.3026,
+      "step": 12072
+    },
+    {
+      "epoch": 0.8877858666078388,
+      "grad_norm": 0.49151188135147095,
+      "learning_rate": 6.540342852829218e-06,
+      "loss": 0.3315,
+      "step": 12073
+    },
+    {
+      "epoch": 0.8878594014265755,
+      "grad_norm": 0.3770902454853058,
+      "learning_rate": 6.531873861044735e-06,
+      "loss": 0.3342,
+      "step": 12074
+    },
+    {
+      "epoch": 0.8879329362453121,
+      "grad_norm": 0.47363612055778503,
+      "learning_rate": 6.523410170871014e-06,
+      "loss": 0.3269,
+      "step": 12075
+    },
+    {
+      "epoch": 0.8880064710640488,
+      "grad_norm": 0.44505301117897034,
+      "learning_rate": 6.514951782788125e-06,
+      "loss": 0.3222,
+      "step": 12076
+    },
+    {
+      "epoch": 0.8880800058827855,
+      "grad_norm": 0.38828587532043457,
+      "learning_rate": 6.506498697275831e-06,
+      "loss": 0.3338,
+      "step": 12077
+    },
+    {
+      "epoch": 0.8881535407015222,
+      "grad_norm": 0.34711384773254395,
+      "learning_rate": 6.498050914813591e-06,
+      "loss": 0.3426,
+      "step": 12078
+    },
+    {
+      "epoch": 0.8882270755202588,
+      "grad_norm": 0.39520955085754395,
+      "learning_rate": 6.489608435880568e-06,
+      "loss": 0.3229,
+      "step": 12079
+    },
+    {
+      "epoch": 0.8883006103389955,
+      "grad_norm": 0.5820334553718567,
+      "learning_rate": 6.481171260955665e-06,
+      "loss": 0.3396,
+      "step": 12080
+    },
+    {
+      "epoch": 0.8883741451577322,
+      "grad_norm": 0.5003470182418823,
+      "learning_rate": 6.4727393905174014e-06,
+      "loss": 0.3514,
+      "step": 12081
+    },
+    {
+      "epoch": 0.8884476799764689,
+      "grad_norm": 0.39697059988975525,
+      "learning_rate": 6.464312825044072e-06,
+      "loss": 0.3134,
+      "step": 12082
+    },
+    {
+      "epoch": 0.8885212147952055,
+      "grad_norm": 0.5527603626251221,
+      "learning_rate": 6.455891565013639e-06,
+      "loss": 0.3284,
+      "step": 12083
+    },
+    {
+      "epoch": 0.8885947496139422,
+      "grad_norm": 0.5512741208076477,
+      "learning_rate": 6.447475610903742e-06,
+      "loss": 0.3182,
+      "step": 12084
+    },
+    {
+      "epoch": 0.8886682844326789,
+      "grad_norm": 0.6610158085823059,
+      "learning_rate": 6.439064963191755e-06,
+      "loss": 0.3093,
+      "step": 12085
+    },
+    {
+      "epoch": 0.8887418192514156,
+      "grad_norm": 0.3784783184528351,
+      "learning_rate": 6.430659622354762e-06,
+      "loss": 0.298,
+      "step": 12086
+    },
+    {
+      "epoch": 0.8888153540701522,
+      "grad_norm": 0.48848778009414673,
+      "learning_rate": 6.422259588869506e-06,
+      "loss": 0.3237,
+      "step": 12087
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 0.7071486711502075,
+      "learning_rate": 6.413864863212415e-06,
+      "loss": 0.3109,
+      "step": 12088
+    },
+    {
+      "epoch": 0.8889624237076256,
+      "grad_norm": 0.4567631781101227,
+      "learning_rate": 6.405475445859688e-06,
+      "loss": 0.3218,
+      "step": 12089
+    },
+    {
+      "epoch": 0.8890359585263622,
+      "grad_norm": 0.45196911692619324,
+      "learning_rate": 6.397091337287176e-06,
+      "loss": 0.3134,
+      "step": 12090
+    },
+    {
+      "epoch": 0.8891094933450989,
+      "grad_norm": 0.7193071842193604,
+      "learning_rate": 6.388712537970421e-06,
+      "loss": 0.3515,
+      "step": 12091
+    },
+    {
+      "epoch": 0.8891830281638355,
+      "grad_norm": 0.45182666182518005,
+      "learning_rate": 6.380339048384665e-06,
+      "loss": 0.3459,
+      "step": 12092
+    },
+    {
+      "epoch": 0.8892565629825723,
+      "grad_norm": 0.4258079528808594,
+      "learning_rate": 6.371970869004906e-06,
+      "loss": 0.3294,
+      "step": 12093
+    },
+    {
+      "epoch": 0.8893300978013089,
+      "grad_norm": 0.4302186667919159,
+      "learning_rate": 6.363608000305754e-06,
+      "loss": 0.3007,
+      "step": 12094
+    },
+    {
+      "epoch": 0.8894036326200456,
+      "grad_norm": 0.4494895040988922,
+      "learning_rate": 6.355250442761574e-06,
+      "loss": 0.3173,
+      "step": 12095
+    },
+    {
+      "epoch": 0.8894771674387822,
+      "grad_norm": 0.5614860653877258,
+      "learning_rate": 6.346898196846429e-06,
+      "loss": 0.3621,
+      "step": 12096
+    },
+    {
+      "epoch": 0.889550702257519,
+      "grad_norm": 0.5139672756195068,
+      "learning_rate": 6.338551263034043e-06,
+      "loss": 0.3141,
+      "step": 12097
+    },
+    {
+      "epoch": 0.8896242370762556,
+      "grad_norm": 0.36834076046943665,
+      "learning_rate": 6.330209641797868e-06,
+      "loss": 0.3287,
+      "step": 12098
+    },
+    {
+      "epoch": 0.8896977718949923,
+      "grad_norm": 0.4075983166694641,
+      "learning_rate": 6.3218733336110726e-06,
+      "loss": 0.334,
+      "step": 12099
+    },
+    {
+      "epoch": 0.889771306713729,
+      "grad_norm": 0.669446587562561,
+      "learning_rate": 6.313542338946488e-06,
+      "loss": 0.3269,
+      "step": 12100
+    },
+    {
+      "epoch": 0.8898448415324656,
+      "grad_norm": 0.3722193241119385,
+      "learning_rate": 6.305216658276636e-06,
+      "loss": 0.3356,
+      "step": 12101
+    },
+    {
+      "epoch": 0.8899183763512023,
+      "grad_norm": 0.39152291417121887,
+      "learning_rate": 6.296896292073773e-06,
+      "loss": 0.3312,
+      "step": 12102
+    },
+    {
+      "epoch": 0.8899919111699389,
+      "grad_norm": 0.3953525424003601,
+      "learning_rate": 6.288581240809854e-06,
+      "loss": 0.3372,
+      "step": 12103
+    },
+    {
+      "epoch": 0.8900654459886757,
+      "grad_norm": 0.39794230461120605,
+      "learning_rate": 6.280271504956503e-06,
+      "loss": 0.3138,
+      "step": 12104
+    },
+    {
+      "epoch": 0.8901389808074123,
+      "grad_norm": 0.43654003739356995,
+      "learning_rate": 6.2719670849850424e-06,
+      "loss": 0.328,
+      "step": 12105
+    },
+    {
+      "epoch": 0.890212515626149,
+      "grad_norm": 0.36434346437454224,
+      "learning_rate": 6.26366798136655e-06,
+      "loss": 0.3398,
+      "step": 12106
+    },
+    {
+      "epoch": 0.8902860504448856,
+      "grad_norm": 0.43549248576164246,
+      "learning_rate": 6.255374194571717e-06,
+      "loss": 0.3152,
+      "step": 12107
+    },
+    {
+      "epoch": 0.8903595852636224,
+      "grad_norm": 0.3923182189464569,
+      "learning_rate": 6.2470857250709895e-06,
+      "loss": 0.3452,
+      "step": 12108
+    },
+    {
+      "epoch": 0.890433120082359,
+      "grad_norm": 0.4755379557609558,
+      "learning_rate": 6.238802573334512e-06,
+      "loss": 0.3324,
+      "step": 12109
+    },
+    {
+      "epoch": 0.8905066549010957,
+      "grad_norm": 0.4153541028499603,
+      "learning_rate": 6.230524739832089e-06,
+      "loss": 0.3348,
+      "step": 12110
+    },
+    {
+      "epoch": 0.8905801897198323,
+      "grad_norm": 0.37494897842407227,
+      "learning_rate": 6.222252225033287e-06,
+      "loss": 0.3172,
+      "step": 12111
+    },
+    {
+      "epoch": 0.8906537245385691,
+      "grad_norm": 0.48641374707221985,
+      "learning_rate": 6.213985029407276e-06,
+      "loss": 0.3455,
+      "step": 12112
+    },
+    {
+      "epoch": 0.8907272593573057,
+      "grad_norm": 0.4049307703971863,
+      "learning_rate": 6.205723153423037e-06,
+      "loss": 0.3367,
+      "step": 12113
+    },
+    {
+      "epoch": 0.8908007941760424,
+      "grad_norm": 0.3546871244907379,
+      "learning_rate": 6.197466597549151e-06,
+      "loss": 0.3246,
+      "step": 12114
+    },
+    {
+      "epoch": 0.890874328994779,
+      "grad_norm": 0.4721483886241913,
+      "learning_rate": 6.189215362253942e-06,
+      "loss": 0.3424,
+      "step": 12115
+    },
+    {
+      "epoch": 0.8909478638135157,
+      "grad_norm": 0.4694400727748871,
+      "learning_rate": 6.180969448005469e-06,
+      "loss": 0.329,
+      "step": 12116
+    },
+    {
+      "epoch": 0.8910213986322524,
+      "grad_norm": 0.3798201382160187,
+      "learning_rate": 6.172728855271392e-06,
+      "loss": 0.3242,
+      "step": 12117
+    },
+    {
+      "epoch": 0.891094933450989,
+      "grad_norm": 0.36733582615852356,
+      "learning_rate": 6.16449358451916e-06,
+      "loss": 0.3224,
+      "step": 12118
+    },
+    {
+      "epoch": 0.8911684682697257,
+      "grad_norm": 0.43835675716400146,
+      "learning_rate": 6.156263636215897e-06,
+      "loss": 0.3359,
+      "step": 12119
+    },
+    {
+      "epoch": 0.8912420030884624,
+      "grad_norm": 0.5328495502471924,
+      "learning_rate": 6.148039010828388e-06,
+      "loss": 0.3397,
+      "step": 12120
+    },
+    {
+      "epoch": 0.8913155379071991,
+      "grad_norm": 0.4005294740200043,
+      "learning_rate": 6.139819708823147e-06,
+      "loss": 0.3257,
+      "step": 12121
+    },
+    {
+      "epoch": 0.8913890727259357,
+      "grad_norm": 0.3825775980949402,
+      "learning_rate": 6.131605730666401e-06,
+      "loss": 0.3216,
+      "step": 12122
+    },
+    {
+      "epoch": 0.8914626075446724,
+      "grad_norm": 0.4263826012611389,
+      "learning_rate": 6.123397076824033e-06,
+      "loss": 0.343,
+      "step": 12123
+    },
+    {
+      "epoch": 0.8915361423634091,
+      "grad_norm": 0.4350526034832001,
+      "learning_rate": 6.115193747761649e-06,
+      "loss": 0.3341,
+      "step": 12124
+    },
+    {
+      "epoch": 0.8916096771821458,
+      "grad_norm": 0.37302783131599426,
+      "learning_rate": 6.106995743944555e-06,
+      "loss": 0.3216,
+      "step": 12125
+    },
+    {
+      "epoch": 0.8916832120008824,
+      "grad_norm": 0.40580254793167114,
+      "learning_rate": 6.098803065837766e-06,
+      "loss": 0.3192,
+      "step": 12126
+    },
+    {
+      "epoch": 0.8917567468196191,
+      "grad_norm": 0.3971993923187256,
+      "learning_rate": 6.090615713905956e-06,
+      "loss": 0.3163,
+      "step": 12127
+    },
+    {
+      "epoch": 0.8918302816383558,
+      "grad_norm": 0.46286246180534363,
+      "learning_rate": 6.08243368861352e-06,
+      "loss": 0.3605,
+      "step": 12128
+    },
+    {
+      "epoch": 0.8919038164570924,
+      "grad_norm": 0.3754488229751587,
+      "learning_rate": 6.074256990424587e-06,
+      "loss": 0.3281,
+      "step": 12129
+    },
+    {
+      "epoch": 0.8919773512758291,
+      "grad_norm": 0.47284212708473206,
+      "learning_rate": 6.066085619802897e-06,
+      "loss": 0.3413,
+      "step": 12130
+    },
+    {
+      "epoch": 0.8920508860945657,
+      "grad_norm": 0.39337968826293945,
+      "learning_rate": 6.057919577211957e-06,
+      "loss": 0.3245,
+      "step": 12131
+    },
+    {
+      "epoch": 0.8921244209133025,
+      "grad_norm": 0.38866695761680603,
+      "learning_rate": 6.049758863114985e-06,
+      "loss": 0.3785,
+      "step": 12132
+    },
+    {
+      "epoch": 0.8921979557320391,
+      "grad_norm": 0.49345672130584717,
+      "learning_rate": 6.041603477974822e-06,
+      "loss": 0.3589,
+      "step": 12133
+    },
+    {
+      "epoch": 0.8922714905507758,
+      "grad_norm": 0.3981233537197113,
+      "learning_rate": 6.033453422254088e-06,
+      "loss": 0.3695,
+      "step": 12134
+    },
+    {
+      "epoch": 0.8923450253695124,
+      "grad_norm": 0.3776891231536865,
+      "learning_rate": 6.0253086964150215e-06,
+      "loss": 0.3094,
+      "step": 12135
+    },
+    {
+      "epoch": 0.8924185601882492,
+      "grad_norm": 0.4066665470600128,
+      "learning_rate": 6.017169300919634e-06,
+      "loss": 0.341,
+      "step": 12136
+    },
+    {
+      "epoch": 0.8924920950069858,
+      "grad_norm": 0.3675583302974701,
+      "learning_rate": 6.009035236229576e-06,
+      "loss": 0.347,
+      "step": 12137
+    },
+    {
+      "epoch": 0.8925656298257225,
+      "grad_norm": 0.4479418396949768,
+      "learning_rate": 6.000906502806236e-06,
+      "loss": 0.3255,
+      "step": 12138
+    },
+    {
+      "epoch": 0.8926391646444591,
+      "grad_norm": 0.41685107350349426,
+      "learning_rate": 5.992783101110689e-06,
+      "loss": 0.366,
+      "step": 12139
+    },
+    {
+      "epoch": 0.8927126994631959,
+      "grad_norm": 0.3777109682559967,
+      "learning_rate": 5.984665031603687e-06,
+      "loss": 0.3182,
+      "step": 12140
+    },
+    {
+      "epoch": 0.8927862342819325,
+      "grad_norm": 0.39727869629859924,
+      "learning_rate": 5.976552294745696e-06,
+      "loss": 0.3354,
+      "step": 12141
+    },
+    {
+      "epoch": 0.8928597691006692,
+      "grad_norm": 0.38765406608581543,
+      "learning_rate": 5.968444890996905e-06,
+      "loss": 0.3398,
+      "step": 12142
+    },
+    {
+      "epoch": 0.8929333039194058,
+      "grad_norm": 0.3978247046470642,
+      "learning_rate": 5.960342820817133e-06,
+      "loss": 0.3122,
+      "step": 12143
+    },
+    {
+      "epoch": 0.8930068387381425,
+      "grad_norm": 0.4194415211677551,
+      "learning_rate": 5.952246084665969e-06,
+      "loss": 0.3275,
+      "step": 12144
+    },
+    {
+      "epoch": 0.8930803735568792,
+      "grad_norm": 0.3700238764286041,
+      "learning_rate": 5.944154683002667e-06,
+      "loss": 0.3102,
+      "step": 12145
+    },
+    {
+      "epoch": 0.8931539083756158,
+      "grad_norm": 0.4378203749656677,
+      "learning_rate": 5.936068616286172e-06,
+      "loss": 0.3339,
+      "step": 12146
+    },
+    {
+      "epoch": 0.8932274431943525,
+      "grad_norm": 0.41214072704315186,
+      "learning_rate": 5.927987884975128e-06,
+      "loss": 0.3491,
+      "step": 12147
+    },
+    {
+      "epoch": 0.8933009780130892,
+      "grad_norm": 0.41789504885673523,
+      "learning_rate": 5.919912489527879e-06,
+      "loss": 0.3424,
+      "step": 12148
+    },
+    {
+      "epoch": 0.8933745128318259,
+      "grad_norm": 0.4017007648944855,
+      "learning_rate": 5.911842430402503e-06,
+      "loss": 0.3282,
+      "step": 12149
+    },
+    {
+      "epoch": 0.8934480476505625,
+      "grad_norm": 0.3500281572341919,
+      "learning_rate": 5.903777708056701e-06,
+      "loss": 0.3214,
+      "step": 12150
+    },
+    {
+      "epoch": 0.8935215824692992,
+      "grad_norm": 0.3951825201511383,
+      "learning_rate": 5.89571832294794e-06,
+      "loss": 0.3404,
+      "step": 12151
+    },
+    {
+      "epoch": 0.8935951172880359,
+      "grad_norm": 0.34924137592315674,
+      "learning_rate": 5.887664275533355e-06,
+      "loss": 0.3047,
+      "step": 12152
+    },
+    {
+      "epoch": 0.8936686521067726,
+      "grad_norm": 0.4211569130420685,
+      "learning_rate": 5.879615566269758e-06,
+      "loss": 0.3378,
+      "step": 12153
+    },
+    {
+      "epoch": 0.8937421869255092,
+      "grad_norm": 0.4558100998401642,
+      "learning_rate": 5.871572195613695e-06,
+      "loss": 0.3145,
+      "step": 12154
+    },
+    {
+      "epoch": 0.893815721744246,
+      "grad_norm": 0.41076478362083435,
+      "learning_rate": 5.863534164021412e-06,
+      "loss": 0.325,
+      "step": 12155
+    },
+    {
+      "epoch": 0.8938892565629826,
+      "grad_norm": 0.4022265672683716,
+      "learning_rate": 5.855501471948799e-06,
+      "loss": 0.3099,
+      "step": 12156
+    },
+    {
+      "epoch": 0.8939627913817192,
+      "grad_norm": 0.40882691740989685,
+      "learning_rate": 5.847474119851504e-06,
+      "loss": 0.3361,
+      "step": 12157
+    },
+    {
+      "epoch": 0.8940363262004559,
+      "grad_norm": 0.37608230113983154,
+      "learning_rate": 5.83945210818484e-06,
+      "loss": 0.3448,
+      "step": 12158
+    },
+    {
+      "epoch": 0.8941098610191925,
+      "grad_norm": 0.3837265372276306,
+      "learning_rate": 5.831435437403843e-06,
+      "loss": 0.3466,
+      "step": 12159
+    },
+    {
+      "epoch": 0.8941833958379293,
+      "grad_norm": 0.36954542994499207,
+      "learning_rate": 5.823424107963183e-06,
+      "loss": 0.3291,
+      "step": 12160
+    },
+    {
+      "epoch": 0.8942569306566659,
+      "grad_norm": 0.43227455019950867,
+      "learning_rate": 5.815418120317295e-06,
+      "loss": 0.3484,
+      "step": 12161
+    },
+    {
+      "epoch": 0.8943304654754026,
+      "grad_norm": 0.36964842677116394,
+      "learning_rate": 5.807417474920307e-06,
+      "loss": 0.3644,
+      "step": 12162
+    },
+    {
+      "epoch": 0.8944040002941392,
+      "grad_norm": 0.3844154179096222,
+      "learning_rate": 5.799422172225988e-06,
+      "loss": 0.3337,
+      "step": 12163
+    },
+    {
+      "epoch": 0.894477535112876,
+      "grad_norm": 0.4462898075580597,
+      "learning_rate": 5.791432212687853e-06,
+      "loss": 0.3588,
+      "step": 12164
+    },
+    {
+      "epoch": 0.8945510699316126,
+      "grad_norm": 0.38679277896881104,
+      "learning_rate": 5.7834475967591285e-06,
+      "loss": 0.3346,
+      "step": 12165
+    },
+    {
+      "epoch": 0.8946246047503493,
+      "grad_norm": 0.3705883324146271,
+      "learning_rate": 5.775468324892663e-06,
+      "loss": 0.3413,
+      "step": 12166
+    },
+    {
+      "epoch": 0.894698139569086,
+      "grad_norm": 0.39610105752944946,
+      "learning_rate": 5.767494397541085e-06,
+      "loss": 0.3243,
+      "step": 12167
+    },
+    {
+      "epoch": 0.8947716743878227,
+      "grad_norm": 0.3637906610965729,
+      "learning_rate": 5.759525815156686e-06,
+      "loss": 0.3407,
+      "step": 12168
+    },
+    {
+      "epoch": 0.8948452092065593,
+      "grad_norm": 0.3610195219516754,
+      "learning_rate": 5.7515625781914274e-06,
+      "loss": 0.3629,
+      "step": 12169
+    },
+    {
+      "epoch": 0.894918744025296,
+      "grad_norm": 0.4312238395214081,
+      "learning_rate": 5.743604687096993e-06,
+      "loss": 0.3246,
+      "step": 12170
+    },
+    {
+      "epoch": 0.8949922788440327,
+      "grad_norm": 0.41626954078674316,
+      "learning_rate": 5.735652142324776e-06,
+      "loss": 0.3355,
+      "step": 12171
+    },
+    {
+      "epoch": 0.8950658136627693,
+      "grad_norm": 0.44940194487571716,
+      "learning_rate": 5.7277049443258624e-06,
+      "loss": 0.3428,
+      "step": 12172
+    },
+    {
+      "epoch": 0.895139348481506,
+      "grad_norm": 0.46545594930648804,
+      "learning_rate": 5.7197630935510116e-06,
+      "loss": 0.3691,
+      "step": 12173
+    },
+    {
+      "epoch": 0.8952128833002426,
+      "grad_norm": 0.45996588468551636,
+      "learning_rate": 5.711826590450675e-06,
+      "loss": 0.337,
+      "step": 12174
+    },
+    {
+      "epoch": 0.8952864181189794,
+      "grad_norm": 0.4508563280105591,
+      "learning_rate": 5.7038954354750705e-06,
+      "loss": 0.3476,
+      "step": 12175
+    },
+    {
+      "epoch": 0.895359952937716,
+      "grad_norm": 0.4273480474948883,
+      "learning_rate": 5.695969629074016e-06,
+      "loss": 0.3669,
+      "step": 12176
+    },
+    {
+      "epoch": 0.8954334877564527,
+      "grad_norm": 0.4094293415546417,
+      "learning_rate": 5.6880491716970854e-06,
+      "loss": 0.3422,
+      "step": 12177
+    },
+    {
+      "epoch": 0.8955070225751893,
+      "grad_norm": 0.3761555254459381,
+      "learning_rate": 5.680134063793552e-06,
+      "loss": 0.3388,
+      "step": 12178
+    },
+    {
+      "epoch": 0.8955805573939261,
+      "grad_norm": 0.3804563283920288,
+      "learning_rate": 5.672224305812335e-06,
+      "loss": 0.3382,
+      "step": 12179
+    },
+    {
+      "epoch": 0.8956540922126627,
+      "grad_norm": 0.36700940132141113,
+      "learning_rate": 5.664319898202108e-06,
+      "loss": 0.3067,
+      "step": 12180
+    },
+    {
+      "epoch": 0.8957276270313994,
+      "grad_norm": 0.44328272342681885,
+      "learning_rate": 5.6564208414112255e-06,
+      "loss": 0.3307,
+      "step": 12181
+    },
+    {
+      "epoch": 0.895801161850136,
+      "grad_norm": 0.3801679015159607,
+      "learning_rate": 5.648527135887716e-06,
+      "loss": 0.3264,
+      "step": 12182
+    },
+    {
+      "epoch": 0.8958746966688728,
+      "grad_norm": 0.40783780813217163,
+      "learning_rate": 5.640638782079299e-06,
+      "loss": 0.3071,
+      "step": 12183
+    },
+    {
+      "epoch": 0.8959482314876094,
+      "grad_norm": 0.4315406382083893,
+      "learning_rate": 5.632755780433441e-06,
+      "loss": 0.3218,
+      "step": 12184
+    },
+    {
+      "epoch": 0.896021766306346,
+      "grad_norm": 0.34394437074661255,
+      "learning_rate": 5.624878131397282e-06,
+      "loss": 0.3098,
+      "step": 12185
+    },
+    {
+      "epoch": 0.8960953011250827,
+      "grad_norm": 0.40551242232322693,
+      "learning_rate": 5.617005835417621e-06,
+      "loss": 0.3373,
+      "step": 12186
+    },
+    {
+      "epoch": 0.8961688359438194,
+      "grad_norm": 0.52838534116745,
+      "learning_rate": 5.60913889294099e-06,
+      "loss": 0.339,
+      "step": 12187
+    },
+    {
+      "epoch": 0.8962423707625561,
+      "grad_norm": 0.48734983801841736,
+      "learning_rate": 5.601277304413621e-06,
+      "loss": 0.3392,
+      "step": 12188
+    },
+    {
+      "epoch": 0.8963159055812927,
+      "grad_norm": 0.3670576214790344,
+      "learning_rate": 5.593421070281423e-06,
+      "loss": 0.3331,
+      "step": 12189
+    },
+    {
+      "epoch": 0.8963894404000294,
+      "grad_norm": 0.4184560477733612,
+      "learning_rate": 5.585570190990008e-06,
+      "loss": 0.3427,
+      "step": 12190
+    },
+    {
+      "epoch": 0.8964629752187661,
+      "grad_norm": 0.45680177211761475,
+      "learning_rate": 5.5777246669847185e-06,
+      "loss": 0.3592,
+      "step": 12191
+    },
+    {
+      "epoch": 0.8965365100375028,
+      "grad_norm": 0.47725391387939453,
+      "learning_rate": 5.569884498710509e-06,
+      "loss": 0.2999,
+      "step": 12192
+    },
+    {
+      "epoch": 0.8966100448562394,
+      "grad_norm": 0.4896572232246399,
+      "learning_rate": 5.562049686612114e-06,
+      "loss": 0.3126,
+      "step": 12193
+    },
+    {
+      "epoch": 0.8966835796749761,
+      "grad_norm": 0.3601611852645874,
+      "learning_rate": 5.554220231133922e-06,
+      "loss": 0.3078,
+      "step": 12194
+    },
+    {
+      "epoch": 0.8967571144937128,
+      "grad_norm": 0.42465871572494507,
+      "learning_rate": 5.546396132720044e-06,
+      "loss": 0.332,
+      "step": 12195
+    },
+    {
+      "epoch": 0.8968306493124495,
+      "grad_norm": 0.37394270300865173,
+      "learning_rate": 5.538577391814237e-06,
+      "loss": 0.3129,
+      "step": 12196
+    },
+    {
+      "epoch": 0.8969041841311861,
+      "grad_norm": 0.4976320266723633,
+      "learning_rate": 5.530764008860012e-06,
+      "loss": 0.3152,
+      "step": 12197
+    },
+    {
+      "epoch": 0.8969777189499228,
+      "grad_norm": 0.40999382734298706,
+      "learning_rate": 5.52295598430057e-06,
+      "loss": 0.3281,
+      "step": 12198
+    },
+    {
+      "epoch": 0.8970512537686595,
+      "grad_norm": 0.43295958638191223,
+      "learning_rate": 5.5151533185787476e-06,
+      "loss": 0.3243,
+      "step": 12199
+    },
+    {
+      "epoch": 0.8971247885873961,
+      "grad_norm": 0.37947818636894226,
+      "learning_rate": 5.507356012137155e-06,
+      "loss": 0.3421,
+      "step": 12200
+    },
+    {
+      "epoch": 0.8971983234061328,
+      "grad_norm": 0.42937490344047546,
+      "learning_rate": 5.49956406541805e-06,
+      "loss": 0.3222,
+      "step": 12201
+    },
+    {
+      "epoch": 0.8972718582248694,
+      "grad_norm": 0.41951248049736023,
+      "learning_rate": 5.491777478863402e-06,
+      "loss": 0.311,
+      "step": 12202
+    },
+    {
+      "epoch": 0.8973453930436062,
+      "grad_norm": 0.46373212337493896,
+      "learning_rate": 5.483996252914858e-06,
+      "loss": 0.3529,
+      "step": 12203
+    },
+    {
+      "epoch": 0.8974189278623428,
+      "grad_norm": 0.4181596338748932,
+      "learning_rate": 5.476220388013819e-06,
+      "loss": 0.3155,
+      "step": 12204
+    },
+    {
+      "epoch": 0.8974924626810795,
+      "grad_norm": 0.35076573491096497,
+      "learning_rate": 5.468449884601312e-06,
+      "loss": 0.3488,
+      "step": 12205
+    },
+    {
+      "epoch": 0.8975659974998161,
+      "grad_norm": 0.5083370208740234,
+      "learning_rate": 5.460684743118083e-06,
+      "loss": 0.3432,
+      "step": 12206
+    },
+    {
+      "epoch": 0.8976395323185529,
+      "grad_norm": 0.45127707719802856,
+      "learning_rate": 5.45292496400458e-06,
+      "loss": 0.3011,
+      "step": 12207
+    },
+    {
+      "epoch": 0.8977130671372895,
+      "grad_norm": 0.41799670457839966,
+      "learning_rate": 5.445170547700962e-06,
+      "loss": 0.3287,
+      "step": 12208
+    },
+    {
+      "epoch": 0.8977866019560262,
+      "grad_norm": 0.34766992926597595,
+      "learning_rate": 5.4374214946470435e-06,
+      "loss": 0.3159,
+      "step": 12209
+    },
+    {
+      "epoch": 0.8978601367747628,
+      "grad_norm": 0.5719279646873474,
+      "learning_rate": 5.429677805282385e-06,
+      "loss": 0.3372,
+      "step": 12210
+    },
+    {
+      "epoch": 0.8979336715934996,
+      "grad_norm": 0.41343966126441956,
+      "learning_rate": 5.421939480046212e-06,
+      "loss": 0.3284,
+      "step": 12211
+    },
+    {
+      "epoch": 0.8980072064122362,
+      "grad_norm": 0.36415496468544006,
+      "learning_rate": 5.41420651937743e-06,
+      "loss": 0.3047,
+      "step": 12212
+    },
+    {
+      "epoch": 0.8980807412309728,
+      "grad_norm": 0.4092578589916229,
+      "learning_rate": 5.406478923714675e-06,
+      "loss": 0.355,
+      "step": 12213
+    },
+    {
+      "epoch": 0.8981542760497095,
+      "grad_norm": 0.44389787316322327,
+      "learning_rate": 5.398756693496276e-06,
+      "loss": 0.3063,
+      "step": 12214
+    },
+    {
+      "epoch": 0.8982278108684462,
+      "grad_norm": 0.42262783646583557,
+      "learning_rate": 5.391039829160227e-06,
+      "loss": 0.2892,
+      "step": 12215
+    },
+    {
+      "epoch": 0.8983013456871829,
+      "grad_norm": 0.4673064053058624,
+      "learning_rate": 5.383328331144255e-06,
+      "loss": 0.326,
+      "step": 12216
+    },
+    {
+      "epoch": 0.8983748805059195,
+      "grad_norm": 0.5213967561721802,
+      "learning_rate": 5.375622199885733e-06,
+      "loss": 0.3319,
+      "step": 12217
+    },
+    {
+      "epoch": 0.8984484153246562,
+      "grad_norm": 0.6143325567245483,
+      "learning_rate": 5.367921435821799e-06,
+      "loss": 0.3157,
+      "step": 12218
+    },
+    {
+      "epoch": 0.8985219501433929,
+      "grad_norm": 0.5441703796386719,
+      "learning_rate": 5.360226039389216e-06,
+      "loss": 0.3227,
+      "step": 12219
+    },
+    {
+      "epoch": 0.8985954849621296,
+      "grad_norm": 0.3748890459537506,
+      "learning_rate": 5.352536011024501e-06,
+      "loss": 0.2929,
+      "step": 12220
+    },
+    {
+      "epoch": 0.8986690197808662,
+      "grad_norm": 0.43756935000419617,
+      "learning_rate": 5.344851351163827e-06,
+      "loss": 0.32,
+      "step": 12221
+    },
+    {
+      "epoch": 0.898742554599603,
+      "grad_norm": 0.6093867421150208,
+      "learning_rate": 5.337172060243067e-06,
+      "loss": 0.3271,
+      "step": 12222
+    },
+    {
+      "epoch": 0.8988160894183396,
+      "grad_norm": 0.5129573345184326,
+      "learning_rate": 5.329498138697808e-06,
+      "loss": 0.3295,
+      "step": 12223
+    },
+    {
+      "epoch": 0.8988896242370763,
+      "grad_norm": 0.4128526747226715,
+      "learning_rate": 5.321829586963345e-06,
+      "loss": 0.305,
+      "step": 12224
+    },
+    {
+      "epoch": 0.8989631590558129,
+      "grad_norm": 0.4200683832168579,
+      "learning_rate": 5.314166405474608e-06,
+      "loss": 0.3397,
+      "step": 12225
+    },
+    {
+      "epoch": 0.8990366938745497,
+      "grad_norm": 0.40440833568573,
+      "learning_rate": 5.306508594666282e-06,
+      "loss": 0.3474,
+      "step": 12226
+    },
+    {
+      "epoch": 0.8991102286932863,
+      "grad_norm": 0.4204501807689667,
+      "learning_rate": 5.298856154972731e-06,
+      "loss": 0.3097,
+      "step": 12227
+    },
+    {
+      "epoch": 0.8991837635120229,
+      "grad_norm": 0.4138278067111969,
+      "learning_rate": 5.291209086828008e-06,
+      "loss": 0.3486,
+      "step": 12228
+    },
+    {
+      "epoch": 0.8992572983307596,
+      "grad_norm": 0.3967941701412201,
+      "learning_rate": 5.283567390665833e-06,
+      "loss": 0.327,
+      "step": 12229
+    },
+    {
+      "epoch": 0.8993308331494962,
+      "grad_norm": 0.5004366040229797,
+      "learning_rate": 5.275931066919682e-06,
+      "loss": 0.3353,
+      "step": 12230
+    },
+    {
+      "epoch": 0.899404367968233,
+      "grad_norm": 0.45890718698501587,
+      "learning_rate": 5.2683001160226954e-06,
+      "loss": 0.2965,
+      "step": 12231
+    },
+    {
+      "epoch": 0.8994779027869696,
+      "grad_norm": 0.40877851843833923,
+      "learning_rate": 5.260674538407684e-06,
+      "loss": 0.322,
+      "step": 12232
+    },
+    {
+      "epoch": 0.8995514376057063,
+      "grad_norm": 0.43145790696144104,
+      "learning_rate": 5.253054334507201e-06,
+      "loss": 0.3207,
+      "step": 12233
+    },
+    {
+      "epoch": 0.899624972424443,
+      "grad_norm": 0.43816477060317993,
+      "learning_rate": 5.24543950475348e-06,
+      "loss": 0.3466,
+      "step": 12234
+    },
+    {
+      "epoch": 0.8996985072431797,
+      "grad_norm": 0.3547910153865814,
+      "learning_rate": 5.2378300495784076e-06,
+      "loss": 0.3301,
+      "step": 12235
+    },
+    {
+      "epoch": 0.8997720420619163,
+      "grad_norm": 0.36371687054634094,
+      "learning_rate": 5.2302259694136265e-06,
+      "loss": 0.3219,
+      "step": 12236
+    },
+    {
+      "epoch": 0.899845576880653,
+      "grad_norm": 0.42376983165740967,
+      "learning_rate": 5.222627264690461e-06,
+      "loss": 0.3227,
+      "step": 12237
+    },
+    {
+      "epoch": 0.8999191116993897,
+      "grad_norm": 0.4092542827129364,
+      "learning_rate": 5.215033935839886e-06,
+      "loss": 0.3248,
+      "step": 12238
+    },
+    {
+      "epoch": 0.8999926465181264,
+      "grad_norm": 0.3986511528491974,
+      "learning_rate": 5.207445983292624e-06,
+      "loss": 0.3431,
+      "step": 12239
+    },
+    {
+      "epoch": 0.900066181336863,
+      "grad_norm": 0.4341167211532593,
+      "learning_rate": 5.1998634074790555e-06,
+      "loss": 0.3569,
+      "step": 12240
+    },
+    {
+      "epoch": 0.9001397161555996,
+      "grad_norm": 0.41129961609840393,
+      "learning_rate": 5.192286208829289e-06,
+      "loss": 0.3409,
+      "step": 12241
+    },
+    {
+      "epoch": 0.9002132509743364,
+      "grad_norm": 0.4803019165992737,
+      "learning_rate": 5.184714387773093e-06,
+      "loss": 0.3606,
+      "step": 12242
+    },
+    {
+      "epoch": 0.900286785793073,
+      "grad_norm": 0.380974143743515,
+      "learning_rate": 5.177147944739968e-06,
+      "loss": 0.2962,
+      "step": 12243
+    },
+    {
+      "epoch": 0.9003603206118097,
+      "grad_norm": 0.3446759879589081,
+      "learning_rate": 5.169586880159083e-06,
+      "loss": 0.3145,
+      "step": 12244
+    },
+    {
+      "epoch": 0.9004338554305463,
+      "grad_norm": 0.4313039481639862,
+      "learning_rate": 5.162031194459305e-06,
+      "loss": 0.2959,
+      "step": 12245
+    },
+    {
+      "epoch": 0.9005073902492831,
+      "grad_norm": 0.42963510751724243,
+      "learning_rate": 5.154480888069191e-06,
+      "loss": 0.3488,
+      "step": 12246
+    },
+    {
+      "epoch": 0.9005809250680197,
+      "grad_norm": 0.40936169028282166,
+      "learning_rate": 5.1469359614170436e-06,
+      "loss": 0.3099,
+      "step": 12247
+    },
+    {
+      "epoch": 0.9006544598867564,
+      "grad_norm": 0.3820749819278717,
+      "learning_rate": 5.139396414930764e-06,
+      "loss": 0.345,
+      "step": 12248
+    },
+    {
+      "epoch": 0.900727994705493,
+      "grad_norm": 0.4815734028816223,
+      "learning_rate": 5.131862249038044e-06,
+      "loss": 0.3648,
+      "step": 12249
+    },
+    {
+      "epoch": 0.9008015295242298,
+      "grad_norm": 0.42807886004447937,
+      "learning_rate": 5.12433346416622e-06,
+      "loss": 0.3233,
+      "step": 12250
+    },
+    {
+      "epoch": 0.9008750643429664,
+      "grad_norm": 0.38714998960494995,
+      "learning_rate": 5.116810060742327e-06,
+      "loss": 0.3063,
+      "step": 12251
+    },
+    {
+      "epoch": 0.9009485991617031,
+      "grad_norm": 0.3813808560371399,
+      "learning_rate": 5.109292039193092e-06,
+      "loss": 0.3432,
+      "step": 12252
+    },
+    {
+      "epoch": 0.9010221339804397,
+      "grad_norm": 0.37811750173568726,
+      "learning_rate": 5.10177939994495e-06,
+      "loss": 0.3508,
+      "step": 12253
+    },
+    {
+      "epoch": 0.9010956687991765,
+      "grad_norm": 0.40216320753097534,
+      "learning_rate": 5.09427214342405e-06,
+      "loss": 0.3136,
+      "step": 12254
+    },
+    {
+      "epoch": 0.9011692036179131,
+      "grad_norm": 0.4344820976257324,
+      "learning_rate": 5.086770270056174e-06,
+      "loss": 0.3212,
+      "step": 12255
+    },
+    {
+      "epoch": 0.9012427384366497,
+      "grad_norm": 0.37653812766075134,
+      "learning_rate": 5.079273780266858e-06,
+      "loss": 0.3356,
+      "step": 12256
+    },
+    {
+      "epoch": 0.9013162732553864,
+      "grad_norm": 0.36919304728507996,
+      "learning_rate": 5.071782674481318e-06,
+      "loss": 0.3437,
+      "step": 12257
+    },
+    {
+      "epoch": 0.9013898080741231,
+      "grad_norm": 0.388128399848938,
+      "learning_rate": 5.064296953124437e-06,
+      "loss": 0.3218,
+      "step": 12258
+    },
+    {
+      "epoch": 0.9014633428928598,
+      "grad_norm": 0.38958117365837097,
+      "learning_rate": 5.056816616620819e-06,
+      "loss": 0.3331,
+      "step": 12259
+    },
+    {
+      "epoch": 0.9015368777115964,
+      "grad_norm": 0.37597349286079407,
+      "learning_rate": 5.04934166539478e-06,
+      "loss": 0.306,
+      "step": 12260
+    },
+    {
+      "epoch": 0.9016104125303331,
+      "grad_norm": 0.3909282386302948,
+      "learning_rate": 5.04187209987026e-06,
+      "loss": 0.3166,
+      "step": 12261
+    },
+    {
+      "epoch": 0.9016839473490698,
+      "grad_norm": 0.42908570170402527,
+      "learning_rate": 5.0344079204709736e-06,
+      "loss": 0.322,
+      "step": 12262
+    },
+    {
+      "epoch": 0.9017574821678065,
+      "grad_norm": 0.3963474631309509,
+      "learning_rate": 5.026949127620306e-06,
+      "loss": 0.3322,
+      "step": 12263
+    },
+    {
+      "epoch": 0.9018310169865431,
+      "grad_norm": 0.40451672673225403,
+      "learning_rate": 5.019495721741318e-06,
+      "loss": 0.3517,
+      "step": 12264
+    },
+    {
+      "epoch": 0.9019045518052798,
+      "grad_norm": 0.4690014719963074,
+      "learning_rate": 5.0120477032567505e-06,
+      "loss": 0.3435,
+      "step": 12265
+    },
+    {
+      "epoch": 0.9019780866240165,
+      "grad_norm": 0.47354766726493835,
+      "learning_rate": 5.004605072589086e-06,
+      "loss": 0.3308,
+      "step": 12266
+    },
+    {
+      "epoch": 0.9020516214427532,
+      "grad_norm": 0.33382993936538696,
+      "learning_rate": 4.997167830160487e-06,
+      "loss": 0.3372,
+      "step": 12267
+    },
+    {
+      "epoch": 0.9021251562614898,
+      "grad_norm": 0.37928077578544617,
+      "learning_rate": 4.989735976392773e-06,
+      "loss": 0.3424,
+      "step": 12268
+    },
+    {
+      "epoch": 0.9021986910802264,
+      "grad_norm": 0.38096994161605835,
+      "learning_rate": 4.9823095117074945e-06,
+      "loss": 0.3457,
+      "step": 12269
+    },
+    {
+      "epoch": 0.9022722258989632,
+      "grad_norm": 0.3688880503177643,
+      "learning_rate": 4.9748884365259266e-06,
+      "loss": 0.3174,
+      "step": 12270
+    },
+    {
+      "epoch": 0.9023457607176998,
+      "grad_norm": 0.34803563356399536,
+      "learning_rate": 4.9674727512689425e-06,
+      "loss": 0.3236,
+      "step": 12271
+    },
+    {
+      "epoch": 0.9024192955364365,
+      "grad_norm": 0.38490965962409973,
+      "learning_rate": 4.960062456357206e-06,
+      "loss": 0.3388,
+      "step": 12272
+    },
+    {
+      "epoch": 0.9024928303551731,
+      "grad_norm": 0.4144085645675659,
+      "learning_rate": 4.952657552211038e-06,
+      "loss": 0.343,
+      "step": 12273
+    },
+    {
+      "epoch": 0.9025663651739099,
+      "grad_norm": 0.3752376139163971,
+      "learning_rate": 4.945258039250445e-06,
+      "loss": 0.334,
+      "step": 12274
+    },
+    {
+      "epoch": 0.9026398999926465,
+      "grad_norm": 0.38848042488098145,
+      "learning_rate": 4.937863917895102e-06,
+      "loss": 0.3167,
+      "step": 12275
+    },
+    {
+      "epoch": 0.9027134348113832,
+      "grad_norm": 0.3678349554538727,
+      "learning_rate": 4.930475188564454e-06,
+      "loss": 0.308,
+      "step": 12276
+    },
+    {
+      "epoch": 0.9027869696301198,
+      "grad_norm": 0.4714556634426117,
+      "learning_rate": 4.923091851677597e-06,
+      "loss": 0.3089,
+      "step": 12277
+    },
+    {
+      "epoch": 0.9028605044488566,
+      "grad_norm": 0.38930946588516235,
+      "learning_rate": 4.915713907653296e-06,
+      "loss": 0.3341,
+      "step": 12278
+    },
+    {
+      "epoch": 0.9029340392675932,
+      "grad_norm": 0.38419944047927856,
+      "learning_rate": 4.90834135691004e-06,
+      "loss": 0.3594,
+      "step": 12279
+    },
+    {
+      "epoch": 0.9030075740863299,
+      "grad_norm": 0.39680224657058716,
+      "learning_rate": 4.900974199866037e-06,
+      "loss": 0.3278,
+      "step": 12280
+    },
+    {
+      "epoch": 0.9030811089050665,
+      "grad_norm": 0.38519757986068726,
+      "learning_rate": 4.893612436939121e-06,
+      "loss": 0.3483,
+      "step": 12281
+    },
+    {
+      "epoch": 0.9031546437238033,
+      "grad_norm": 0.48108789324760437,
+      "learning_rate": 4.886256068546891e-06,
+      "loss": 0.3753,
+      "step": 12282
+    },
+    {
+      "epoch": 0.9032281785425399,
+      "grad_norm": 0.35637858510017395,
+      "learning_rate": 4.878905095106589e-06,
+      "loss": 0.3318,
+      "step": 12283
+    },
+    {
+      "epoch": 0.9033017133612765,
+      "grad_norm": 0.4122523367404938,
+      "learning_rate": 4.8715595170351735e-06,
+      "loss": 0.3328,
+      "step": 12284
+    },
+    {
+      "epoch": 0.9033752481800132,
+      "grad_norm": 0.4944498538970947,
+      "learning_rate": 4.864219334749298e-06,
+      "loss": 0.3268,
+      "step": 12285
+    },
+    {
+      "epoch": 0.9034487829987499,
+      "grad_norm": 0.3390939235687256,
+      "learning_rate": 4.856884548665308e-06,
+      "loss": 0.3115,
+      "step": 12286
+    },
+    {
+      "epoch": 0.9035223178174866,
+      "grad_norm": 0.45525938272476196,
+      "learning_rate": 4.849555159199237e-06,
+      "loss": 0.3502,
+      "step": 12287
+    },
+    {
+      "epoch": 0.9035958526362232,
+      "grad_norm": 0.4112241566181183,
+      "learning_rate": 4.8422311667667975e-06,
+      "loss": 0.3432,
+      "step": 12288
+    },
+    {
+      "epoch": 0.90366938745496,
+      "grad_norm": 0.4072153866291046,
+      "learning_rate": 4.834912571783434e-06,
+      "loss": 0.3104,
+      "step": 12289
+    },
+    {
+      "epoch": 0.9037429222736966,
+      "grad_norm": 0.4448481500148773,
+      "learning_rate": 4.827599374664271e-06,
+      "loss": 0.3306,
+      "step": 12290
+    },
+    {
+      "epoch": 0.9038164570924333,
+      "grad_norm": 0.5866658687591553,
+      "learning_rate": 4.8202915758240985e-06,
+      "loss": 0.3703,
+      "step": 12291
+    },
+    {
+      "epoch": 0.9038899919111699,
+      "grad_norm": 0.3893630802631378,
+      "learning_rate": 4.8129891756774406e-06,
+      "loss": 0.3158,
+      "step": 12292
+    },
+    {
+      "epoch": 0.9039635267299067,
+      "grad_norm": 0.3538910746574402,
+      "learning_rate": 4.805692174638499e-06,
+      "loss": 0.3204,
+      "step": 12293
+    },
+    {
+      "epoch": 0.9040370615486433,
+      "grad_norm": 0.41515931487083435,
+      "learning_rate": 4.798400573121154e-06,
+      "loss": 0.3274,
+      "step": 12294
+    },
+    {
+      "epoch": 0.90411059636738,
+      "grad_norm": 0.4705389738082886,
+      "learning_rate": 4.791114371538996e-06,
+      "loss": 0.3659,
+      "step": 12295
+    },
+    {
+      "epoch": 0.9041841311861166,
+      "grad_norm": 0.43706148862838745,
+      "learning_rate": 4.783833570305318e-06,
+      "loss": 0.3286,
+      "step": 12296
+    },
+    {
+      "epoch": 0.9042576660048532,
+      "grad_norm": 0.4621943533420563,
+      "learning_rate": 4.776558169833067e-06,
+      "loss": 0.3641,
+      "step": 12297
+    },
+    {
+      "epoch": 0.90433120082359,
+      "grad_norm": 0.4062458872795105,
+      "learning_rate": 4.7692881705349575e-06,
+      "loss": 0.3619,
+      "step": 12298
+    },
+    {
+      "epoch": 0.9044047356423266,
+      "grad_norm": 0.47583961486816406,
+      "learning_rate": 4.762023572823293e-06,
+      "loss": 0.3172,
+      "step": 12299
+    },
+    {
+      "epoch": 0.9044782704610633,
+      "grad_norm": 0.39618170261383057,
+      "learning_rate": 4.754764377110188e-06,
+      "loss": 0.3384,
+      "step": 12300
+    },
+    {
+      "epoch": 0.9045518052798,
+      "grad_norm": 0.3695598542690277,
+      "learning_rate": 4.747510583807347e-06,
+      "loss": 0.3347,
+      "step": 12301
+    },
+    {
+      "epoch": 0.9046253400985367,
+      "grad_norm": 0.41830700635910034,
+      "learning_rate": 4.7402621933262285e-06,
+      "loss": 0.3221,
+      "step": 12302
+    },
+    {
+      "epoch": 0.9046988749172733,
+      "grad_norm": 0.41394171118736267,
+      "learning_rate": 4.733019206077982e-06,
+      "loss": 0.3147,
+      "step": 12303
+    },
+    {
+      "epoch": 0.90477240973601,
+      "grad_norm": 0.43831366300582886,
+      "learning_rate": 4.725781622473413e-06,
+      "loss": 0.3431,
+      "step": 12304
+    },
+    {
+      "epoch": 0.9048459445547467,
+      "grad_norm": 0.4557306468486786,
+      "learning_rate": 4.7185494429230595e-06,
+      "loss": 0.3366,
+      "step": 12305
+    },
+    {
+      "epoch": 0.9049194793734834,
+      "grad_norm": 0.38420382142066956,
+      "learning_rate": 4.711322667837148e-06,
+      "loss": 0.3272,
+      "step": 12306
+    },
+    {
+      "epoch": 0.90499301419222,
+      "grad_norm": 0.3779398500919342,
+      "learning_rate": 4.704101297625563e-06,
+      "loss": 0.3486,
+      "step": 12307
+    },
+    {
+      "epoch": 0.9050665490109567,
+      "grad_norm": 0.47544732689857483,
+      "learning_rate": 4.69688533269792e-06,
+      "loss": 0.3351,
+      "step": 12308
+    },
+    {
+      "epoch": 0.9051400838296934,
+      "grad_norm": 0.40497535467147827,
+      "learning_rate": 4.689674773463537e-06,
+      "loss": 0.338,
+      "step": 12309
+    },
+    {
+      "epoch": 0.9052136186484301,
+      "grad_norm": 0.5008670687675476,
+      "learning_rate": 4.682469620331375e-06,
+      "loss": 0.3655,
+      "step": 12310
+    },
+    {
+      "epoch": 0.9052871534671667,
+      "grad_norm": 0.4107423424720764,
+      "learning_rate": 4.675269873710131e-06,
+      "loss": 0.3298,
+      "step": 12311
+    },
+    {
+      "epoch": 0.9053606882859033,
+      "grad_norm": 0.4208897054195404,
+      "learning_rate": 4.668075534008165e-06,
+      "loss": 0.3025,
+      "step": 12312
+    },
+    {
+      "epoch": 0.9054342231046401,
+      "grad_norm": 0.42510664463043213,
+      "learning_rate": 4.6608866016335854e-06,
+      "loss": 0.34,
+      "step": 12313
+    },
+    {
+      "epoch": 0.9055077579233767,
+      "grad_norm": 0.37682756781578064,
+      "learning_rate": 4.65370307699412e-06,
+      "loss": 0.3298,
+      "step": 12314
+    },
+    {
+      "epoch": 0.9055812927421134,
+      "grad_norm": 0.4242480397224426,
+      "learning_rate": 4.646524960497234e-06,
+      "loss": 0.3214,
+      "step": 12315
+    },
+    {
+      "epoch": 0.90565482756085,
+      "grad_norm": 0.48263803124427795,
+      "learning_rate": 4.6393522525501e-06,
+      "loss": 0.3386,
+      "step": 12316
+    },
+    {
+      "epoch": 0.9057283623795868,
+      "grad_norm": 0.463515043258667,
+      "learning_rate": 4.6321849535595265e-06,
+      "loss": 0.3155,
+      "step": 12317
+    },
+    {
+      "epoch": 0.9058018971983234,
+      "grad_norm": 0.41226983070373535,
+      "learning_rate": 4.625023063932066e-06,
+      "loss": 0.3322,
+      "step": 12318
+    },
+    {
+      "epoch": 0.9058754320170601,
+      "grad_norm": 0.3975716829299927,
+      "learning_rate": 4.617866584073971e-06,
+      "loss": 0.3346,
+      "step": 12319
+    },
+    {
+      "epoch": 0.9059489668357967,
+      "grad_norm": 0.43680816888809204,
+      "learning_rate": 4.610715514391128e-06,
+      "loss": 0.3703,
+      "step": 12320
+    },
+    {
+      "epoch": 0.9060225016545335,
+      "grad_norm": 0.49334561824798584,
+      "learning_rate": 4.603569855289191e-06,
+      "loss": 0.3416,
+      "step": 12321
+    },
+    {
+      "epoch": 0.9060960364732701,
+      "grad_norm": 0.42487600445747375,
+      "learning_rate": 4.596429607173425e-06,
+      "loss": 0.3177,
+      "step": 12322
+    },
+    {
+      "epoch": 0.9061695712920068,
+      "grad_norm": 0.3886333107948303,
+      "learning_rate": 4.589294770448871e-06,
+      "loss": 0.3229,
+      "step": 12323
+    },
+    {
+      "epoch": 0.9062431061107434,
+      "grad_norm": 0.4167497158050537,
+      "learning_rate": 4.582165345520195e-06,
+      "loss": 0.3429,
+      "step": 12324
+    },
+    {
+      "epoch": 0.90631664092948,
+      "grad_norm": 0.44024458527565,
+      "learning_rate": 4.575041332791796e-06,
+      "loss": 0.3329,
+      "step": 12325
+    },
+    {
+      "epoch": 0.9063901757482168,
+      "grad_norm": 0.41468292474746704,
+      "learning_rate": 4.5679227326677735e-06,
+      "loss": 0.3306,
+      "step": 12326
+    },
+    {
+      "epoch": 0.9064637105669534,
+      "grad_norm": 0.5193233489990234,
+      "learning_rate": 4.56080954555187e-06,
+      "loss": 0.3411,
+      "step": 12327
+    },
+    {
+      "epoch": 0.9065372453856901,
+      "grad_norm": 0.3500364422798157,
+      "learning_rate": 4.553701771847574e-06,
+      "loss": 0.3451,
+      "step": 12328
+    },
+    {
+      "epoch": 0.9066107802044268,
+      "grad_norm": 0.39741620421409607,
+      "learning_rate": 4.546599411958063e-06,
+      "loss": 0.3173,
+      "step": 12329
+    },
+    {
+      "epoch": 0.9066843150231635,
+      "grad_norm": 0.4095485806465149,
+      "learning_rate": 4.5395024662861385e-06,
+      "loss": 0.3372,
+      "step": 12330
+    },
+    {
+      "epoch": 0.9067578498419001,
+      "grad_norm": 0.5281849503517151,
+      "learning_rate": 4.532410935234388e-06,
+      "loss": 0.3247,
+      "step": 12331
+    },
+    {
+      "epoch": 0.9068313846606368,
+      "grad_norm": 0.45893728733062744,
+      "learning_rate": 4.5253248192050565e-06,
+      "loss": 0.3245,
+      "step": 12332
+    },
+    {
+      "epoch": 0.9069049194793735,
+      "grad_norm": 0.4347682297229767,
+      "learning_rate": 4.518244118600057e-06,
+      "loss": 0.3211,
+      "step": 12333
+    },
+    {
+      "epoch": 0.9069784542981102,
+      "grad_norm": 0.5005847811698914,
+      "learning_rate": 4.511168833821e-06,
+      "loss": 0.3641,
+      "step": 12334
+    },
+    {
+      "epoch": 0.9070519891168468,
+      "grad_norm": 0.42408087849617004,
+      "learning_rate": 4.504098965269232e-06,
+      "loss": 0.3315,
+      "step": 12335
+    },
+    {
+      "epoch": 0.9071255239355835,
+      "grad_norm": 0.5094184875488281,
+      "learning_rate": 4.497034513345755e-06,
+      "loss": 0.315,
+      "step": 12336
+    },
+    {
+      "epoch": 0.9071990587543202,
+      "grad_norm": 0.44381213188171387,
+      "learning_rate": 4.489975478451258e-06,
+      "loss": 0.2948,
+      "step": 12337
+    },
+    {
+      "epoch": 0.9072725935730569,
+      "grad_norm": 0.40118831396102905,
+      "learning_rate": 4.4829218609861445e-06,
+      "loss": 0.2822,
+      "step": 12338
+    },
+    {
+      "epoch": 0.9073461283917935,
+      "grad_norm": 0.40534865856170654,
+      "learning_rate": 4.475873661350516e-06,
+      "loss": 0.3453,
+      "step": 12339
+    },
+    {
+      "epoch": 0.9074196632105301,
+      "grad_norm": 0.37381526827812195,
+      "learning_rate": 4.468830879944141e-06,
+      "loss": 0.2885,
+      "step": 12340
+    },
+    {
+      "epoch": 0.9074931980292669,
+      "grad_norm": 0.5412638187408447,
+      "learning_rate": 4.4617935171664906e-06,
+      "loss": 0.3318,
+      "step": 12341
+    },
+    {
+      "epoch": 0.9075667328480035,
+      "grad_norm": 0.44171276688575745,
+      "learning_rate": 4.454761573416755e-06,
+      "loss": 0.3169,
+      "step": 12342
+    },
+    {
+      "epoch": 0.9076402676667402,
+      "grad_norm": 0.5502289533615112,
+      "learning_rate": 4.447735049093749e-06,
+      "loss": 0.3296,
+      "step": 12343
+    },
+    {
+      "epoch": 0.9077138024854768,
+      "grad_norm": 0.3543931245803833,
+      "learning_rate": 4.440713944596075e-06,
+      "loss": 0.3008,
+      "step": 12344
+    },
+    {
+      "epoch": 0.9077873373042136,
+      "grad_norm": 0.41880667209625244,
+      "learning_rate": 4.433698260321939e-06,
+      "loss": 0.3068,
+      "step": 12345
+    },
+    {
+      "epoch": 0.9078608721229502,
+      "grad_norm": 0.5409342646598816,
+      "learning_rate": 4.42668799666931e-06,
+      "loss": 0.31,
+      "step": 12346
+    },
+    {
+      "epoch": 0.9079344069416869,
+      "grad_norm": 0.38681283593177795,
+      "learning_rate": 4.419683154035792e-06,
+      "loss": 0.3151,
+      "step": 12347
+    },
+    {
+      "epoch": 0.9080079417604235,
+      "grad_norm": 0.3947303891181946,
+      "learning_rate": 4.412683732818701e-06,
+      "loss": 0.3054,
+      "step": 12348
+    },
+    {
+      "epoch": 0.9080814765791603,
+      "grad_norm": 0.3997407555580139,
+      "learning_rate": 4.405689733415086e-06,
+      "loss": 0.3122,
+      "step": 12349
+    },
+    {
+      "epoch": 0.9081550113978969,
+      "grad_norm": 0.49641963839530945,
+      "learning_rate": 4.398701156221619e-06,
+      "loss": 0.3364,
+      "step": 12350
+    },
+    {
+      "epoch": 0.9082285462166336,
+      "grad_norm": 0.3735009431838989,
+      "learning_rate": 4.391718001634715e-06,
+      "loss": 0.3369,
+      "step": 12351
+    },
+    {
+      "epoch": 0.9083020810353702,
+      "grad_norm": 0.546679675579071,
+      "learning_rate": 4.38474027005048e-06,
+      "loss": 0.3389,
+      "step": 12352
+    },
+    {
+      "epoch": 0.9083756158541069,
+      "grad_norm": 0.41463690996170044,
+      "learning_rate": 4.377767961864676e-06,
+      "loss": 0.3176,
+      "step": 12353
+    },
+    {
+      "epoch": 0.9084491506728436,
+      "grad_norm": 0.38297685980796814,
+      "learning_rate": 4.370801077472775e-06,
+      "loss": 0.3018,
+      "step": 12354
+    },
+    {
+      "epoch": 0.9085226854915802,
+      "grad_norm": 0.40302371978759766,
+      "learning_rate": 4.363839617269982e-06,
+      "loss": 0.3342,
+      "step": 12355
+    },
+    {
+      "epoch": 0.908596220310317,
+      "grad_norm": 0.4044061303138733,
+      "learning_rate": 4.356883581651128e-06,
+      "loss": 0.3367,
+      "step": 12356
+    },
+    {
+      "epoch": 0.9086697551290536,
+      "grad_norm": 0.43211984634399414,
+      "learning_rate": 4.3499329710107725e-06,
+      "loss": 0.3126,
+      "step": 12357
+    },
+    {
+      "epoch": 0.9087432899477903,
+      "grad_norm": 0.41504794359207153,
+      "learning_rate": 4.342987785743147e-06,
+      "loss": 0.3202,
+      "step": 12358
+    },
+    {
+      "epoch": 0.9088168247665269,
+      "grad_norm": 0.4771023690700531,
+      "learning_rate": 4.3360480262422235e-06,
+      "loss": 0.3432,
+      "step": 12359
+    },
+    {
+      "epoch": 0.9088903595852637,
+      "grad_norm": 0.45240676403045654,
+      "learning_rate": 4.32911369290161e-06,
+      "loss": 0.3413,
+      "step": 12360
+    },
+    {
+      "epoch": 0.9089638944040003,
+      "grad_norm": 0.37014785408973694,
+      "learning_rate": 4.322184786114625e-06,
+      "loss": 0.3417,
+      "step": 12361
+    },
+    {
+      "epoch": 0.909037429222737,
+      "grad_norm": 0.3973746597766876,
+      "learning_rate": 4.315261306274299e-06,
+      "loss": 0.3627,
+      "step": 12362
+    },
+    {
+      "epoch": 0.9091109640414736,
+      "grad_norm": 0.44140180945396423,
+      "learning_rate": 4.308343253773328e-06,
+      "loss": 0.337,
+      "step": 12363
+    },
+    {
+      "epoch": 0.9091844988602104,
+      "grad_norm": 0.37748292088508606,
+      "learning_rate": 4.301430629004111e-06,
+      "loss": 0.33,
+      "step": 12364
+    },
+    {
+      "epoch": 0.909258033678947,
+      "grad_norm": 0.4581148624420166,
+      "learning_rate": 4.2945234323587545e-06,
+      "loss": 0.3161,
+      "step": 12365
+    },
+    {
+      "epoch": 0.9093315684976837,
+      "grad_norm": 0.406579852104187,
+      "learning_rate": 4.287621664229024e-06,
+      "loss": 0.315,
+      "step": 12366
+    },
+    {
+      "epoch": 0.9094051033164203,
+      "grad_norm": 0.3354578912258148,
+      "learning_rate": 4.280725325006396e-06,
+      "loss": 0.3061,
+      "step": 12367
+    },
+    {
+      "epoch": 0.909478638135157,
+      "grad_norm": 0.36020344495773315,
+      "learning_rate": 4.273834415082057e-06,
+      "loss": 0.3235,
+      "step": 12368
+    },
+    {
+      "epoch": 0.9095521729538937,
+      "grad_norm": 0.4023871421813965,
+      "learning_rate": 4.26694893484686e-06,
+      "loss": 0.3188,
+      "step": 12369
+    },
+    {
+      "epoch": 0.9096257077726303,
+      "grad_norm": 0.33774617314338684,
+      "learning_rate": 4.2600688846913376e-06,
+      "loss": 0.3039,
+      "step": 12370
+    },
+    {
+      "epoch": 0.909699242591367,
+      "grad_norm": 0.4185110628604889,
+      "learning_rate": 4.253194265005745e-06,
+      "loss": 0.334,
+      "step": 12371
+    },
+    {
+      "epoch": 0.9097727774101037,
+      "grad_norm": 0.3783149719238281,
+      "learning_rate": 4.246325076180024e-06,
+      "loss": 0.3314,
+      "step": 12372
+    },
+    {
+      "epoch": 0.9098463122288404,
+      "grad_norm": 0.3857842981815338,
+      "learning_rate": 4.239461318603788e-06,
+      "loss": 0.3369,
+      "step": 12373
+    },
+    {
+      "epoch": 0.909919847047577,
+      "grad_norm": 0.4323977828025818,
+      "learning_rate": 4.232602992666379e-06,
+      "loss": 0.3414,
+      "step": 12374
+    },
+    {
+      "epoch": 0.9099933818663137,
+      "grad_norm": 0.38833433389663696,
+      "learning_rate": 4.225750098756798e-06,
+      "loss": 0.3144,
+      "step": 12375
+    },
+    {
+      "epoch": 0.9100669166850504,
+      "grad_norm": 0.4973389208316803,
+      "learning_rate": 4.218902637263733e-06,
+      "loss": 0.3181,
+      "step": 12376
+    },
+    {
+      "epoch": 0.9101404515037871,
+      "grad_norm": 0.3987575173377991,
+      "learning_rate": 4.212060608575585e-06,
+      "loss": 0.3039,
+      "step": 12377
+    },
+    {
+      "epoch": 0.9102139863225237,
+      "grad_norm": 0.3880026042461395,
+      "learning_rate": 4.205224013080478e-06,
+      "loss": 0.2917,
+      "step": 12378
+    },
+    {
+      "epoch": 0.9102875211412604,
+      "grad_norm": 0.4149444103240967,
+      "learning_rate": 4.198392851166133e-06,
+      "loss": 0.324,
+      "step": 12379
+    },
+    {
+      "epoch": 0.9103610559599971,
+      "grad_norm": 0.35679104924201965,
+      "learning_rate": 4.1915671232200745e-06,
+      "loss": 0.3064,
+      "step": 12380
+    },
+    {
+      "epoch": 0.9104345907787337,
+      "grad_norm": 0.4423486590385437,
+      "learning_rate": 4.184746829629416e-06,
+      "loss": 0.3294,
+      "step": 12381
+    },
+    {
+      "epoch": 0.9105081255974704,
+      "grad_norm": 0.41194313764572144,
+      "learning_rate": 4.1779319707810575e-06,
+      "loss": 0.2886,
+      "step": 12382
+    },
+    {
+      "epoch": 0.910581660416207,
+      "grad_norm": 0.3674742877483368,
+      "learning_rate": 4.171122547061501e-06,
+      "loss": 0.3007,
+      "step": 12383
+    },
+    {
+      "epoch": 0.9106551952349438,
+      "grad_norm": 0.4422037899494171,
+      "learning_rate": 4.164318558857006e-06,
+      "loss": 0.3016,
+      "step": 12384
+    },
+    {
+      "epoch": 0.9107287300536804,
+      "grad_norm": 0.3555283844470978,
+      "learning_rate": 4.157520006553517e-06,
+      "loss": 0.311,
+      "step": 12385
+    },
+    {
+      "epoch": 0.9108022648724171,
+      "grad_norm": 0.40977275371551514,
+      "learning_rate": 4.150726890536627e-06,
+      "loss": 0.2925,
+      "step": 12386
+    },
+    {
+      "epoch": 0.9108757996911537,
+      "grad_norm": 0.3645288348197937,
+      "learning_rate": 4.143939211191672e-06,
+      "loss": 0.3109,
+      "step": 12387
+    },
+    {
+      "epoch": 0.9109493345098905,
+      "grad_norm": 0.3789367973804474,
+      "learning_rate": 4.137156968903644e-06,
+      "loss": 0.3137,
+      "step": 12388
+    },
+    {
+      "epoch": 0.9110228693286271,
+      "grad_norm": 0.4595945477485657,
+      "learning_rate": 4.130380164057235e-06,
+      "loss": 0.3209,
+      "step": 12389
+    },
+    {
+      "epoch": 0.9110964041473638,
+      "grad_norm": 0.47762107849121094,
+      "learning_rate": 4.123608797036838e-06,
+      "loss": 0.3068,
+      "step": 12390
+    },
+    {
+      "epoch": 0.9111699389661004,
+      "grad_norm": 0.38172879815101624,
+      "learning_rate": 4.116842868226545e-06,
+      "loss": 0.3511,
+      "step": 12391
+    },
+    {
+      "epoch": 0.9112434737848372,
+      "grad_norm": 0.5774811506271362,
+      "learning_rate": 4.110082378010105e-06,
+      "loss": 0.282,
+      "step": 12392
+    },
+    {
+      "epoch": 0.9113170086035738,
+      "grad_norm": 0.4071440100669861,
+      "learning_rate": 4.103327326770989e-06,
+      "loss": 0.3259,
+      "step": 12393
+    },
+    {
+      "epoch": 0.9113905434223105,
+      "grad_norm": 0.5204576253890991,
+      "learning_rate": 4.096577714892347e-06,
+      "loss": 0.319,
+      "step": 12394
+    },
+    {
+      "epoch": 0.9114640782410471,
+      "grad_norm": 0.42858704924583435,
+      "learning_rate": 4.0898335427570375e-06,
+      "loss": 0.3184,
+      "step": 12395
+    },
+    {
+      "epoch": 0.9115376130597838,
+      "grad_norm": 0.37450799345970154,
+      "learning_rate": 4.083094810747579e-06,
+      "loss": 0.3118,
+      "step": 12396
+    },
+    {
+      "epoch": 0.9116111478785205,
+      "grad_norm": 0.41867202520370483,
+      "learning_rate": 4.076361519246208e-06,
+      "loss": 0.3322,
+      "step": 12397
+    },
+    {
+      "epoch": 0.9116846826972571,
+      "grad_norm": 0.3654906153678894,
+      "learning_rate": 4.0696336686348534e-06,
+      "loss": 0.3113,
+      "step": 12398
+    },
+    {
+      "epoch": 0.9117582175159938,
+      "grad_norm": 0.37004315853118896,
+      "learning_rate": 4.0629112592950995e-06,
+      "loss": 0.3633,
+      "step": 12399
+    },
+    {
+      "epoch": 0.9118317523347305,
+      "grad_norm": 0.5044448971748352,
+      "learning_rate": 4.056194291608273e-06,
+      "loss": 0.3092,
+      "step": 12400
+    },
+    {
+      "epoch": 0.9119052871534672,
+      "grad_norm": 0.3766420781612396,
+      "learning_rate": 4.04948276595537e-06,
+      "loss": 0.3172,
+      "step": 12401
+    },
+    {
+      "epoch": 0.9119788219722038,
+      "grad_norm": 0.39086443185806274,
+      "learning_rate": 4.042776682717054e-06,
+      "loss": 0.3141,
+      "step": 12402
+    },
+    {
+      "epoch": 0.9120523567909405,
+      "grad_norm": 0.40908294916152954,
+      "learning_rate": 4.036076042273728e-06,
+      "loss": 0.3374,
+      "step": 12403
+    },
+    {
+      "epoch": 0.9121258916096772,
+      "grad_norm": 0.35348647832870483,
+      "learning_rate": 4.029380845005426e-06,
+      "loss": 0.3015,
+      "step": 12404
+    },
+    {
+      "epoch": 0.9121994264284139,
+      "grad_norm": 0.3739855885505676,
+      "learning_rate": 4.022691091291941e-06,
+      "loss": 0.3002,
+      "step": 12405
+    },
+    {
+      "epoch": 0.9122729612471505,
+      "grad_norm": 0.3743153214454651,
+      "learning_rate": 4.016006781512682e-06,
+      "loss": 0.3247,
+      "step": 12406
+    },
+    {
+      "epoch": 0.9123464960658872,
+      "grad_norm": 0.41579514741897583,
+      "learning_rate": 4.009327916046823e-06,
+      "loss": 0.2967,
+      "step": 12407
+    },
+    {
+      "epoch": 0.9124200308846239,
+      "grad_norm": 0.3787899315357208,
+      "learning_rate": 4.002654495273206e-06,
+      "loss": 0.3027,
+      "step": 12408
+    },
+    {
+      "epoch": 0.9124935657033605,
+      "grad_norm": 0.4838390052318573,
+      "learning_rate": 3.995986519570316e-06,
+      "loss": 0.3412,
+      "step": 12409
+    },
+    {
+      "epoch": 0.9125671005220972,
+      "grad_norm": 0.4374041259288788,
+      "learning_rate": 3.989323989316384e-06,
+      "loss": 0.2978,
+      "step": 12410
+    },
+    {
+      "epoch": 0.9126406353408338,
+      "grad_norm": 0.4229987561702728,
+      "learning_rate": 3.9826669048893425e-06,
+      "loss": 0.3097,
+      "step": 12411
+    },
+    {
+      "epoch": 0.9127141701595706,
+      "grad_norm": 0.47174331545829773,
+      "learning_rate": 3.9760152666667435e-06,
+      "loss": 0.3336,
+      "step": 12412
+    },
+    {
+      "epoch": 0.9127877049783072,
+      "grad_norm": 0.4055821895599365,
+      "learning_rate": 3.969369075025897e-06,
+      "loss": 0.3176,
+      "step": 12413
+    },
+    {
+      "epoch": 0.9128612397970439,
+      "grad_norm": 0.40028902888298035,
+      "learning_rate": 3.962728330343802e-06,
+      "loss": 0.3043,
+      "step": 12414
+    },
+    {
+      "epoch": 0.9129347746157805,
+      "grad_norm": 0.4045827388763428,
+      "learning_rate": 3.9560930329970995e-06,
+      "loss": 0.2951,
+      "step": 12415
+    },
+    {
+      "epoch": 0.9130083094345173,
+      "grad_norm": 0.3966497480869293,
+      "learning_rate": 3.9494631833621566e-06,
+      "loss": 0.3428,
+      "step": 12416
+    },
+    {
+      "epoch": 0.9130818442532539,
+      "grad_norm": 0.3666483759880066,
+      "learning_rate": 3.942838781815017e-06,
+      "loss": 0.2946,
+      "step": 12417
+    },
+    {
+      "epoch": 0.9131553790719906,
+      "grad_norm": 0.38779330253601074,
+      "learning_rate": 3.936219828731458e-06,
+      "loss": 0.3158,
+      "step": 12418
+    },
+    {
+      "epoch": 0.9132289138907272,
+      "grad_norm": 0.38527557253837585,
+      "learning_rate": 3.929606324486868e-06,
+      "loss": 0.3342,
+      "step": 12419
+    },
+    {
+      "epoch": 0.913302448709464,
+      "grad_norm": 0.37070325016975403,
+      "learning_rate": 3.922998269456402e-06,
+      "loss": 0.3085,
+      "step": 12420
+    },
+    {
+      "epoch": 0.9133759835282006,
+      "grad_norm": 0.38730576634407043,
+      "learning_rate": 3.916395664014882e-06,
+      "loss": 0.3014,
+      "step": 12421
+    },
+    {
+      "epoch": 0.9134495183469373,
+      "grad_norm": 0.37935590744018555,
+      "learning_rate": 3.909798508536777e-06,
+      "loss": 0.3225,
+      "step": 12422
+    },
+    {
+      "epoch": 0.913523053165674,
+      "grad_norm": 0.430261492729187,
+      "learning_rate": 3.9032068033963195e-06,
+      "loss": 0.2998,
+      "step": 12423
+    },
+    {
+      "epoch": 0.9135965879844106,
+      "grad_norm": 0.39038699865341187,
+      "learning_rate": 3.8966205489674e-06,
+      "loss": 0.3022,
+      "step": 12424
+    },
+    {
+      "epoch": 0.9136701228031473,
+      "grad_norm": 0.450810045003891,
+      "learning_rate": 3.890039745623564e-06,
+      "loss": 0.3348,
+      "step": 12425
+    },
+    {
+      "epoch": 0.9137436576218839,
+      "grad_norm": 0.4049530029296875,
+      "learning_rate": 3.883464393738123e-06,
+      "loss": 0.3194,
+      "step": 12426
+    },
+    {
+      "epoch": 0.9138171924406207,
+      "grad_norm": 0.4183083772659302,
+      "learning_rate": 3.876894493684002e-06,
+      "loss": 0.32,
+      "step": 12427
+    },
+    {
+      "epoch": 0.9138907272593573,
+      "grad_norm": 0.349718302488327,
+      "learning_rate": 3.870330045833881e-06,
+      "loss": 0.38,
+      "step": 12428
+    },
+    {
+      "epoch": 0.913964262078094,
+      "grad_norm": 0.3825940787792206,
+      "learning_rate": 3.8637710505600726e-06,
+      "loss": 0.3077,
+      "step": 12429
+    },
+    {
+      "epoch": 0.9140377968968306,
+      "grad_norm": 0.4328307509422302,
+      "learning_rate": 3.857217508234634e-06,
+      "loss": 0.3136,
+      "step": 12430
+    },
+    {
+      "epoch": 0.9141113317155674,
+      "grad_norm": 0.370234876871109,
+      "learning_rate": 3.850669419229291e-06,
+      "loss": 0.3093,
+      "step": 12431
+    },
+    {
+      "epoch": 0.914184866534304,
+      "grad_norm": 0.3730565905570984,
+      "learning_rate": 3.844126783915436e-06,
+      "loss": 0.3275,
+      "step": 12432
+    },
+    {
+      "epoch": 0.9142584013530407,
+      "grad_norm": 0.46163612604141235,
+      "learning_rate": 3.837589602664182e-06,
+      "loss": 0.3189,
+      "step": 12433
+    },
+    {
+      "epoch": 0.9143319361717773,
+      "grad_norm": 0.41708049178123474,
+      "learning_rate": 3.8310578758463445e-06,
+      "loss": 0.309,
+      "step": 12434
+    },
+    {
+      "epoch": 0.9144054709905141,
+      "grad_norm": 0.38195526599884033,
+      "learning_rate": 3.824531603832393e-06,
+      "loss": 0.3256,
+      "step": 12435
+    },
+    {
+      "epoch": 0.9144790058092507,
+      "grad_norm": 0.39229607582092285,
+      "learning_rate": 3.818010786992499e-06,
+      "loss": 0.3041,
+      "step": 12436
+    },
+    {
+      "epoch": 0.9145525406279873,
+      "grad_norm": 0.5532341003417969,
+      "learning_rate": 3.8114954256965542e-06,
+      "loss": 0.3177,
+      "step": 12437
+    },
+    {
+      "epoch": 0.914626075446724,
+      "grad_norm": 0.35372570157051086,
+      "learning_rate": 3.804985520314097e-06,
+      "loss": 0.3148,
+      "step": 12438
+    },
+    {
+      "epoch": 0.9146996102654606,
+      "grad_norm": 0.4024505019187927,
+      "learning_rate": 3.798481071214366e-06,
+      "loss": 0.314,
+      "step": 12439
+    },
+    {
+      "epoch": 0.9147731450841974,
+      "grad_norm": 0.40003079175949097,
+      "learning_rate": 3.7919820787663097e-06,
+      "loss": 0.3245,
+      "step": 12440
+    },
+    {
+      "epoch": 0.914846679902934,
+      "grad_norm": 0.43571245670318604,
+      "learning_rate": 3.7854885433385888e-06,
+      "loss": 0.2979,
+      "step": 12441
+    },
+    {
+      "epoch": 0.9149202147216707,
+      "grad_norm": 0.41171473264694214,
+      "learning_rate": 3.7790004652994757e-06,
+      "loss": 0.3296,
+      "step": 12442
+    },
+    {
+      "epoch": 0.9149937495404074,
+      "grad_norm": 0.39947640895843506,
+      "learning_rate": 3.7725178450169984e-06,
+      "loss": 0.3246,
+      "step": 12443
+    },
+    {
+      "epoch": 0.9150672843591441,
+      "grad_norm": 0.48186755180358887,
+      "learning_rate": 3.7660406828588846e-06,
+      "loss": 0.3346,
+      "step": 12444
+    },
+    {
+      "epoch": 0.9151408191778807,
+      "grad_norm": 0.41475167870521545,
+      "learning_rate": 3.759568979192485e-06,
+      "loss": 0.2989,
+      "step": 12445
+    },
+    {
+      "epoch": 0.9152143539966174,
+      "grad_norm": 0.39201492071151733,
+      "learning_rate": 3.7531027343848944e-06,
+      "loss": 0.3286,
+      "step": 12446
+    },
+    {
+      "epoch": 0.9152878888153541,
+      "grad_norm": 0.36774030327796936,
+      "learning_rate": 3.7466419488029093e-06,
+      "loss": 0.3006,
+      "step": 12447
+    },
+    {
+      "epoch": 0.9153614236340908,
+      "grad_norm": 0.42248961329460144,
+      "learning_rate": 3.740186622812958e-06,
+      "loss": 0.3156,
+      "step": 12448
+    },
+    {
+      "epoch": 0.9154349584528274,
+      "grad_norm": 0.43139779567718506,
+      "learning_rate": 3.733736756781214e-06,
+      "loss": 0.3146,
+      "step": 12449
+    },
+    {
+      "epoch": 0.9155084932715641,
+      "grad_norm": 0.4576491415500641,
+      "learning_rate": 3.7272923510735193e-06,
+      "loss": 0.2996,
+      "step": 12450
+    },
+    {
+      "epoch": 0.9155820280903008,
+      "grad_norm": 0.4606596529483795,
+      "learning_rate": 3.7208534060554133e-06,
+      "loss": 0.312,
+      "step": 12451
+    },
+    {
+      "epoch": 0.9156555629090374,
+      "grad_norm": 0.3906884789466858,
+      "learning_rate": 3.7144199220920937e-06,
+      "loss": 0.2834,
+      "step": 12452
+    },
+    {
+      "epoch": 0.9157290977277741,
+      "grad_norm": 0.40887317061424255,
+      "learning_rate": 3.7079918995484795e-06,
+      "loss": 0.3059,
+      "step": 12453
+    },
+    {
+      "epoch": 0.9158026325465107,
+      "grad_norm": 0.5036699771881104,
+      "learning_rate": 3.7015693387892126e-06,
+      "loss": 0.3203,
+      "step": 12454
+    },
+    {
+      "epoch": 0.9158761673652475,
+      "grad_norm": 0.4407394528388977,
+      "learning_rate": 3.6951522401785344e-06,
+      "loss": 0.315,
+      "step": 12455
+    },
+    {
+      "epoch": 0.9159497021839841,
+      "grad_norm": 0.4326542019844055,
+      "learning_rate": 3.688740604080465e-06,
+      "loss": 0.3189,
+      "step": 12456
+    },
+    {
+      "epoch": 0.9160232370027208,
+      "grad_norm": 0.37702569365501404,
+      "learning_rate": 3.6823344308586805e-06,
+      "loss": 0.2844,
+      "step": 12457
+    },
+    {
+      "epoch": 0.9160967718214574,
+      "grad_norm": 0.43788009881973267,
+      "learning_rate": 3.6759337208765234e-06,
+      "loss": 0.3271,
+      "step": 12458
+    },
+    {
+      "epoch": 0.9161703066401942,
+      "grad_norm": 0.39360660314559937,
+      "learning_rate": 3.6695384744970586e-06,
+      "loss": 0.3179,
+      "step": 12459
+    },
+    {
+      "epoch": 0.9162438414589308,
+      "grad_norm": 0.43125101923942566,
+      "learning_rate": 3.6631486920830517e-06,
+      "loss": 0.3145,
+      "step": 12460
+    },
+    {
+      "epoch": 0.9163173762776675,
+      "grad_norm": 0.43140435218811035,
+      "learning_rate": 3.6567643739969126e-06,
+      "loss": 0.2988,
+      "step": 12461
+    },
+    {
+      "epoch": 0.9163909110964041,
+      "grad_norm": 0.4313714802265167,
+      "learning_rate": 3.650385520600763e-06,
+      "loss": 0.3188,
+      "step": 12462
+    },
+    {
+      "epoch": 0.9164644459151409,
+      "grad_norm": 0.4288271367549896,
+      "learning_rate": 3.6440121322564245e-06,
+      "loss": 0.287,
+      "step": 12463
+    },
+    {
+      "epoch": 0.9165379807338775,
+      "grad_norm": 0.3784022629261017,
+      "learning_rate": 3.6376442093254193e-06,
+      "loss": 0.3178,
+      "step": 12464
+    },
+    {
+      "epoch": 0.9166115155526141,
+      "grad_norm": 0.39759528636932373,
+      "learning_rate": 3.6312817521689248e-06,
+      "loss": 0.309,
+      "step": 12465
+    },
+    {
+      "epoch": 0.9166850503713508,
+      "grad_norm": 0.4431919753551483,
+      "learning_rate": 3.62492476114783e-06,
+      "loss": 0.2969,
+      "step": 12466
+    },
+    {
+      "epoch": 0.9167585851900875,
+      "grad_norm": 0.4033741354942322,
+      "learning_rate": 3.618573236622713e-06,
+      "loss": 0.3143,
+      "step": 12467
+    },
+    {
+      "epoch": 0.9168321200088242,
+      "grad_norm": 0.37059158086776733,
+      "learning_rate": 3.6122271789538307e-06,
+      "loss": 0.3146,
+      "step": 12468
+    },
+    {
+      "epoch": 0.9169056548275608,
+      "grad_norm": 0.3822936713695526,
+      "learning_rate": 3.6058865885011506e-06,
+      "loss": 0.3021,
+      "step": 12469
+    },
+    {
+      "epoch": 0.9169791896462975,
+      "grad_norm": 0.4060378670692444,
+      "learning_rate": 3.5995514656243177e-06,
+      "loss": 0.3181,
+      "step": 12470
+    },
+    {
+      "epoch": 0.9170527244650342,
+      "grad_norm": 0.39533427357673645,
+      "learning_rate": 3.5932218106826563e-06,
+      "loss": 0.3037,
+      "step": 12471
+    },
+    {
+      "epoch": 0.9171262592837709,
+      "grad_norm": 0.4834331274032593,
+      "learning_rate": 3.58689762403519e-06,
+      "loss": 0.3204,
+      "step": 12472
+    },
+    {
+      "epoch": 0.9171997941025075,
+      "grad_norm": 0.35594531893730164,
+      "learning_rate": 3.5805789060406546e-06,
+      "loss": 0.3054,
+      "step": 12473
+    },
+    {
+      "epoch": 0.9172733289212442,
+      "grad_norm": 0.4161876142024994,
+      "learning_rate": 3.5742656570574297e-06,
+      "loss": 0.2905,
+      "step": 12474
+    },
+    {
+      "epoch": 0.9173468637399809,
+      "grad_norm": 0.3419899344444275,
+      "learning_rate": 3.5679578774436174e-06,
+      "loss": 0.3003,
+      "step": 12475
+    },
+    {
+      "epoch": 0.9174203985587176,
+      "grad_norm": 0.3864138424396515,
+      "learning_rate": 3.5616555675569983e-06,
+      "loss": 0.3098,
+      "step": 12476
+    },
+    {
+      "epoch": 0.9174939333774542,
+      "grad_norm": 0.3989678621292114,
+      "learning_rate": 3.555358727755065e-06,
+      "loss": 0.2956,
+      "step": 12477
+    },
+    {
+      "epoch": 0.917567468196191,
+      "grad_norm": 0.3974872827529907,
+      "learning_rate": 3.549067358394953e-06,
+      "loss": 0.304,
+      "step": 12478
+    },
+    {
+      "epoch": 0.9176410030149276,
+      "grad_norm": 0.4101930558681488,
+      "learning_rate": 3.542781459833533e-06,
+      "loss": 0.3124,
+      "step": 12479
+    },
+    {
+      "epoch": 0.9177145378336642,
+      "grad_norm": 0.4162314236164093,
+      "learning_rate": 3.536501032427353e-06,
+      "loss": 0.2885,
+      "step": 12480
+    },
+    {
+      "epoch": 0.9177880726524009,
+      "grad_norm": 0.37092435359954834,
+      "learning_rate": 3.5302260765326277e-06,
+      "loss": 0.3035,
+      "step": 12481
+    },
+    {
+      "epoch": 0.9178616074711375,
+      "grad_norm": 0.3876136541366577,
+      "learning_rate": 3.523956592505284e-06,
+      "loss": 0.3121,
+      "step": 12482
+    },
+    {
+      "epoch": 0.9179351422898743,
+      "grad_norm": 0.4062784016132355,
+      "learning_rate": 3.5176925807009485e-06,
+      "loss": 0.2993,
+      "step": 12483
+    },
+    {
+      "epoch": 0.9180086771086109,
+      "grad_norm": 0.3610823452472687,
+      "learning_rate": 3.5114340414749035e-06,
+      "loss": 0.2888,
+      "step": 12484
+    },
+    {
+      "epoch": 0.9180822119273476,
+      "grad_norm": 0.4039155840873718,
+      "learning_rate": 3.505180975182154e-06,
+      "loss": 0.3254,
+      "step": 12485
+    },
+    {
+      "epoch": 0.9181557467460842,
+      "grad_norm": 0.40599581599235535,
+      "learning_rate": 3.498933382177372e-06,
+      "loss": 0.3111,
+      "step": 12486
+    },
+    {
+      "epoch": 0.918229281564821,
+      "grad_norm": 0.41686299443244934,
+      "learning_rate": 3.492691262814929e-06,
+      "loss": 0.3083,
+      "step": 12487
+    },
+    {
+      "epoch": 0.9183028163835576,
+      "grad_norm": 0.4268966019153595,
+      "learning_rate": 3.4864546174488864e-06,
+      "loss": 0.3071,
+      "step": 12488
+    },
+    {
+      "epoch": 0.9183763512022943,
+      "grad_norm": 0.4592595100402832,
+      "learning_rate": 3.480223446432984e-06,
+      "loss": 0.3131,
+      "step": 12489
+    },
+    {
+      "epoch": 0.918449886021031,
+      "grad_norm": 0.45335501432418823,
+      "learning_rate": 3.4739977501206833e-06,
+      "loss": 0.2953,
+      "step": 12490
+    },
+    {
+      "epoch": 0.9185234208397677,
+      "grad_norm": 0.46626558899879456,
+      "learning_rate": 3.4677775288650903e-06,
+      "loss": 0.3282,
+      "step": 12491
+    },
+    {
+      "epoch": 0.9185969556585043,
+      "grad_norm": 0.3825525641441345,
+      "learning_rate": 3.4615627830190343e-06,
+      "loss": 0.2985,
+      "step": 12492
+    },
+    {
+      "epoch": 0.9186704904772409,
+      "grad_norm": 0.42326027154922485,
+      "learning_rate": 3.4553535129350222e-06,
+      "loss": 0.3104,
+      "step": 12493
+    },
+    {
+      "epoch": 0.9187440252959777,
+      "grad_norm": 0.4211302697658539,
+      "learning_rate": 3.449149718965239e-06,
+      "loss": 0.3039,
+      "step": 12494
+    },
+    {
+      "epoch": 0.9188175601147143,
+      "grad_norm": 0.365681916475296,
+      "learning_rate": 3.4429514014615695e-06,
+      "loss": 0.3298,
+      "step": 12495
+    },
+    {
+      "epoch": 0.918891094933451,
+      "grad_norm": 0.42569392919540405,
+      "learning_rate": 3.43675856077561e-06,
+      "loss": 0.3174,
+      "step": 12496
+    },
+    {
+      "epoch": 0.9189646297521876,
+      "grad_norm": 0.39004355669021606,
+      "learning_rate": 3.4305711972586253e-06,
+      "loss": 0.3255,
+      "step": 12497
+    },
+    {
+      "epoch": 0.9190381645709244,
+      "grad_norm": 0.3648390769958496,
+      "learning_rate": 3.4243893112615333e-06,
+      "loss": 0.2884,
+      "step": 12498
+    },
+    {
+      "epoch": 0.919111699389661,
+      "grad_norm": 0.4174000918865204,
+      "learning_rate": 3.4182129031349985e-06,
+      "loss": 0.3266,
+      "step": 12499
+    },
+    {
+      "epoch": 0.9191852342083977,
+      "grad_norm": 0.36228662729263306,
+      "learning_rate": 3.412041973229374e-06,
+      "loss": 0.2998,
+      "step": 12500
+    },
+    {
+      "epoch": 0.9192587690271343,
+      "grad_norm": 0.43555736541748047,
+      "learning_rate": 3.4058765218946465e-06,
+      "loss": 0.3031,
+      "step": 12501
+    },
+    {
+      "epoch": 0.9193323038458711,
+      "grad_norm": 0.39482876658439636,
+      "learning_rate": 3.3997165494805472e-06,
+      "loss": 0.3013,
+      "step": 12502
+    },
+    {
+      "epoch": 0.9194058386646077,
+      "grad_norm": 0.49014195799827576,
+      "learning_rate": 3.3935620563364745e-06,
+      "loss": 0.3013,
+      "step": 12503
+    },
+    {
+      "epoch": 0.9194793734833444,
+      "grad_norm": 0.387024849653244,
+      "learning_rate": 3.3874130428115046e-06,
+      "loss": 0.3253,
+      "step": 12504
+    },
+    {
+      "epoch": 0.919552908302081,
+      "grad_norm": 0.4492528438568115,
+      "learning_rate": 3.381269509254437e-06,
+      "loss": 0.32,
+      "step": 12505
+    },
+    {
+      "epoch": 0.9196264431208178,
+      "grad_norm": 0.4507391154766083,
+      "learning_rate": 3.3751314560137247e-06,
+      "loss": 0.3207,
+      "step": 12506
+    },
+    {
+      "epoch": 0.9196999779395544,
+      "grad_norm": 0.45294150710105896,
+      "learning_rate": 3.368998883437524e-06,
+      "loss": 0.3333,
+      "step": 12507
+    },
+    {
+      "epoch": 0.919773512758291,
+      "grad_norm": 0.4811859428882599,
+      "learning_rate": 3.362871791873701e-06,
+      "loss": 0.3,
+      "step": 12508
+    },
+    {
+      "epoch": 0.9198470475770277,
+      "grad_norm": 0.3753063976764679,
+      "learning_rate": 3.3567501816697543e-06,
+      "loss": 0.291,
+      "step": 12509
+    },
+    {
+      "epoch": 0.9199205823957644,
+      "grad_norm": 0.4204675257205963,
+      "learning_rate": 3.3506340531729518e-06,
+      "loss": 0.3183,
+      "step": 12510
+    },
+    {
+      "epoch": 0.9199941172145011,
+      "grad_norm": 0.4535338878631592,
+      "learning_rate": 3.344523406730171e-06,
+      "loss": 0.3167,
+      "step": 12511
+    },
+    {
+      "epoch": 0.9200676520332377,
+      "grad_norm": 0.40330275893211365,
+      "learning_rate": 3.3384182426880127e-06,
+      "loss": 0.2803,
+      "step": 12512
+    },
+    {
+      "epoch": 0.9201411868519744,
+      "grad_norm": 0.3880481719970703,
+      "learning_rate": 3.332318561392811e-06,
+      "loss": 0.3033,
+      "step": 12513
+    },
+    {
+      "epoch": 0.9202147216707111,
+      "grad_norm": 0.3858894109725952,
+      "learning_rate": 3.3262243631905e-06,
+      "loss": 0.2802,
+      "step": 12514
+    },
+    {
+      "epoch": 0.9202882564894478,
+      "grad_norm": 0.4791482388973236,
+      "learning_rate": 3.32013564842677e-06,
+      "loss": 0.3019,
+      "step": 12515
+    },
+    {
+      "epoch": 0.9203617913081844,
+      "grad_norm": 0.38610804080963135,
+      "learning_rate": 3.314052417446978e-06,
+      "loss": 0.3644,
+      "step": 12516
+    },
+    {
+      "epoch": 0.9204353261269211,
+      "grad_norm": 0.5227265357971191,
+      "learning_rate": 3.30797467059617e-06,
+      "loss": 0.2985,
+      "step": 12517
+    },
+    {
+      "epoch": 0.9205088609456578,
+      "grad_norm": 0.3696402609348297,
+      "learning_rate": 3.3019024082190706e-06,
+      "loss": 0.2897,
+      "step": 12518
+    },
+    {
+      "epoch": 0.9205823957643945,
+      "grad_norm": 0.36798691749572754,
+      "learning_rate": 3.295835630660138e-06,
+      "loss": 0.3194,
+      "step": 12519
+    },
+    {
+      "epoch": 0.9206559305831311,
+      "grad_norm": 0.4637029469013214,
+      "learning_rate": 3.2897743382634625e-06,
+      "loss": 0.3037,
+      "step": 12520
+    },
+    {
+      "epoch": 0.9207294654018677,
+      "grad_norm": 0.4008382260799408,
+      "learning_rate": 3.283718531372837e-06,
+      "loss": 0.3551,
+      "step": 12521
+    },
+    {
+      "epoch": 0.9208030002206045,
+      "grad_norm": 0.4361030161380768,
+      "learning_rate": 3.2776682103317634e-06,
+      "loss": 0.3396,
+      "step": 12522
+    },
+    {
+      "epoch": 0.9208765350393411,
+      "grad_norm": 0.5320006012916565,
+      "learning_rate": 3.2716233754834347e-06,
+      "loss": 0.3017,
+      "step": 12523
+    },
+    {
+      "epoch": 0.9209500698580778,
+      "grad_norm": 0.453865110874176,
+      "learning_rate": 3.265584027170698e-06,
+      "loss": 0.3086,
+      "step": 12524
+    },
+    {
+      "epoch": 0.9210236046768144,
+      "grad_norm": 0.41679641604423523,
+      "learning_rate": 3.259550165736125e-06,
+      "loss": 0.2992,
+      "step": 12525
+    },
+    {
+      "epoch": 0.9210971394955512,
+      "grad_norm": 0.4866052567958832,
+      "learning_rate": 3.253521791521963e-06,
+      "loss": 0.2972,
+      "step": 12526
+    },
+    {
+      "epoch": 0.9211706743142878,
+      "grad_norm": 0.38841328024864197,
+      "learning_rate": 3.24749890487015e-06,
+      "loss": 0.3136,
+      "step": 12527
+    },
+    {
+      "epoch": 0.9212442091330245,
+      "grad_norm": 0.43513405323028564,
+      "learning_rate": 3.2414815061222902e-06,
+      "loss": 0.3186,
+      "step": 12528
+    },
+    {
+      "epoch": 0.9213177439517611,
+      "grad_norm": 0.3930714726448059,
+      "learning_rate": 3.2354695956197335e-06,
+      "loss": 0.3176,
+      "step": 12529
+    },
+    {
+      "epoch": 0.9213912787704979,
+      "grad_norm": 0.4270578622817993,
+      "learning_rate": 3.229463173703451e-06,
+      "loss": 0.2943,
+      "step": 12530
+    },
+    {
+      "epoch": 0.9214648135892345,
+      "grad_norm": 0.39195600152015686,
+      "learning_rate": 3.223462240714159e-06,
+      "loss": 0.3071,
+      "step": 12531
+    },
+    {
+      "epoch": 0.9215383484079712,
+      "grad_norm": 0.39136576652526855,
+      "learning_rate": 3.217466796992197e-06,
+      "loss": 0.3143,
+      "step": 12532
+    },
+    {
+      "epoch": 0.9216118832267078,
+      "grad_norm": 0.36886072158813477,
+      "learning_rate": 3.21147684287767e-06,
+      "loss": 0.3115,
+      "step": 12533
+    },
+    {
+      "epoch": 0.9216854180454446,
+      "grad_norm": 0.407470166683197,
+      "learning_rate": 3.2054923787103173e-06,
+      "loss": 0.3319,
+      "step": 12534
+    },
+    {
+      "epoch": 0.9217589528641812,
+      "grad_norm": 0.43305402994155884,
+      "learning_rate": 3.1995134048295904e-06,
+      "loss": 0.2902,
+      "step": 12535
+    },
+    {
+      "epoch": 0.9218324876829178,
+      "grad_norm": 0.4049653112888336,
+      "learning_rate": 3.1935399215746287e-06,
+      "loss": 0.305,
+      "step": 12536
+    },
+    {
+      "epoch": 0.9219060225016545,
+      "grad_norm": 0.3584001660346985,
+      "learning_rate": 3.1875719292842277e-06,
+      "loss": 0.2794,
+      "step": 12537
+    },
+    {
+      "epoch": 0.9219795573203912,
+      "grad_norm": 0.3664934039115906,
+      "learning_rate": 3.181609428296928e-06,
+      "loss": 0.2994,
+      "step": 12538
+    },
+    {
+      "epoch": 0.9220530921391279,
+      "grad_norm": 0.39442920684814453,
+      "learning_rate": 3.1756524189509253e-06,
+      "loss": 0.3095,
+      "step": 12539
+    },
+    {
+      "epoch": 0.9221266269578645,
+      "grad_norm": 0.3939058184623718,
+      "learning_rate": 3.1697009015840827e-06,
+      "loss": 0.3179,
+      "step": 12540
+    },
+    {
+      "epoch": 0.9222001617766012,
+      "grad_norm": 0.50575852394104,
+      "learning_rate": 3.1637548765339974e-06,
+      "loss": 0.2852,
+      "step": 12541
+    },
+    {
+      "epoch": 0.9222736965953379,
+      "grad_norm": 0.47058790922164917,
+      "learning_rate": 3.1578143441379436e-06,
+      "loss": 0.3369,
+      "step": 12542
+    },
+    {
+      "epoch": 0.9223472314140746,
+      "grad_norm": 0.3721347749233246,
+      "learning_rate": 3.1518793047328632e-06,
+      "loss": 0.3082,
+      "step": 12543
+    },
+    {
+      "epoch": 0.9224207662328112,
+      "grad_norm": 0.42583027482032776,
+      "learning_rate": 3.1459497586553863e-06,
+      "loss": 0.3236,
+      "step": 12544
+    },
+    {
+      "epoch": 0.922494301051548,
+      "grad_norm": 0.38098499178886414,
+      "learning_rate": 3.1400257062418557e-06,
+      "loss": 0.2836,
+      "step": 12545
+    },
+    {
+      "epoch": 0.9225678358702846,
+      "grad_norm": 0.616654098033905,
+      "learning_rate": 3.1341071478282915e-06,
+      "loss": 0.3451,
+      "step": 12546
+    },
+    {
+      "epoch": 0.9226413706890213,
+      "grad_norm": 0.4547909200191498,
+      "learning_rate": 3.128194083750391e-06,
+      "loss": 0.3054,
+      "step": 12547
+    },
+    {
+      "epoch": 0.9227149055077579,
+      "grad_norm": 0.41336581110954285,
+      "learning_rate": 3.122286514343542e-06,
+      "loss": 0.299,
+      "step": 12548
+    },
+    {
+      "epoch": 0.9227884403264945,
+      "grad_norm": 0.47918015718460083,
+      "learning_rate": 3.1163844399428653e-06,
+      "loss": 0.3476,
+      "step": 12549
+    },
+    {
+      "epoch": 0.9228619751452313,
+      "grad_norm": 0.41955244541168213,
+      "learning_rate": 3.1104878608830934e-06,
+      "loss": 0.3039,
+      "step": 12550
+    },
+    {
+      "epoch": 0.9229355099639679,
+      "grad_norm": 0.3985159993171692,
+      "learning_rate": 3.104596777498703e-06,
+      "loss": 0.3003,
+      "step": 12551
+    },
+    {
+      "epoch": 0.9230090447827046,
+      "grad_norm": 0.37710142135620117,
+      "learning_rate": 3.0987111901238487e-06,
+      "loss": 0.2896,
+      "step": 12552
+    },
+    {
+      "epoch": 0.9230825796014412,
+      "grad_norm": 0.3680197298526764,
+      "learning_rate": 3.0928310990923525e-06,
+      "loss": 0.2791,
+      "step": 12553
+    },
+    {
+      "epoch": 0.923156114420178,
+      "grad_norm": 0.38727277517318726,
+      "learning_rate": 3.086956504737748e-06,
+      "loss": 0.2815,
+      "step": 12554
+    },
+    {
+      "epoch": 0.9232296492389146,
+      "grad_norm": 0.3743537664413452,
+      "learning_rate": 3.0810874073932573e-06,
+      "loss": 0.3153,
+      "step": 12555
+    },
+    {
+      "epoch": 0.9233031840576513,
+      "grad_norm": 0.3990650177001953,
+      "learning_rate": 3.0752238073917695e-06,
+      "loss": 0.2993,
+      "step": 12556
+    },
+    {
+      "epoch": 0.923376718876388,
+      "grad_norm": 0.41652435064315796,
+      "learning_rate": 3.069365705065863e-06,
+      "loss": 0.306,
+      "step": 12557
+    },
+    {
+      "epoch": 0.9234502536951247,
+      "grad_norm": 0.40294569730758667,
+      "learning_rate": 3.063513100747828e-06,
+      "loss": 0.2882,
+      "step": 12558
+    },
+    {
+      "epoch": 0.9235237885138613,
+      "grad_norm": 0.3958832323551178,
+      "learning_rate": 3.057665994769654e-06,
+      "loss": 0.2864,
+      "step": 12559
+    },
+    {
+      "epoch": 0.923597323332598,
+      "grad_norm": 0.49150729179382324,
+      "learning_rate": 3.0518243874629427e-06,
+      "loss": 0.2959,
+      "step": 12560
+    },
+    {
+      "epoch": 0.9236708581513347,
+      "grad_norm": 0.3807083070278168,
+      "learning_rate": 3.0459882791590843e-06,
+      "loss": 0.2987,
+      "step": 12561
+    },
+    {
+      "epoch": 0.9237443929700714,
+      "grad_norm": 0.40860626101493835,
+      "learning_rate": 3.040157670189092e-06,
+      "loss": 0.2803,
+      "step": 12562
+    },
+    {
+      "epoch": 0.923817927788808,
+      "grad_norm": 0.46117886900901794,
+      "learning_rate": 3.0343325608836682e-06,
+      "loss": 0.2948,
+      "step": 12563
+    },
+    {
+      "epoch": 0.9238914626075446,
+      "grad_norm": 0.43663543462753296,
+      "learning_rate": 3.0285129515732367e-06,
+      "loss": 0.3104,
+      "step": 12564
+    },
+    {
+      "epoch": 0.9239649974262814,
+      "grad_norm": 0.3736487627029419,
+      "learning_rate": 3.022698842587901e-06,
+      "loss": 0.2863,
+      "step": 12565
+    },
+    {
+      "epoch": 0.924038532245018,
+      "grad_norm": 0.39043936133384705,
+      "learning_rate": 3.01689023425743e-06,
+      "loss": 0.3004,
+      "step": 12566
+    },
+    {
+      "epoch": 0.9241120670637547,
+      "grad_norm": 0.3816046416759491,
+      "learning_rate": 3.011087126911283e-06,
+      "loss": 0.315,
+      "step": 12567
+    },
+    {
+      "epoch": 0.9241856018824913,
+      "grad_norm": 0.4305155277252197,
+      "learning_rate": 3.0052895208786293e-06,
+      "loss": 0.3158,
+      "step": 12568
+    },
+    {
+      "epoch": 0.9242591367012281,
+      "grad_norm": 0.39069896936416626,
+      "learning_rate": 2.9994974164883282e-06,
+      "loss": 0.303,
+      "step": 12569
+    },
+    {
+      "epoch": 0.9243326715199647,
+      "grad_norm": 0.44182702898979187,
+      "learning_rate": 2.9937108140688954e-06,
+      "loss": 0.3126,
+      "step": 12570
+    },
+    {
+      "epoch": 0.9244062063387014,
+      "grad_norm": 0.4033292829990387,
+      "learning_rate": 2.987929713948556e-06,
+      "loss": 0.3063,
+      "step": 12571
+    },
+    {
+      "epoch": 0.924479741157438,
+      "grad_norm": 0.4600944519042969,
+      "learning_rate": 2.982154116455238e-06,
+      "loss": 0.3132,
+      "step": 12572
+    },
+    {
+      "epoch": 0.9245532759761748,
+      "grad_norm": 0.4010082483291626,
+      "learning_rate": 2.9763840219165116e-06,
+      "loss": 0.2988,
+      "step": 12573
+    },
+    {
+      "epoch": 0.9246268107949114,
+      "grad_norm": 0.4551703631877899,
+      "learning_rate": 2.9706194306596823e-06,
+      "loss": 0.3025,
+      "step": 12574
+    },
+    {
+      "epoch": 0.9247003456136481,
+      "grad_norm": 0.37216436862945557,
+      "learning_rate": 2.9648603430117217e-06,
+      "loss": 0.3195,
+      "step": 12575
+    },
+    {
+      "epoch": 0.9247738804323847,
+      "grad_norm": 0.3897989094257355,
+      "learning_rate": 2.9591067592992794e-06,
+      "loss": 0.2918,
+      "step": 12576
+    },
+    {
+      "epoch": 0.9248474152511214,
+      "grad_norm": 0.41530898213386536,
+      "learning_rate": 2.9533586798487165e-06,
+      "loss": 0.3135,
+      "step": 12577
+    },
+    {
+      "epoch": 0.9249209500698581,
+      "grad_norm": 0.42115870118141174,
+      "learning_rate": 2.947616104986073e-06,
+      "loss": 0.3219,
+      "step": 12578
+    },
+    {
+      "epoch": 0.9249944848885947,
+      "grad_norm": 0.3936525285243988,
+      "learning_rate": 2.9418790350370763e-06,
+      "loss": 0.3252,
+      "step": 12579
+    },
+    {
+      "epoch": 0.9250680197073314,
+      "grad_norm": 0.3856762647628784,
+      "learning_rate": 2.9361474703271108e-06,
+      "loss": 0.3056,
+      "step": 12580
+    },
+    {
+      "epoch": 0.925141554526068,
+      "grad_norm": 0.4213770031929016,
+      "learning_rate": 2.9304214111812945e-06,
+      "loss": 0.3065,
+      "step": 12581
+    },
+    {
+      "epoch": 0.9252150893448048,
+      "grad_norm": 0.4456462264060974,
+      "learning_rate": 2.924700857924434e-06,
+      "loss": 0.3355,
+      "step": 12582
+    },
+    {
+      "epoch": 0.9252886241635414,
+      "grad_norm": 0.4170643389225006,
+      "learning_rate": 2.918985810880981e-06,
+      "loss": 0.3121,
+      "step": 12583
+    },
+    {
+      "epoch": 0.9253621589822781,
+      "grad_norm": 0.46571120619773865,
+      "learning_rate": 2.913276270375098e-06,
+      "loss": 0.3076,
+      "step": 12584
+    },
+    {
+      "epoch": 0.9254356938010148,
+      "grad_norm": 0.38390854001045227,
+      "learning_rate": 2.9075722367306604e-06,
+      "loss": 0.2988,
+      "step": 12585
+    },
+    {
+      "epoch": 0.9255092286197515,
+      "grad_norm": 0.3796612620353699,
+      "learning_rate": 2.9018737102711745e-06,
+      "loss": 0.294,
+      "step": 12586
+    },
+    {
+      "epoch": 0.9255827634384881,
+      "grad_norm": 0.4449961483478546,
+      "learning_rate": 2.896180691319883e-06,
+      "loss": 0.2972,
+      "step": 12587
+    },
+    {
+      "epoch": 0.9256562982572248,
+      "grad_norm": 0.39639243483543396,
+      "learning_rate": 2.890493180199716e-06,
+      "loss": 0.3225,
+      "step": 12588
+    },
+    {
+      "epoch": 0.9257298330759615,
+      "grad_norm": 0.4099622964859009,
+      "learning_rate": 2.884811177233249e-06,
+      "loss": 0.3263,
+      "step": 12589
+    },
+    {
+      "epoch": 0.9258033678946982,
+      "grad_norm": 0.39708411693573,
+      "learning_rate": 2.87913468274279e-06,
+      "loss": 0.3037,
+      "step": 12590
+    },
+    {
+      "epoch": 0.9258769027134348,
+      "grad_norm": 0.395522803068161,
+      "learning_rate": 2.873463697050305e-06,
+      "loss": 0.3328,
+      "step": 12591
+    },
+    {
+      "epoch": 0.9259504375321714,
+      "grad_norm": 0.3920922875404358,
+      "learning_rate": 2.867798220477469e-06,
+      "loss": 0.3522,
+      "step": 12592
+    },
+    {
+      "epoch": 0.9260239723509082,
+      "grad_norm": 0.40847334265708923,
+      "learning_rate": 2.8621382533456033e-06,
+      "loss": 0.3471,
+      "step": 12593
+    },
+    {
+      "epoch": 0.9260975071696448,
+      "grad_norm": 0.3897688090801239,
+      "learning_rate": 2.856483795975784e-06,
+      "loss": 0.3589,
+      "step": 12594
+    },
+    {
+      "epoch": 0.9261710419883815,
+      "grad_norm": 0.4040391445159912,
+      "learning_rate": 2.8508348486887217e-06,
+      "loss": 0.3159,
+      "step": 12595
+    },
+    {
+      "epoch": 0.9262445768071181,
+      "grad_norm": 0.41659459471702576,
+      "learning_rate": 2.8451914118048263e-06,
+      "loss": 0.3557,
+      "step": 12596
+    },
+    {
+      "epoch": 0.9263181116258549,
+      "grad_norm": 0.4273473620414734,
+      "learning_rate": 2.8395534856442084e-06,
+      "loss": 0.2943,
+      "step": 12597
+    },
+    {
+      "epoch": 0.9263916464445915,
+      "grad_norm": 0.4538755714893341,
+      "learning_rate": 2.8339210705266682e-06,
+      "loss": 0.3474,
+      "step": 12598
+    },
+    {
+      "epoch": 0.9264651812633282,
+      "grad_norm": 0.40069469809532166,
+      "learning_rate": 2.82829416677165e-06,
+      "loss": 0.3241,
+      "step": 12599
+    },
+    {
+      "epoch": 0.9265387160820648,
+      "grad_norm": 0.5075940489768982,
+      "learning_rate": 2.822672774698343e-06,
+      "loss": 0.3028,
+      "step": 12600
+    },
+    {
+      "epoch": 0.9266122509008016,
+      "grad_norm": 0.4134211540222168,
+      "learning_rate": 2.8170568946255917e-06,
+      "loss": 0.3101,
+      "step": 12601
+    },
+    {
+      "epoch": 0.9266857857195382,
+      "grad_norm": 0.38034477829933167,
+      "learning_rate": 2.8114465268719415e-06,
+      "loss": 0.3263,
+      "step": 12602
+    },
+    {
+      "epoch": 0.9267593205382749,
+      "grad_norm": 0.45565247535705566,
+      "learning_rate": 2.8058416717556046e-06,
+      "loss": 0.3134,
+      "step": 12603
+    },
+    {
+      "epoch": 0.9268328553570115,
+      "grad_norm": 0.35809287428855896,
+      "learning_rate": 2.8002423295944935e-06,
+      "loss": 0.3701,
+      "step": 12604
+    },
+    {
+      "epoch": 0.9269063901757482,
+      "grad_norm": 0.3757377564907074,
+      "learning_rate": 2.7946485007062316e-06,
+      "loss": 0.3131,
+      "step": 12605
+    },
+    {
+      "epoch": 0.9269799249944849,
+      "grad_norm": 0.3758884370326996,
+      "learning_rate": 2.7890601854080768e-06,
+      "loss": 0.2833,
+      "step": 12606
+    },
+    {
+      "epoch": 0.9270534598132215,
+      "grad_norm": 0.40869179368019104,
+      "learning_rate": 2.7834773840170192e-06,
+      "loss": 0.3416,
+      "step": 12607
+    },
+    {
+      "epoch": 0.9271269946319582,
+      "grad_norm": 0.46668121218681335,
+      "learning_rate": 2.7779000968497393e-06,
+      "loss": 0.3273,
+      "step": 12608
+    },
+    {
+      "epoch": 0.9272005294506949,
+      "grad_norm": 0.4114082157611847,
+      "learning_rate": 2.7723283242225507e-06,
+      "loss": 0.3238,
+      "step": 12609
+    },
+    {
+      "epoch": 0.9272740642694316,
+      "grad_norm": 0.4520261287689209,
+      "learning_rate": 2.7667620664515115e-06,
+      "loss": 0.323,
+      "step": 12610
+    },
+    {
+      "epoch": 0.9273475990881682,
+      "grad_norm": 0.5085795521736145,
+      "learning_rate": 2.761201323852347e-06,
+      "loss": 0.3321,
+      "step": 12611
+    },
+    {
+      "epoch": 0.927421133906905,
+      "grad_norm": 0.3922363817691803,
+      "learning_rate": 2.755646096740461e-06,
+      "loss": 0.3165,
+      "step": 12612
+    },
+    {
+      "epoch": 0.9274946687256416,
+      "grad_norm": 0.4747444689273834,
+      "learning_rate": 2.750096385430967e-06,
+      "loss": 0.3468,
+      "step": 12613
+    },
+    {
+      "epoch": 0.9275682035443783,
+      "grad_norm": 0.4456048905849457,
+      "learning_rate": 2.7445521902386138e-06,
+      "loss": 0.3461,
+      "step": 12614
+    },
+    {
+      "epoch": 0.9276417383631149,
+      "grad_norm": 0.40030208230018616,
+      "learning_rate": 2.739013511477917e-06,
+      "loss": 0.3438,
+      "step": 12615
+    },
+    {
+      "epoch": 0.9277152731818517,
+      "grad_norm": 0.43622955679893494,
+      "learning_rate": 2.7334803494630136e-06,
+      "loss": 0.3292,
+      "step": 12616
+    },
+    {
+      "epoch": 0.9277888080005883,
+      "grad_norm": 0.4212262034416199,
+      "learning_rate": 2.727952704507741e-06,
+      "loss": 0.3143,
+      "step": 12617
+    },
+    {
+      "epoch": 0.927862342819325,
+      "grad_norm": 0.3864156901836395,
+      "learning_rate": 2.722430576925661e-06,
+      "loss": 0.3447,
+      "step": 12618
+    },
+    {
+      "epoch": 0.9279358776380616,
+      "grad_norm": 0.42168161273002625,
+      "learning_rate": 2.716913967029977e-06,
+      "loss": 0.3505,
+      "step": 12619
+    },
+    {
+      "epoch": 0.9280094124567982,
+      "grad_norm": 0.41601425409317017,
+      "learning_rate": 2.7114028751335952e-06,
+      "loss": 0.316,
+      "step": 12620
+    },
+    {
+      "epoch": 0.928082947275535,
+      "grad_norm": 0.3661242723464966,
+      "learning_rate": 2.7058973015491205e-06,
+      "loss": 0.3063,
+      "step": 12621
+    },
+    {
+      "epoch": 0.9281564820942716,
+      "grad_norm": 0.4625905454158783,
+      "learning_rate": 2.700397246588826e-06,
+      "loss": 0.3373,
+      "step": 12622
+    },
+    {
+      "epoch": 0.9282300169130083,
+      "grad_norm": 0.4525626301765442,
+      "learning_rate": 2.694902710564684e-06,
+      "loss": 0.3392,
+      "step": 12623
+    },
+    {
+      "epoch": 0.928303551731745,
+      "grad_norm": 0.3719938099384308,
+      "learning_rate": 2.6894136937883565e-06,
+      "loss": 0.3283,
+      "step": 12624
+    },
+    {
+      "epoch": 0.9283770865504817,
+      "grad_norm": 0.3795459270477295,
+      "learning_rate": 2.683930196571183e-06,
+      "loss": 0.298,
+      "step": 12625
+    },
+    {
+      "epoch": 0.9284506213692183,
+      "grad_norm": 0.39353248476982117,
+      "learning_rate": 2.67845221922417e-06,
+      "loss": 0.293,
+      "step": 12626
+    },
+    {
+      "epoch": 0.928524156187955,
+      "grad_norm": 0.3910842537879944,
+      "learning_rate": 2.6729797620580698e-06,
+      "loss": 0.3099,
+      "step": 12627
+    },
+    {
+      "epoch": 0.9285976910066917,
+      "grad_norm": 0.37500718235969543,
+      "learning_rate": 2.6675128253832673e-06,
+      "loss": 0.2997,
+      "step": 12628
+    },
+    {
+      "epoch": 0.9286712258254284,
+      "grad_norm": 0.4105183184146881,
+      "learning_rate": 2.662051409509858e-06,
+      "loss": 0.2998,
+      "step": 12629
+    },
+    {
+      "epoch": 0.928744760644165,
+      "grad_norm": 0.4050842821598053,
+      "learning_rate": 2.6565955147476173e-06,
+      "loss": 0.3227,
+      "step": 12630
+    },
+    {
+      "epoch": 0.9288182954629017,
+      "grad_norm": 0.3782486021518707,
+      "learning_rate": 2.6511451414060084e-06,
+      "loss": 0.3249,
+      "step": 12631
+    },
+    {
+      "epoch": 0.9288918302816384,
+      "grad_norm": 0.3876364827156067,
+      "learning_rate": 2.645700289794184e-06,
+      "loss": 0.3009,
+      "step": 12632
+    },
+    {
+      "epoch": 0.928965365100375,
+      "grad_norm": 0.4013892710208893,
+      "learning_rate": 2.640260960220975e-06,
+      "loss": 0.3348,
+      "step": 12633
+    },
+    {
+      "epoch": 0.9290388999191117,
+      "grad_norm": 0.4261789917945862,
+      "learning_rate": 2.634827152994923e-06,
+      "loss": 0.301,
+      "step": 12634
+    },
+    {
+      "epoch": 0.9291124347378483,
+      "grad_norm": 0.4308265447616577,
+      "learning_rate": 2.6293988684242266e-06,
+      "loss": 0.3,
+      "step": 12635
+    },
+    {
+      "epoch": 0.9291859695565851,
+      "grad_norm": 0.4358551800251007,
+      "learning_rate": 2.6239761068167724e-06,
+      "loss": 0.3208,
+      "step": 12636
+    },
+    {
+      "epoch": 0.9292595043753217,
+      "grad_norm": 0.3945387005805969,
+      "learning_rate": 2.618558868480181e-06,
+      "loss": 0.3046,
+      "step": 12637
+    },
+    {
+      "epoch": 0.9293330391940584,
+      "grad_norm": 0.3930168151855469,
+      "learning_rate": 2.6131471537216956e-06,
+      "loss": 0.2963,
+      "step": 12638
+    },
+    {
+      "epoch": 0.929406574012795,
+      "grad_norm": 0.3529595136642456,
+      "learning_rate": 2.6077409628482706e-06,
+      "loss": 0.3441,
+      "step": 12639
+    },
+    {
+      "epoch": 0.9294801088315318,
+      "grad_norm": 0.43910402059555054,
+      "learning_rate": 2.602340296166561e-06,
+      "loss": 0.3491,
+      "step": 12640
+    },
+    {
+      "epoch": 0.9295536436502684,
+      "grad_norm": 0.4294656813144684,
+      "learning_rate": 2.5969451539828994e-06,
+      "loss": 0.295,
+      "step": 12641
+    },
+    {
+      "epoch": 0.9296271784690051,
+      "grad_norm": 0.43740537762641907,
+      "learning_rate": 2.591555536603296e-06,
+      "loss": 0.326,
+      "step": 12642
+    },
+    {
+      "epoch": 0.9297007132877417,
+      "grad_norm": 0.4188920855522156,
+      "learning_rate": 2.586171444333474e-06,
+      "loss": 0.3446,
+      "step": 12643
+    },
+    {
+      "epoch": 0.9297742481064785,
+      "grad_norm": 0.4197244644165039,
+      "learning_rate": 2.58079287747881e-06,
+      "loss": 0.3323,
+      "step": 12644
+    },
+    {
+      "epoch": 0.9298477829252151,
+      "grad_norm": 0.39543431997299194,
+      "learning_rate": 2.5754198363443726e-06,
+      "loss": 0.3317,
+      "step": 12645
+    },
+    {
+      "epoch": 0.9299213177439518,
+      "grad_norm": 0.5921205878257751,
+      "learning_rate": 2.5700523212349503e-06,
+      "loss": 0.3129,
+      "step": 12646
+    },
+    {
+      "epoch": 0.9299948525626884,
+      "grad_norm": 0.4030892550945282,
+      "learning_rate": 2.5646903324549887e-06,
+      "loss": 0.3049,
+      "step": 12647
+    },
+    {
+      "epoch": 0.930068387381425,
+      "grad_norm": 0.40432223677635193,
+      "learning_rate": 2.559333870308611e-06,
+      "loss": 0.3266,
+      "step": 12648
+    },
+    {
+      "epoch": 0.9301419222001618,
+      "grad_norm": 0.4155685603618622,
+      "learning_rate": 2.553982935099641e-06,
+      "loss": 0.291,
+      "step": 12649
+    },
+    {
+      "epoch": 0.9302154570188984,
+      "grad_norm": 0.3919181525707245,
+      "learning_rate": 2.548637527131603e-06,
+      "loss": 0.3426,
+      "step": 12650
+    },
+    {
+      "epoch": 0.9302889918376351,
+      "grad_norm": 0.38122454285621643,
+      "learning_rate": 2.5432976467076986e-06,
+      "loss": 0.3284,
+      "step": 12651
+    },
+    {
+      "epoch": 0.9303625266563718,
+      "grad_norm": 0.480056494474411,
+      "learning_rate": 2.5379632941307964e-06,
+      "loss": 0.3006,
+      "step": 12652
+    },
+    {
+      "epoch": 0.9304360614751085,
+      "grad_norm": 0.525952160358429,
+      "learning_rate": 2.532634469703465e-06,
+      "loss": 0.3157,
+      "step": 12653
+    },
+    {
+      "epoch": 0.9305095962938451,
+      "grad_norm": 0.4624314308166504,
+      "learning_rate": 2.5273111737279754e-06,
+      "loss": 0.3148,
+      "step": 12654
+    },
+    {
+      "epoch": 0.9305831311125818,
+      "grad_norm": 0.39725443720817566,
+      "learning_rate": 2.521993406506251e-06,
+      "loss": 0.3284,
+      "step": 12655
+    },
+    {
+      "epoch": 0.9306566659313185,
+      "grad_norm": 0.38574692606925964,
+      "learning_rate": 2.5166811683399404e-06,
+      "loss": 0.3113,
+      "step": 12656
+    },
+    {
+      "epoch": 0.9307302007500552,
+      "grad_norm": 0.4867246747016907,
+      "learning_rate": 2.511374459530369e-06,
+      "loss": 0.3194,
+      "step": 12657
+    },
+    {
+      "epoch": 0.9308037355687918,
+      "grad_norm": 0.5428057909011841,
+      "learning_rate": 2.506073280378496e-06,
+      "loss": 0.3424,
+      "step": 12658
+    },
+    {
+      "epoch": 0.9308772703875285,
+      "grad_norm": 0.4286310076713562,
+      "learning_rate": 2.500777631185047e-06,
+      "loss": 0.3453,
+      "step": 12659
+    },
+    {
+      "epoch": 0.9309508052062652,
+      "grad_norm": 0.5961957573890686,
+      "learning_rate": 2.495487512250394e-06,
+      "loss": 0.3227,
+      "step": 12660
+    },
+    {
+      "epoch": 0.9310243400250018,
+      "grad_norm": 0.39832717180252075,
+      "learning_rate": 2.490202923874585e-06,
+      "loss": 0.3012,
+      "step": 12661
+    },
+    {
+      "epoch": 0.9310978748437385,
+      "grad_norm": 0.40125834941864014,
+      "learning_rate": 2.4849238663573692e-06,
+      "loss": 0.3178,
+      "step": 12662
+    },
+    {
+      "epoch": 0.9311714096624751,
+      "grad_norm": 0.3737126290798187,
+      "learning_rate": 2.4796503399981742e-06,
+      "loss": 0.2841,
+      "step": 12663
+    },
+    {
+      "epoch": 0.9312449444812119,
+      "grad_norm": 0.3661714196205139,
+      "learning_rate": 2.474382345096149e-06,
+      "loss": 0.3033,
+      "step": 12664
+    },
+    {
+      "epoch": 0.9313184792999485,
+      "grad_norm": 0.39344683289527893,
+      "learning_rate": 2.469119881950066e-06,
+      "loss": 0.3248,
+      "step": 12665
+    },
+    {
+      "epoch": 0.9313920141186852,
+      "grad_norm": 0.36999598145484924,
+      "learning_rate": 2.463862950858431e-06,
+      "loss": 0.3259,
+      "step": 12666
+    },
+    {
+      "epoch": 0.9314655489374218,
+      "grad_norm": 0.4255140423774719,
+      "learning_rate": 2.4586115521194274e-06,
+      "loss": 0.3066,
+      "step": 12667
+    },
+    {
+      "epoch": 0.9315390837561586,
+      "grad_norm": 0.3791583776473999,
+      "learning_rate": 2.4533656860309175e-06,
+      "loss": 0.2894,
+      "step": 12668
+    },
+    {
+      "epoch": 0.9316126185748952,
+      "grad_norm": 0.3744094669818878,
+      "learning_rate": 2.4481253528904514e-06,
+      "loss": 0.3378,
+      "step": 12669
+    },
+    {
+      "epoch": 0.9316861533936319,
+      "grad_norm": 0.370572566986084,
+      "learning_rate": 2.4428905529952696e-06,
+      "loss": 0.2952,
+      "step": 12670
+    },
+    {
+      "epoch": 0.9317596882123685,
+      "grad_norm": 0.4141451120376587,
+      "learning_rate": 2.4376612866422788e-06,
+      "loss": 0.297,
+      "step": 12671
+    },
+    {
+      "epoch": 0.9318332230311053,
+      "grad_norm": 0.4008753299713135,
+      "learning_rate": 2.4324375541281195e-06,
+      "loss": 0.2977,
+      "step": 12672
+    },
+    {
+      "epoch": 0.9319067578498419,
+      "grad_norm": 0.42221757769584656,
+      "learning_rate": 2.427219355749055e-06,
+      "loss": 0.3133,
+      "step": 12673
+    },
+    {
+      "epoch": 0.9319802926685786,
+      "grad_norm": 0.35533300042152405,
+      "learning_rate": 2.422006691801082e-06,
+      "loss": 0.3106,
+      "step": 12674
+    },
+    {
+      "epoch": 0.9320538274873152,
+      "grad_norm": 0.4142398238182068,
+      "learning_rate": 2.4167995625798743e-06,
+      "loss": 0.3357,
+      "step": 12675
+    },
+    {
+      "epoch": 0.9321273623060519,
+      "grad_norm": 0.4538228213787079,
+      "learning_rate": 2.4115979683807745e-06,
+      "loss": 0.327,
+      "step": 12676
+    },
+    {
+      "epoch": 0.9322008971247886,
+      "grad_norm": 0.40959346294403076,
+      "learning_rate": 2.406401909498834e-06,
+      "loss": 0.29,
+      "step": 12677
+    },
+    {
+      "epoch": 0.9322744319435252,
+      "grad_norm": 0.4675796627998352,
+      "learning_rate": 2.4012113862287633e-06,
+      "loss": 0.3099,
+      "step": 12678
+    },
+    {
+      "epoch": 0.932347966762262,
+      "grad_norm": 0.43242987990379333,
+      "learning_rate": 2.396026398864981e-06,
+      "loss": 0.3488,
+      "step": 12679
+    },
+    {
+      "epoch": 0.9324215015809986,
+      "grad_norm": 0.43072831630706787,
+      "learning_rate": 2.390846947701597e-06,
+      "loss": 0.2895,
+      "step": 12680
+    },
+    {
+      "epoch": 0.9324950363997353,
+      "grad_norm": 0.3497699201107025,
+      "learning_rate": 2.3856730330323763e-06,
+      "loss": 0.2996,
+      "step": 12681
+    },
+    {
+      "epoch": 0.9325685712184719,
+      "grad_norm": 0.42607128620147705,
+      "learning_rate": 2.3805046551507947e-06,
+      "loss": 0.3132,
+      "step": 12682
+    },
+    {
+      "epoch": 0.9326421060372087,
+      "grad_norm": 0.4205682873725891,
+      "learning_rate": 2.3753418143500184e-06,
+      "loss": 0.3436,
+      "step": 12683
+    },
+    {
+      "epoch": 0.9327156408559453,
+      "grad_norm": 0.468954861164093,
+      "learning_rate": 2.3701845109228904e-06,
+      "loss": 0.3089,
+      "step": 12684
+    },
+    {
+      "epoch": 0.932789175674682,
+      "grad_norm": 0.41086405515670776,
+      "learning_rate": 2.3650327451619213e-06,
+      "loss": 0.3279,
+      "step": 12685
+    },
+    {
+      "epoch": 0.9328627104934186,
+      "grad_norm": 0.43541219830513,
+      "learning_rate": 2.359886517359322e-06,
+      "loss": 0.3186,
+      "step": 12686
+    },
+    {
+      "epoch": 0.9329362453121554,
+      "grad_norm": 0.38499715924263,
+      "learning_rate": 2.3547458278070255e-06,
+      "loss": 0.3049,
+      "step": 12687
+    },
+    {
+      "epoch": 0.933009780130892,
+      "grad_norm": 0.39422476291656494,
+      "learning_rate": 2.3496106767965766e-06,
+      "loss": 0.3014,
+      "step": 12688
+    },
+    {
+      "epoch": 0.9330833149496286,
+      "grad_norm": 0.38022294640541077,
+      "learning_rate": 2.3444810646192638e-06,
+      "loss": 0.3396,
+      "step": 12689
+    },
+    {
+      "epoch": 0.9331568497683653,
+      "grad_norm": 0.40269675850868225,
+      "learning_rate": 2.339356991566055e-06,
+      "loss": 0.3073,
+      "step": 12690
+    },
+    {
+      "epoch": 0.933230384587102,
+      "grad_norm": 0.49384069442749023,
+      "learning_rate": 2.334238457927573e-06,
+      "loss": 0.3103,
+      "step": 12691
+    },
+    {
+      "epoch": 0.9333039194058387,
+      "grad_norm": 0.4364045262336731,
+      "learning_rate": 2.329125463994164e-06,
+      "loss": 0.3076,
+      "step": 12692
+    },
+    {
+      "epoch": 0.9333774542245753,
+      "grad_norm": 0.45160093903541565,
+      "learning_rate": 2.3240180100558394e-06,
+      "loss": 0.3253,
+      "step": 12693
+    },
+    {
+      "epoch": 0.933450989043312,
+      "grad_norm": 0.37220802903175354,
+      "learning_rate": 2.3189160964022906e-06,
+      "loss": 0.3142,
+      "step": 12694
+    },
+    {
+      "epoch": 0.9335245238620486,
+      "grad_norm": 0.47034505009651184,
+      "learning_rate": 2.3138197233229074e-06,
+      "loss": 0.3101,
+      "step": 12695
+    },
+    {
+      "epoch": 0.9335980586807854,
+      "grad_norm": 0.45476022362709045,
+      "learning_rate": 2.3087288911067595e-06,
+      "loss": 0.331,
+      "step": 12696
+    },
+    {
+      "epoch": 0.933671593499522,
+      "grad_norm": 0.46938061714172363,
+      "learning_rate": 2.3036436000426155e-06,
+      "loss": 0.3221,
+      "step": 12697
+    },
+    {
+      "epoch": 0.9337451283182587,
+      "grad_norm": 0.42547741532325745,
+      "learning_rate": 2.2985638504189e-06,
+      "loss": 0.3358,
+      "step": 12698
+    },
+    {
+      "epoch": 0.9338186631369954,
+      "grad_norm": 0.5363611578941345,
+      "learning_rate": 2.2934896425237493e-06,
+      "loss": 0.3108,
+      "step": 12699
+    },
+    {
+      "epoch": 0.9338921979557321,
+      "grad_norm": 0.47727200388908386,
+      "learning_rate": 2.288420976644989e-06,
+      "loss": 0.3021,
+      "step": 12700
+    },
+    {
+      "epoch": 0.9339657327744687,
+      "grad_norm": 0.4377639889717102,
+      "learning_rate": 2.2833578530701004e-06,
+      "loss": 0.3344,
+      "step": 12701
+    },
+    {
+      "epoch": 0.9340392675932054,
+      "grad_norm": 0.41499462723731995,
+      "learning_rate": 2.278300272086276e-06,
+      "loss": 0.3101,
+      "step": 12702
+    },
+    {
+      "epoch": 0.9341128024119421,
+      "grad_norm": 0.4532848000526428,
+      "learning_rate": 2.2732482339804072e-06,
+      "loss": 0.3042,
+      "step": 12703
+    },
+    {
+      "epoch": 0.9341863372306787,
+      "grad_norm": 0.4306369125843048,
+      "learning_rate": 2.268201739039022e-06,
+      "loss": 0.3078,
+      "step": 12704
+    },
+    {
+      "epoch": 0.9342598720494154,
+      "grad_norm": 0.7333908677101135,
+      "learning_rate": 2.2631607875483686e-06,
+      "loss": 0.3215,
+      "step": 12705
+    },
+    {
+      "epoch": 0.934333406868152,
+      "grad_norm": 0.3949889838695526,
+      "learning_rate": 2.2581253797943957e-06,
+      "loss": 0.3052,
+      "step": 12706
+    },
+    {
+      "epoch": 0.9344069416868888,
+      "grad_norm": 0.41159898042678833,
+      "learning_rate": 2.2530955160626977e-06,
+      "loss": 0.3097,
+      "step": 12707
+    },
+    {
+      "epoch": 0.9344804765056254,
+      "grad_norm": 0.4328388273715973,
+      "learning_rate": 2.2480711966385792e-06,
+      "loss": 0.3094,
+      "step": 12708
+    },
+    {
+      "epoch": 0.9345540113243621,
+      "grad_norm": 0.37786707282066345,
+      "learning_rate": 2.2430524218070126e-06,
+      "loss": 0.3111,
+      "step": 12709
+    },
+    {
+      "epoch": 0.9346275461430987,
+      "grad_norm": 0.36371809244155884,
+      "learning_rate": 2.2380391918527034e-06,
+      "loss": 0.3158,
+      "step": 12710
+    },
+    {
+      "epoch": 0.9347010809618355,
+      "grad_norm": 0.4139184057712555,
+      "learning_rate": 2.2330315070599573e-06,
+      "loss": 0.3527,
+      "step": 12711
+    },
+    {
+      "epoch": 0.9347746157805721,
+      "grad_norm": 0.3842453360557556,
+      "learning_rate": 2.2280293677128473e-06,
+      "loss": 0.3206,
+      "step": 12712
+    },
+    {
+      "epoch": 0.9348481505993088,
+      "grad_norm": 0.47639384865760803,
+      "learning_rate": 2.2230327740951127e-06,
+      "loss": 0.3178,
+      "step": 12713
+    },
+    {
+      "epoch": 0.9349216854180454,
+      "grad_norm": 0.3866749703884125,
+      "learning_rate": 2.2180417264901277e-06,
+      "loss": 0.3204,
+      "step": 12714
+    },
+    {
+      "epoch": 0.9349952202367822,
+      "grad_norm": 0.5215319395065308,
+      "learning_rate": 2.213056225181009e-06,
+      "loss": 0.3305,
+      "step": 12715
+    },
+    {
+      "epoch": 0.9350687550555188,
+      "grad_norm": 0.41819825768470764,
+      "learning_rate": 2.2080762704505544e-06,
+      "loss": 0.3523,
+      "step": 12716
+    },
+    {
+      "epoch": 0.9351422898742554,
+      "grad_norm": 0.3988860845565796,
+      "learning_rate": 2.203101862581214e-06,
+      "loss": 0.2974,
+      "step": 12717
+    },
+    {
+      "epoch": 0.9352158246929921,
+      "grad_norm": 0.36510366201400757,
+      "learning_rate": 2.1981330018551407e-06,
+      "loss": 0.3344,
+      "step": 12718
+    },
+    {
+      "epoch": 0.9352893595117288,
+      "grad_norm": 0.4062545597553253,
+      "learning_rate": 2.1931696885541753e-06,
+      "loss": 0.3496,
+      "step": 12719
+    },
+    {
+      "epoch": 0.9353628943304655,
+      "grad_norm": 0.42962646484375,
+      "learning_rate": 2.1882119229598598e-06,
+      "loss": 0.3216,
+      "step": 12720
+    },
+    {
+      "epoch": 0.9354364291492021,
+      "grad_norm": 0.3963826596736908,
+      "learning_rate": 2.1832597053533688e-06,
+      "loss": 0.3365,
+      "step": 12721
+    },
+    {
+      "epoch": 0.9355099639679388,
+      "grad_norm": 0.3872268497943878,
+      "learning_rate": 2.1783130360156225e-06,
+      "loss": 0.3083,
+      "step": 12722
+    },
+    {
+      "epoch": 0.9355834987866755,
+      "grad_norm": 0.40570685267448425,
+      "learning_rate": 2.173371915227207e-06,
+      "loss": 0.336,
+      "step": 12723
+    },
+    {
+      "epoch": 0.9356570336054122,
+      "grad_norm": 0.404642790555954,
+      "learning_rate": 2.168436343268365e-06,
+      "loss": 0.3241,
+      "step": 12724
+    },
+    {
+      "epoch": 0.9357305684241488,
+      "grad_norm": 0.40955179929733276,
+      "learning_rate": 2.163506320419062e-06,
+      "loss": 0.3208,
+      "step": 12725
+    },
+    {
+      "epoch": 0.9358041032428855,
+      "grad_norm": 0.3807671368122101,
+      "learning_rate": 2.1585818469589404e-06,
+      "loss": 0.3003,
+      "step": 12726
+    },
+    {
+      "epoch": 0.9358776380616222,
+      "grad_norm": 0.3643546402454376,
+      "learning_rate": 2.153662923167299e-06,
+      "loss": 0.3075,
+      "step": 12727
+    },
+    {
+      "epoch": 0.9359511728803589,
+      "grad_norm": 0.4754871726036072,
+      "learning_rate": 2.14874954932317e-06,
+      "loss": 0.3252,
+      "step": 12728
+    },
+    {
+      "epoch": 0.9360247076990955,
+      "grad_norm": 0.4323866665363312,
+      "learning_rate": 2.1438417257052313e-06,
+      "loss": 0.3234,
+      "step": 12729
+    },
+    {
+      "epoch": 0.9360982425178322,
+      "grad_norm": 0.4110676944255829,
+      "learning_rate": 2.1389394525918595e-06,
+      "loss": 0.312,
+      "step": 12730
+    },
+    {
+      "epoch": 0.9361717773365689,
+      "grad_norm": 0.39710694551467896,
+      "learning_rate": 2.134042730261121e-06,
+      "loss": 0.3706,
+      "step": 12731
+    },
+    {
+      "epoch": 0.9362453121553055,
+      "grad_norm": 0.415497362613678,
+      "learning_rate": 2.1291515589907496e-06,
+      "loss": 0.317,
+      "step": 12732
+    },
+    {
+      "epoch": 0.9363188469740422,
+      "grad_norm": 0.376547634601593,
+      "learning_rate": 2.124265939058201e-06,
+      "loss": 0.312,
+      "step": 12733
+    },
+    {
+      "epoch": 0.9363923817927788,
+      "grad_norm": 0.40726184844970703,
+      "learning_rate": 2.1193858707405755e-06,
+      "loss": 0.2975,
+      "step": 12734
+    },
+    {
+      "epoch": 0.9364659166115156,
+      "grad_norm": 0.3920994699001312,
+      "learning_rate": 2.1145113543146746e-06,
+      "loss": 0.3142,
+      "step": 12735
+    },
+    {
+      "epoch": 0.9365394514302522,
+      "grad_norm": 0.38863492012023926,
+      "learning_rate": 2.1096423900570095e-06,
+      "loss": 0.3372,
+      "step": 12736
+    },
+    {
+      "epoch": 0.9366129862489889,
+      "grad_norm": 0.39700421690940857,
+      "learning_rate": 2.1047789782437265e-06,
+      "loss": 0.3153,
+      "step": 12737
+    },
+    {
+      "epoch": 0.9366865210677255,
+      "grad_norm": 0.48076707124710083,
+      "learning_rate": 2.099921119150683e-06,
+      "loss": 0.3212,
+      "step": 12738
+    },
+    {
+      "epoch": 0.9367600558864623,
+      "grad_norm": 0.4469163119792938,
+      "learning_rate": 2.095068813053447e-06,
+      "loss": 0.3135,
+      "step": 12739
+    },
+    {
+      "epoch": 0.9368335907051989,
+      "grad_norm": 0.4279829263687134,
+      "learning_rate": 2.0902220602272203e-06,
+      "loss": 0.3237,
+      "step": 12740
+    },
+    {
+      "epoch": 0.9369071255239356,
+      "grad_norm": 0.3890632390975952,
+      "learning_rate": 2.085380860946928e-06,
+      "loss": 0.3385,
+      "step": 12741
+    },
+    {
+      "epoch": 0.9369806603426722,
+      "grad_norm": 0.3993648886680603,
+      "learning_rate": 2.080545215487173e-06,
+      "loss": 0.3398,
+      "step": 12742
+    },
+    {
+      "epoch": 0.937054195161409,
+      "grad_norm": 0.4384815990924835,
+      "learning_rate": 2.075715124122235e-06,
+      "loss": 0.3623,
+      "step": 12743
+    },
+    {
+      "epoch": 0.9371277299801456,
+      "grad_norm": 0.4413893222808838,
+      "learning_rate": 2.070890587126073e-06,
+      "loss": 0.3275,
+      "step": 12744
+    },
+    {
+      "epoch": 0.9372012647988822,
+      "grad_norm": 0.3791849613189697,
+      "learning_rate": 2.066071604772346e-06,
+      "loss": 0.3153,
+      "step": 12745
+    },
+    {
+      "epoch": 0.937274799617619,
+      "grad_norm": 0.4130481779575348,
+      "learning_rate": 2.0612581773343907e-06,
+      "loss": 0.3045,
+      "step": 12746
+    },
+    {
+      "epoch": 0.9373483344363556,
+      "grad_norm": 0.3984481692314148,
+      "learning_rate": 2.0564503050852337e-06,
+      "loss": 0.3018,
+      "step": 12747
+    },
+    {
+      "epoch": 0.9374218692550923,
+      "grad_norm": 0.38589218258857727,
+      "learning_rate": 2.051647988297567e-06,
+      "loss": 0.323,
+      "step": 12748
+    },
+    {
+      "epoch": 0.9374954040738289,
+      "grad_norm": 0.40712812542915344,
+      "learning_rate": 2.0468512272438066e-06,
+      "loss": 0.3283,
+      "step": 12749
+    },
+    {
+      "epoch": 0.9375689388925657,
+      "grad_norm": 0.42833101749420166,
+      "learning_rate": 2.042060022196013e-06,
+      "loss": 0.2916,
+      "step": 12750
+    },
+    {
+      "epoch": 0.9376424737113023,
+      "grad_norm": 0.3651220202445984,
+      "learning_rate": 2.037274373425957e-06,
+      "loss": 0.2964,
+      "step": 12751
+    },
+    {
+      "epoch": 0.937716008530039,
+      "grad_norm": 0.4552348852157593,
+      "learning_rate": 2.0324942812050886e-06,
+      "loss": 0.3059,
+      "step": 12752
+    },
+    {
+      "epoch": 0.9377895433487756,
+      "grad_norm": 0.5083942413330078,
+      "learning_rate": 2.027719745804535e-06,
+      "loss": 0.313,
+      "step": 12753
+    },
+    {
+      "epoch": 0.9378630781675124,
+      "grad_norm": 0.431449830532074,
+      "learning_rate": 2.022950767495102e-06,
+      "loss": 0.3114,
+      "step": 12754
+    },
+    {
+      "epoch": 0.937936612986249,
+      "grad_norm": 0.41020116209983826,
+      "learning_rate": 2.0181873465473066e-06,
+      "loss": 0.3477,
+      "step": 12755
+    },
+    {
+      "epoch": 0.9380101478049857,
+      "grad_norm": 0.530211329460144,
+      "learning_rate": 2.0134294832313327e-06,
+      "loss": 0.3329,
+      "step": 12756
+    },
+    {
+      "epoch": 0.9380836826237223,
+      "grad_norm": 0.39091718196868896,
+      "learning_rate": 2.008677177817031e-06,
+      "loss": 0.3037,
+      "step": 12757
+    },
+    {
+      "epoch": 0.9381572174424591,
+      "grad_norm": 0.43393832445144653,
+      "learning_rate": 2.003930430573986e-06,
+      "loss": 0.3551,
+      "step": 12758
+    },
+    {
+      "epoch": 0.9382307522611957,
+      "grad_norm": 0.3970661759376526,
+      "learning_rate": 1.9991892417714265e-06,
+      "loss": 0.3052,
+      "step": 12759
+    },
+    {
+      "epoch": 0.9383042870799323,
+      "grad_norm": 0.40515807271003723,
+      "learning_rate": 1.9944536116782706e-06,
+      "loss": 0.3166,
+      "step": 12760
+    },
+    {
+      "epoch": 0.938377821898669,
+      "grad_norm": 0.36905577778816223,
+      "learning_rate": 1.9897235405631376e-06,
+      "loss": 0.3031,
+      "step": 12761
+    },
+    {
+      "epoch": 0.9384513567174056,
+      "grad_norm": 0.42813384532928467,
+      "learning_rate": 1.9849990286943344e-06,
+      "loss": 0.2983,
+      "step": 12762
+    },
+    {
+      "epoch": 0.9385248915361424,
+      "grad_norm": 0.38111090660095215,
+      "learning_rate": 1.980280076339813e-06,
+      "loss": 0.3113,
+      "step": 12763
+    },
+    {
+      "epoch": 0.938598426354879,
+      "grad_norm": 0.3734404742717743,
+      "learning_rate": 1.9755666837672494e-06,
+      "loss": 0.3091,
+      "step": 12764
+    },
+    {
+      "epoch": 0.9386719611736157,
+      "grad_norm": 0.43311843276023865,
+      "learning_rate": 1.970858851243995e-06,
+      "loss": 0.2767,
+      "step": 12765
+    },
+    {
+      "epoch": 0.9387454959923524,
+      "grad_norm": 0.3793749213218689,
+      "learning_rate": 1.966156579037082e-06,
+      "loss": 0.2673,
+      "step": 12766
+    },
+    {
+      "epoch": 0.9388190308110891,
+      "grad_norm": 0.47921279072761536,
+      "learning_rate": 1.961459867413229e-06,
+      "loss": 0.3085,
+      "step": 12767
+    },
+    {
+      "epoch": 0.9388925656298257,
+      "grad_norm": 0.37172314524650574,
+      "learning_rate": 1.956768716638824e-06,
+      "loss": 0.3259,
+      "step": 12768
+    },
+    {
+      "epoch": 0.9389661004485624,
+      "grad_norm": 0.370879590511322,
+      "learning_rate": 1.9520831269799865e-06,
+      "loss": 0.3412,
+      "step": 12769
+    },
+    {
+      "epoch": 0.939039635267299,
+      "grad_norm": 0.36972135305404663,
+      "learning_rate": 1.9474030987024493e-06,
+      "loss": 0.3155,
+      "step": 12770
+    },
+    {
+      "epoch": 0.9391131700860358,
+      "grad_norm": 0.4086836576461792,
+      "learning_rate": 1.9427286320716888e-06,
+      "loss": 0.3234,
+      "step": 12771
+    },
+    {
+      "epoch": 0.9391867049047724,
+      "grad_norm": 0.40255117416381836,
+      "learning_rate": 1.938059727352859e-06,
+      "loss": 0.3308,
+      "step": 12772
+    },
+    {
+      "epoch": 0.939260239723509,
+      "grad_norm": 0.3817150592803955,
+      "learning_rate": 1.9333963848107596e-06,
+      "loss": 0.3034,
+      "step": 12773
+    },
+    {
+      "epoch": 0.9393337745422458,
+      "grad_norm": 0.3833003342151642,
+      "learning_rate": 1.928738604709912e-06,
+      "loss": 0.3363,
+      "step": 12774
+    },
+    {
+      "epoch": 0.9394073093609824,
+      "grad_norm": 0.4290448725223541,
+      "learning_rate": 1.9240863873145167e-06,
+      "loss": 0.3312,
+      "step": 12775
+    },
+    {
+      "epoch": 0.9394808441797191,
+      "grad_norm": 0.42154693603515625,
+      "learning_rate": 1.9194397328884396e-06,
+      "loss": 0.3239,
+      "step": 12776
+    },
+    {
+      "epoch": 0.9395543789984557,
+      "grad_norm": 0.3923797607421875,
+      "learning_rate": 1.914798641695248e-06,
+      "loss": 0.3093,
+      "step": 12777
+    },
+    {
+      "epoch": 0.9396279138171925,
+      "grad_norm": 0.38550642132759094,
+      "learning_rate": 1.9101631139981867e-06,
+      "loss": 0.3182,
+      "step": 12778
+    },
+    {
+      "epoch": 0.9397014486359291,
+      "grad_norm": 0.42698895931243896,
+      "learning_rate": 1.9055331500602014e-06,
+      "loss": 0.3199,
+      "step": 12779
+    },
+    {
+      "epoch": 0.9397749834546658,
+      "grad_norm": 0.44420483708381653,
+      "learning_rate": 1.9009087501438928e-06,
+      "loss": 0.3154,
+      "step": 12780
+    },
+    {
+      "epoch": 0.9398485182734024,
+      "grad_norm": 0.36959052085876465,
+      "learning_rate": 1.8962899145115732e-06,
+      "loss": 0.2962,
+      "step": 12781
+    },
+    {
+      "epoch": 0.9399220530921392,
+      "grad_norm": 0.40331682562828064,
+      "learning_rate": 1.891676643425222e-06,
+      "loss": 0.3095,
+      "step": 12782
+    },
+    {
+      "epoch": 0.9399955879108758,
+      "grad_norm": 0.4151085615158081,
+      "learning_rate": 1.8870689371465079e-06,
+      "loss": 0.3241,
+      "step": 12783
+    },
+    {
+      "epoch": 0.9400691227296125,
+      "grad_norm": 0.3800581395626068,
+      "learning_rate": 1.882466795936777e-06,
+      "loss": 0.3043,
+      "step": 12784
+    },
+    {
+      "epoch": 0.9401426575483491,
+      "grad_norm": 0.41361504793167114,
+      "learning_rate": 1.8778702200570875e-06,
+      "loss": 0.2943,
+      "step": 12785
+    },
+    {
+      "epoch": 0.9402161923670859,
+      "grad_norm": 0.4519699215888977,
+      "learning_rate": 1.8732792097681528e-06,
+      "loss": 0.2823,
+      "step": 12786
+    },
+    {
+      "epoch": 0.9402897271858225,
+      "grad_norm": 0.3630177080631256,
+      "learning_rate": 1.8686937653303648e-06,
+      "loss": 0.2845,
+      "step": 12787
+    },
+    {
+      "epoch": 0.9403632620045591,
+      "grad_norm": 0.40905553102493286,
+      "learning_rate": 1.8641138870038489e-06,
+      "loss": 0.3335,
+      "step": 12788
+    },
+    {
+      "epoch": 0.9404367968232958,
+      "grad_norm": 0.4006040692329407,
+      "learning_rate": 1.8595395750483525e-06,
+      "loss": 0.3053,
+      "step": 12789
+    },
+    {
+      "epoch": 0.9405103316420325,
+      "grad_norm": 0.3924902677536011,
+      "learning_rate": 1.8549708297233348e-06,
+      "loss": 0.3419,
+      "step": 12790
+    },
+    {
+      "epoch": 0.9405838664607692,
+      "grad_norm": 0.4172593057155609,
+      "learning_rate": 1.8504076512879554e-06,
+      "loss": 0.317,
+      "step": 12791
+    },
+    {
+      "epoch": 0.9406574012795058,
+      "grad_norm": 0.45915961265563965,
+      "learning_rate": 1.8458500400010404e-06,
+      "loss": 0.3286,
+      "step": 12792
+    },
+    {
+      "epoch": 0.9407309360982425,
+      "grad_norm": 0.4251067638397217,
+      "learning_rate": 1.841297996121083e-06,
+      "loss": 0.3229,
+      "step": 12793
+    },
+    {
+      "epoch": 0.9408044709169792,
+      "grad_norm": 0.43154871463775635,
+      "learning_rate": 1.836751519906299e-06,
+      "loss": 0.3258,
+      "step": 12794
+    },
+    {
+      "epoch": 0.9408780057357159,
+      "grad_norm": 0.39525800943374634,
+      "learning_rate": 1.832210611614582e-06,
+      "loss": 0.2943,
+      "step": 12795
+    },
+    {
+      "epoch": 0.9409515405544525,
+      "grad_norm": 0.43302708864212036,
+      "learning_rate": 1.8276752715034595e-06,
+      "loss": 0.3119,
+      "step": 12796
+    },
+    {
+      "epoch": 0.9410250753731892,
+      "grad_norm": 0.3672832250595093,
+      "learning_rate": 1.8231454998302034e-06,
+      "loss": 0.3477,
+      "step": 12797
+    },
+    {
+      "epoch": 0.9410986101919259,
+      "grad_norm": 0.4604876935482025,
+      "learning_rate": 1.8186212968517524e-06,
+      "loss": 0.3444,
+      "step": 12798
+    },
+    {
+      "epoch": 0.9411721450106626,
+      "grad_norm": 0.33965685963630676,
+      "learning_rate": 1.8141026628247126e-06,
+      "loss": 0.3081,
+      "step": 12799
+    },
+    {
+      "epoch": 0.9412456798293992,
+      "grad_norm": 0.4083092510700226,
+      "learning_rate": 1.80958959800539e-06,
+      "loss": 0.3278,
+      "step": 12800
+    },
+    {
+      "epoch": 0.9413192146481358,
+      "grad_norm": 0.37185537815093994,
+      "learning_rate": 1.8050821026497689e-06,
+      "loss": 0.3065,
+      "step": 12801
+    },
+    {
+      "epoch": 0.9413927494668726,
+      "grad_norm": 0.3713577687740326,
+      "learning_rate": 1.8005801770135222e-06,
+      "loss": 0.3007,
+      "step": 12802
+    },
+    {
+      "epoch": 0.9414662842856092,
+      "grad_norm": 0.42081591486930847,
+      "learning_rate": 1.7960838213520014e-06,
+      "loss": 0.3317,
+      "step": 12803
+    },
+    {
+      "epoch": 0.9415398191043459,
+      "grad_norm": 0.4024532735347748,
+      "learning_rate": 1.791593035920236e-06,
+      "loss": 0.2925,
+      "step": 12804
+    },
+    {
+      "epoch": 0.9416133539230825,
+      "grad_norm": 0.3721790015697479,
+      "learning_rate": 1.7871078209729774e-06,
+      "loss": 0.3396,
+      "step": 12805
+    },
+    {
+      "epoch": 0.9416868887418193,
+      "grad_norm": 0.48170459270477295,
+      "learning_rate": 1.7826281767645891e-06,
+      "loss": 0.3059,
+      "step": 12806
+    },
+    {
+      "epoch": 0.9417604235605559,
+      "grad_norm": 0.3827381134033203,
+      "learning_rate": 1.77815410354919e-06,
+      "loss": 0.3103,
+      "step": 12807
+    },
+    {
+      "epoch": 0.9418339583792926,
+      "grad_norm": 0.4013620615005493,
+      "learning_rate": 1.7736856015805547e-06,
+      "loss": 0.3082,
+      "step": 12808
+    },
+    {
+      "epoch": 0.9419074931980292,
+      "grad_norm": 0.4015626311302185,
+      "learning_rate": 1.769222671112125e-06,
+      "loss": 0.3611,
+      "step": 12809
+    },
+    {
+      "epoch": 0.941981028016766,
+      "grad_norm": 0.37216198444366455,
+      "learning_rate": 1.7647653123970653e-06,
+      "loss": 0.3295,
+      "step": 12810
+    },
+    {
+      "epoch": 0.9420545628355026,
+      "grad_norm": 0.3586808145046234,
+      "learning_rate": 1.7603135256881841e-06,
+      "loss": 0.3149,
+      "step": 12811
+    },
+    {
+      "epoch": 0.9421280976542393,
+      "grad_norm": 0.3850356638431549,
+      "learning_rate": 1.755867311238002e-06,
+      "loss": 0.3293,
+      "step": 12812
+    },
+    {
+      "epoch": 0.942201632472976,
+      "grad_norm": 0.3935440480709076,
+      "learning_rate": 1.7514266692986947e-06,
+      "loss": 0.3287,
+      "step": 12813
+    },
+    {
+      "epoch": 0.9422751672917127,
+      "grad_norm": 0.3444681763648987,
+      "learning_rate": 1.7469916001221498e-06,
+      "loss": 0.3127,
+      "step": 12814
+    },
+    {
+      "epoch": 0.9423487021104493,
+      "grad_norm": 0.40328601002693176,
+      "learning_rate": 1.7425621039599438e-06,
+      "loss": 0.3329,
+      "step": 12815
+    },
+    {
+      "epoch": 0.9424222369291859,
+      "grad_norm": 0.38368502259254456,
+      "learning_rate": 1.738138181063298e-06,
+      "loss": 0.3564,
+      "step": 12816
+    },
+    {
+      "epoch": 0.9424957717479227,
+      "grad_norm": 0.43345651030540466,
+      "learning_rate": 1.733719831683167e-06,
+      "loss": 0.3279,
+      "step": 12817
+    },
+    {
+      "epoch": 0.9425693065666593,
+      "grad_norm": 0.38951122760772705,
+      "learning_rate": 1.7293070560701507e-06,
+      "loss": 0.3292,
+      "step": 12818
+    },
+    {
+      "epoch": 0.942642841385396,
+      "grad_norm": 0.3704914450645447,
+      "learning_rate": 1.7248998544745376e-06,
+      "loss": 0.327,
+      "step": 12819
+    },
+    {
+      "epoch": 0.9427163762041326,
+      "grad_norm": 0.4196280837059021,
+      "learning_rate": 1.7204982271463166e-06,
+      "loss": 0.3308,
+      "step": 12820
+    },
+    {
+      "epoch": 0.9427899110228694,
+      "grad_norm": 0.3756386637687683,
+      "learning_rate": 1.7161021743351657e-06,
+      "loss": 0.329,
+      "step": 12821
+    },
+    {
+      "epoch": 0.942863445841606,
+      "grad_norm": 0.42642146348953247,
+      "learning_rate": 1.711711696290419e-06,
+      "loss": 0.3425,
+      "step": 12822
+    },
+    {
+      "epoch": 0.9429369806603427,
+      "grad_norm": 0.4008386731147766,
+      "learning_rate": 1.7073267932611102e-06,
+      "loss": 0.3463,
+      "step": 12823
+    },
+    {
+      "epoch": 0.9430105154790793,
+      "grad_norm": 0.41766008734703064,
+      "learning_rate": 1.7029474654959632e-06,
+      "loss": 0.3166,
+      "step": 12824
+    },
+    {
+      "epoch": 0.9430840502978161,
+      "grad_norm": 0.45997491478919983,
+      "learning_rate": 1.6985737132433788e-06,
+      "loss": 0.3163,
+      "step": 12825
+    },
+    {
+      "epoch": 0.9431575851165527,
+      "grad_norm": 0.4391177296638489,
+      "learning_rate": 1.6942055367514254e-06,
+      "loss": 0.2994,
+      "step": 12826
+    },
+    {
+      "epoch": 0.9432311199352894,
+      "grad_norm": 0.3864947259426117,
+      "learning_rate": 1.6898429362678714e-06,
+      "loss": 0.3094,
+      "step": 12827
+    },
+    {
+      "epoch": 0.943304654754026,
+      "grad_norm": 0.468308687210083,
+      "learning_rate": 1.6854859120401967e-06,
+      "loss": 0.305,
+      "step": 12828
+    },
+    {
+      "epoch": 0.9433781895727626,
+      "grad_norm": 0.42177459597587585,
+      "learning_rate": 1.6811344643155036e-06,
+      "loss": 0.3438,
+      "step": 12829
+    },
+    {
+      "epoch": 0.9434517243914994,
+      "grad_norm": 0.5491878986358643,
+      "learning_rate": 1.676788593340628e-06,
+      "loss": 0.338,
+      "step": 12830
+    },
+    {
+      "epoch": 0.943525259210236,
+      "grad_norm": 0.3820493817329407,
+      "learning_rate": 1.6724482993620728e-06,
+      "loss": 0.297,
+      "step": 12831
+    },
+    {
+      "epoch": 0.9435987940289727,
+      "grad_norm": 0.3946339786052704,
+      "learning_rate": 1.6681135826260074e-06,
+      "loss": 0.3091,
+      "step": 12832
+    },
+    {
+      "epoch": 0.9436723288477094,
+      "grad_norm": 0.41127029061317444,
+      "learning_rate": 1.6637844433783134e-06,
+      "loss": 0.3199,
+      "step": 12833
+    },
+    {
+      "epoch": 0.9437458636664461,
+      "grad_norm": 0.391390323638916,
+      "learning_rate": 1.6594608818645606e-06,
+      "loss": 0.3111,
+      "step": 12834
+    },
+    {
+      "epoch": 0.9438193984851827,
+      "grad_norm": 0.3985998332500458,
+      "learning_rate": 1.6551428983299532e-06,
+      "loss": 0.3218,
+      "step": 12835
+    },
+    {
+      "epoch": 0.9438929333039194,
+      "grad_norm": 0.42911070585250854,
+      "learning_rate": 1.6508304930194395e-06,
+      "loss": 0.3295,
+      "step": 12836
+    },
+    {
+      "epoch": 0.943966468122656,
+      "grad_norm": 0.5014127492904663,
+      "learning_rate": 1.6465236661776018e-06,
+      "loss": 0.3213,
+      "step": 12837
+    },
+    {
+      "epoch": 0.9440400029413928,
+      "grad_norm": 0.4594709277153015,
+      "learning_rate": 1.6422224180487443e-06,
+      "loss": 0.3288,
+      "step": 12838
+    },
+    {
+      "epoch": 0.9441135377601294,
+      "grad_norm": 0.40935561060905457,
+      "learning_rate": 1.637926748876828e-06,
+      "loss": 0.3206,
+      "step": 12839
+    },
+    {
+      "epoch": 0.9441870725788661,
+      "grad_norm": 0.7014015913009644,
+      "learning_rate": 1.633636658905513e-06,
+      "loss": 0.3248,
+      "step": 12840
+    },
+    {
+      "epoch": 0.9442606073976028,
+      "grad_norm": 0.3882495164871216,
+      "learning_rate": 1.6293521483781493e-06,
+      "loss": 0.3255,
+      "step": 12841
+    },
+    {
+      "epoch": 0.9443341422163395,
+      "grad_norm": 0.3733471930027008,
+      "learning_rate": 1.6250732175377315e-06,
+      "loss": 0.3388,
+      "step": 12842
+    },
+    {
+      "epoch": 0.9444076770350761,
+      "grad_norm": 0.4350942075252533,
+      "learning_rate": 1.6207998666269763e-06,
+      "loss": 0.3397,
+      "step": 12843
+    },
+    {
+      "epoch": 0.9444812118538127,
+      "grad_norm": 0.44014081358909607,
+      "learning_rate": 1.6165320958882902e-06,
+      "loss": 0.333,
+      "step": 12844
+    },
+    {
+      "epoch": 0.9445547466725495,
+      "grad_norm": 0.4580763578414917,
+      "learning_rate": 1.6122699055637235e-06,
+      "loss": 0.3409,
+      "step": 12845
+    },
+    {
+      "epoch": 0.9446282814912861,
+      "grad_norm": 0.4077715277671814,
+      "learning_rate": 1.608013295895039e-06,
+      "loss": 0.3192,
+      "step": 12846
+    },
+    {
+      "epoch": 0.9447018163100228,
+      "grad_norm": 0.4243755638599396,
+      "learning_rate": 1.6037622671236874e-06,
+      "loss": 0.3271,
+      "step": 12847
+    },
+    {
+      "epoch": 0.9447753511287594,
+      "grad_norm": 0.41547298431396484,
+      "learning_rate": 1.5995168194907872e-06,
+      "loss": 0.3284,
+      "step": 12848
+    },
+    {
+      "epoch": 0.9448488859474962,
+      "grad_norm": 0.4217284917831421,
+      "learning_rate": 1.5952769532371236e-06,
+      "loss": 0.3332,
+      "step": 12849
+    },
+    {
+      "epoch": 0.9449224207662328,
+      "grad_norm": 0.4117189645767212,
+      "learning_rate": 1.5910426686032042e-06,
+      "loss": 0.3145,
+      "step": 12850
+    },
+    {
+      "epoch": 0.9449959555849695,
+      "grad_norm": 0.39995384216308594,
+      "learning_rate": 1.5868139658292037e-06,
+      "loss": 0.328,
+      "step": 12851
+    },
+    {
+      "epoch": 0.9450694904037061,
+      "grad_norm": 0.3735046684741974,
+      "learning_rate": 1.5825908451549632e-06,
+      "loss": 0.3546,
+      "step": 12852
+    },
+    {
+      "epoch": 0.9451430252224429,
+      "grad_norm": 0.3762568533420563,
+      "learning_rate": 1.5783733068200468e-06,
+      "loss": 0.3228,
+      "step": 12853
+    },
+    {
+      "epoch": 0.9452165600411795,
+      "grad_norm": 0.45244839787483215,
+      "learning_rate": 1.5741613510636632e-06,
+      "loss": 0.3301,
+      "step": 12854
+    },
+    {
+      "epoch": 0.9452900948599162,
+      "grad_norm": 0.4668266177177429,
+      "learning_rate": 1.5699549781247213e-06,
+      "loss": 0.3198,
+      "step": 12855
+    },
+    {
+      "epoch": 0.9453636296786528,
+      "grad_norm": 0.5311218500137329,
+      "learning_rate": 1.565754188241797e-06,
+      "loss": 0.3488,
+      "step": 12856
+    },
+    {
+      "epoch": 0.9454371644973895,
+      "grad_norm": 0.42301133275032043,
+      "learning_rate": 1.5615589816531994e-06,
+      "loss": 0.3428,
+      "step": 12857
+    },
+    {
+      "epoch": 0.9455106993161262,
+      "grad_norm": 0.4198768138885498,
+      "learning_rate": 1.5573693585968496e-06,
+      "loss": 0.3339,
+      "step": 12858
+    },
+    {
+      "epoch": 0.9455842341348628,
+      "grad_norm": 0.4028914272785187,
+      "learning_rate": 1.5531853193104018e-06,
+      "loss": 0.3014,
+      "step": 12859
+    },
+    {
+      "epoch": 0.9456577689535995,
+      "grad_norm": 0.3689059019088745,
+      "learning_rate": 1.5490068640311772e-06,
+      "loss": 0.3501,
+      "step": 12860
+    },
+    {
+      "epoch": 0.9457313037723362,
+      "grad_norm": 0.5125718116760254,
+      "learning_rate": 1.5448339929961864e-06,
+      "loss": 0.3396,
+      "step": 12861
+    },
+    {
+      "epoch": 0.9458048385910729,
+      "grad_norm": 0.3874472677707672,
+      "learning_rate": 1.5406667064421066e-06,
+      "loss": 0.321,
+      "step": 12862
+    },
+    {
+      "epoch": 0.9458783734098095,
+      "grad_norm": 0.3881443738937378,
+      "learning_rate": 1.5365050046053263e-06,
+      "loss": 0.3141,
+      "step": 12863
+    },
+    {
+      "epoch": 0.9459519082285462,
+      "grad_norm": 0.418083131313324,
+      "learning_rate": 1.5323488877219017e-06,
+      "loss": 0.3123,
+      "step": 12864
+    },
+    {
+      "epoch": 0.9460254430472829,
+      "grad_norm": 0.37440380454063416,
+      "learning_rate": 1.5281983560275548e-06,
+      "loss": 0.3244,
+      "step": 12865
+    },
+    {
+      "epoch": 0.9460989778660196,
+      "grad_norm": 0.4016686975955963,
+      "learning_rate": 1.5240534097577197e-06,
+      "loss": 0.3207,
+      "step": 12866
+    },
+    {
+      "epoch": 0.9461725126847562,
+      "grad_norm": 0.37666648626327515,
+      "learning_rate": 1.5199140491474973e-06,
+      "loss": 0.318,
+      "step": 12867
+    },
+    {
+      "epoch": 0.946246047503493,
+      "grad_norm": 0.3812563717365265,
+      "learning_rate": 1.5157802744316884e-06,
+      "loss": 0.3162,
+      "step": 12868
+    },
+    {
+      "epoch": 0.9463195823222296,
+      "grad_norm": 0.37492579221725464,
+      "learning_rate": 1.5116520858447502e-06,
+      "loss": 0.3465,
+      "step": 12869
+    },
+    {
+      "epoch": 0.9463931171409663,
+      "grad_norm": 0.44046539068222046,
+      "learning_rate": 1.5075294836208509e-06,
+      "loss": 0.3324,
+      "step": 12870
+    },
+    {
+      "epoch": 0.9464666519597029,
+      "grad_norm": 0.3576050102710724,
+      "learning_rate": 1.5034124679938255e-06,
+      "loss": 0.3337,
+      "step": 12871
+    },
+    {
+      "epoch": 0.9465401867784395,
+      "grad_norm": 0.35909318923950195,
+      "learning_rate": 1.4993010391971874e-06,
+      "loss": 0.3134,
+      "step": 12872
+    },
+    {
+      "epoch": 0.9466137215971763,
+      "grad_norm": 0.5033913850784302,
+      "learning_rate": 1.4951951974641386e-06,
+      "loss": 0.2971,
+      "step": 12873
+    },
+    {
+      "epoch": 0.9466872564159129,
+      "grad_norm": 0.3577425479888916,
+      "learning_rate": 1.4910949430275823e-06,
+      "loss": 0.3444,
+      "step": 12874
+    },
+    {
+      "epoch": 0.9467607912346496,
+      "grad_norm": 0.48727473616600037,
+      "learning_rate": 1.4870002761200873e-06,
+      "loss": 0.3097,
+      "step": 12875
+    },
+    {
+      "epoch": 0.9468343260533862,
+      "grad_norm": 0.4036377966403961,
+      "learning_rate": 1.4829111969738907e-06,
+      "loss": 0.3248,
+      "step": 12876
+    },
+    {
+      "epoch": 0.946907860872123,
+      "grad_norm": 0.43332913517951965,
+      "learning_rate": 1.478827705820951e-06,
+      "loss": 0.3155,
+      "step": 12877
+    },
+    {
+      "epoch": 0.9469813956908596,
+      "grad_norm": 0.3897373080253601,
+      "learning_rate": 1.4747498028928609e-06,
+      "loss": 0.314,
+      "step": 12878
+    },
+    {
+      "epoch": 0.9470549305095963,
+      "grad_norm": 0.4704689681529999,
+      "learning_rate": 1.4706774884209462e-06,
+      "loss": 0.3877,
+      "step": 12879
+    },
+    {
+      "epoch": 0.947128465328333,
+      "grad_norm": 0.37957996129989624,
+      "learning_rate": 1.4666107626362002e-06,
+      "loss": 0.325,
+      "step": 12880
+    },
+    {
+      "epoch": 0.9472020001470697,
+      "grad_norm": 0.37616974115371704,
+      "learning_rate": 1.4625496257692606e-06,
+      "loss": 0.3269,
+      "step": 12881
+    },
+    {
+      "epoch": 0.9472755349658063,
+      "grad_norm": 0.39036059379577637,
+      "learning_rate": 1.4584940780505096e-06,
+      "loss": 0.3164,
+      "step": 12882
+    },
+    {
+      "epoch": 0.947349069784543,
+      "grad_norm": 0.3706148564815521,
+      "learning_rate": 1.4544441197099634e-06,
+      "loss": 0.3425,
+      "step": 12883
+    },
+    {
+      "epoch": 0.9474226046032796,
+      "grad_norm": 0.3970668911933899,
+      "learning_rate": 1.4503997509773492e-06,
+      "loss": 0.3276,
+      "step": 12884
+    },
+    {
+      "epoch": 0.9474961394220163,
+      "grad_norm": 0.6417344212532043,
+      "learning_rate": 1.4463609720820615e-06,
+      "loss": 0.3287,
+      "step": 12885
+    },
+    {
+      "epoch": 0.947569674240753,
+      "grad_norm": 0.366873174905777,
+      "learning_rate": 1.4423277832531835e-06,
+      "loss": 0.3282,
+      "step": 12886
+    },
+    {
+      "epoch": 0.9476432090594896,
+      "grad_norm": 0.41999951004981995,
+      "learning_rate": 1.4383001847194989e-06,
+      "loss": 0.2979,
+      "step": 12887
+    },
+    {
+      "epoch": 0.9477167438782264,
+      "grad_norm": 0.38108906149864197,
+      "learning_rate": 1.4342781767094248e-06,
+      "loss": 0.3206,
+      "step": 12888
+    },
+    {
+      "epoch": 0.947790278696963,
+      "grad_norm": 0.4021269679069519,
+      "learning_rate": 1.4302617594511236e-06,
+      "loss": 0.348,
+      "step": 12889
+    },
+    {
+      "epoch": 0.9478638135156997,
+      "grad_norm": 0.39505353569984436,
+      "learning_rate": 1.4262509331724017e-06,
+      "loss": 0.3249,
+      "step": 12890
+    },
+    {
+      "epoch": 0.9479373483344363,
+      "grad_norm": 0.42123401165008545,
+      "learning_rate": 1.4222456981007549e-06,
+      "loss": 0.3127,
+      "step": 12891
+    },
+    {
+      "epoch": 0.9480108831531731,
+      "grad_norm": 0.42893552780151367,
+      "learning_rate": 1.418246054463368e-06,
+      "loss": 0.3284,
+      "step": 12892
+    },
+    {
+      "epoch": 0.9480844179719097,
+      "grad_norm": 0.39829590916633606,
+      "learning_rate": 1.4142520024871042e-06,
+      "loss": 0.3209,
+      "step": 12893
+    },
+    {
+      "epoch": 0.9481579527906464,
+      "grad_norm": 0.3820008337497711,
+      "learning_rate": 1.4102635423985044e-06,
+      "loss": 0.3226,
+      "step": 12894
+    },
+    {
+      "epoch": 0.948231487609383,
+      "grad_norm": 0.35966068506240845,
+      "learning_rate": 1.406280674423799e-06,
+      "loss": 0.3405,
+      "step": 12895
+    },
+    {
+      "epoch": 0.9483050224281198,
+      "grad_norm": 0.35142749547958374,
+      "learning_rate": 1.4023033987889067e-06,
+      "loss": 0.3381,
+      "step": 12896
+    },
+    {
+      "epoch": 0.9483785572468564,
+      "grad_norm": 0.3709595799446106,
+      "learning_rate": 1.3983317157194143e-06,
+      "loss": 0.3209,
+      "step": 12897
+    },
+    {
+      "epoch": 0.9484520920655931,
+      "grad_norm": 0.4498422145843506,
+      "learning_rate": 1.394365625440608e-06,
+      "loss": 0.3286,
+      "step": 12898
+    },
+    {
+      "epoch": 0.9485256268843297,
+      "grad_norm": 0.41318240761756897,
+      "learning_rate": 1.3904051281774522e-06,
+      "loss": 0.3372,
+      "step": 12899
+    },
+    {
+      "epoch": 0.9485991617030664,
+      "grad_norm": 0.3850596249103546,
+      "learning_rate": 1.3864502241545785e-06,
+      "loss": 0.3125,
+      "step": 12900
+    },
+    {
+      "epoch": 0.9486726965218031,
+      "grad_norm": 0.4190703332424164,
+      "learning_rate": 1.3825009135963185e-06,
+      "loss": 0.3238,
+      "step": 12901
+    },
+    {
+      "epoch": 0.9487462313405397,
+      "grad_norm": 0.4199944734573364,
+      "learning_rate": 1.378557196726682e-06,
+      "loss": 0.3075,
+      "step": 12902
+    },
+    {
+      "epoch": 0.9488197661592764,
+      "grad_norm": 0.5277564525604248,
+      "learning_rate": 1.3746190737693786e-06,
+      "loss": 0.3393,
+      "step": 12903
+    },
+    {
+      "epoch": 0.948893300978013,
+      "grad_norm": 0.4092503786087036,
+      "learning_rate": 1.3706865449477413e-06,
+      "loss": 0.3257,
+      "step": 12904
+    },
+    {
+      "epoch": 0.9489668357967498,
+      "grad_norm": 0.6113131046295166,
+      "learning_rate": 1.366759610484869e-06,
+      "loss": 0.3386,
+      "step": 12905
+    },
+    {
+      "epoch": 0.9490403706154864,
+      "grad_norm": 0.3803170621395111,
+      "learning_rate": 1.3628382706034614e-06,
+      "loss": 0.3308,
+      "step": 12906
+    },
+    {
+      "epoch": 0.9491139054342231,
+      "grad_norm": 0.3846963047981262,
+      "learning_rate": 1.3589225255259851e-06,
+      "loss": 0.327,
+      "step": 12907
+    },
+    {
+      "epoch": 0.9491874402529598,
+      "grad_norm": 0.44805923104286194,
+      "learning_rate": 1.355012375474507e-06,
+      "loss": 0.323,
+      "step": 12908
+    },
+    {
+      "epoch": 0.9492609750716965,
+      "grad_norm": 0.40272048115730286,
+      "learning_rate": 1.3511078206708383e-06,
+      "loss": 0.3107,
+      "step": 12909
+    },
+    {
+      "epoch": 0.9493345098904331,
+      "grad_norm": 0.37177592515945435,
+      "learning_rate": 1.3472088613364464e-06,
+      "loss": 0.3424,
+      "step": 12910
+    },
+    {
+      "epoch": 0.9494080447091698,
+      "grad_norm": 0.4952111542224884,
+      "learning_rate": 1.3433154976924767e-06,
+      "loss": 0.3231,
+      "step": 12911
+    },
+    {
+      "epoch": 0.9494815795279065,
+      "grad_norm": 0.424684077501297,
+      "learning_rate": 1.3394277299597634e-06,
+      "loss": 0.3435,
+      "step": 12912
+    },
+    {
+      "epoch": 0.9495551143466431,
+      "grad_norm": 0.4449852406978607,
+      "learning_rate": 1.3355455583588305e-06,
+      "loss": 0.3199,
+      "step": 12913
+    },
+    {
+      "epoch": 0.9496286491653798,
+      "grad_norm": 0.45825129747390747,
+      "learning_rate": 1.3316689831098794e-06,
+      "loss": 0.3407,
+      "step": 12914
+    },
+    {
+      "epoch": 0.9497021839841164,
+      "grad_norm": 0.38972869515419006,
+      "learning_rate": 1.327798004432801e-06,
+      "loss": 0.3153,
+      "step": 12915
+    },
+    {
+      "epoch": 0.9497757188028532,
+      "grad_norm": 0.34823644161224365,
+      "learning_rate": 1.3239326225471415e-06,
+      "loss": 0.3319,
+      "step": 12916
+    },
+    {
+      "epoch": 0.9498492536215898,
+      "grad_norm": 0.3917728662490845,
+      "learning_rate": 1.3200728376721705e-06,
+      "loss": 0.3405,
+      "step": 12917
+    },
+    {
+      "epoch": 0.9499227884403265,
+      "grad_norm": 0.43596649169921875,
+      "learning_rate": 1.3162186500268015e-06,
+      "loss": 0.3022,
+      "step": 12918
+    },
+    {
+      "epoch": 0.9499963232590631,
+      "grad_norm": 0.40561890602111816,
+      "learning_rate": 1.3123700598296595e-06,
+      "loss": 0.3026,
+      "step": 12919
+    },
+    {
+      "epoch": 0.9500698580777999,
+      "grad_norm": 0.3867628276348114,
+      "learning_rate": 1.3085270672990368e-06,
+      "loss": 0.2883,
+      "step": 12920
+    },
+    {
+      "epoch": 0.9501433928965365,
+      "grad_norm": 0.3502522110939026,
+      "learning_rate": 1.3046896726529146e-06,
+      "loss": 0.3272,
+      "step": 12921
+    },
+    {
+      "epoch": 0.9502169277152732,
+      "grad_norm": 0.4699590802192688,
+      "learning_rate": 1.3008578761089407e-06,
+      "loss": 0.3218,
+      "step": 12922
+    },
+    {
+      "epoch": 0.9502904625340098,
+      "grad_norm": 0.4222637712955475,
+      "learning_rate": 1.297031677884486e-06,
+      "loss": 0.3357,
+      "step": 12923
+    },
+    {
+      "epoch": 0.9503639973527466,
+      "grad_norm": 0.40961721539497375,
+      "learning_rate": 1.2932110781965433e-06,
+      "loss": 0.3189,
+      "step": 12924
+    },
+    {
+      "epoch": 0.9504375321714832,
+      "grad_norm": 0.3989596366882324,
+      "learning_rate": 1.2893960772618396e-06,
+      "loss": 0.338,
+      "step": 12925
+    },
+    {
+      "epoch": 0.9505110669902199,
+      "grad_norm": 0.4223251938819885,
+      "learning_rate": 1.2855866752967792e-06,
+      "loss": 0.3035,
+      "step": 12926
+    },
+    {
+      "epoch": 0.9505846018089565,
+      "grad_norm": 0.41485801339149475,
+      "learning_rate": 1.2817828725174008e-06,
+      "loss": 0.3126,
+      "step": 12927
+    },
+    {
+      "epoch": 0.9506581366276932,
+      "grad_norm": 0.35869088768959045,
+      "learning_rate": 1.277984669139487e-06,
+      "loss": 0.3114,
+      "step": 12928
+    },
+    {
+      "epoch": 0.9507316714464299,
+      "grad_norm": 0.40036046504974365,
+      "learning_rate": 1.2741920653784767e-06,
+      "loss": 0.3179,
+      "step": 12929
+    },
+    {
+      "epoch": 0.9508052062651665,
+      "grad_norm": 0.46308979392051697,
+      "learning_rate": 1.2704050614494755e-06,
+      "loss": 0.3041,
+      "step": 12930
+    },
+    {
+      "epoch": 0.9508787410839032,
+      "grad_norm": 0.38731423020362854,
+      "learning_rate": 1.2666236575672895e-06,
+      "loss": 0.3189,
+      "step": 12931
+    },
+    {
+      "epoch": 0.9509522759026399,
+      "grad_norm": 0.4089534878730774,
+      "learning_rate": 1.262847853946414e-06,
+      "loss": 0.2976,
+      "step": 12932
+    },
+    {
+      "epoch": 0.9510258107213766,
+      "grad_norm": 0.39161378145217896,
+      "learning_rate": 1.2590776508010104e-06,
+      "loss": 0.3347,
+      "step": 12933
+    },
+    {
+      "epoch": 0.9510993455401132,
+      "grad_norm": 0.4281909763813019,
+      "learning_rate": 1.2553130483449194e-06,
+      "loss": 0.3492,
+      "step": 12934
+    },
+    {
+      "epoch": 0.95117288035885,
+      "grad_norm": 0.3774206340312958,
+      "learning_rate": 1.2515540467916808e-06,
+      "loss": 0.3025,
+      "step": 12935
+    },
+    {
+      "epoch": 0.9512464151775866,
+      "grad_norm": 0.4368830621242523,
+      "learning_rate": 1.2478006463545245e-06,
+      "loss": 0.348,
+      "step": 12936
+    },
+    {
+      "epoch": 0.9513199499963233,
+      "grad_norm": 0.42124712467193604,
+      "learning_rate": 1.2440528472463132e-06,
+      "loss": 0.3062,
+      "step": 12937
+    },
+    {
+      "epoch": 0.9513934848150599,
+      "grad_norm": 0.4003596305847168,
+      "learning_rate": 1.2403106496796546e-06,
+      "loss": 0.311,
+      "step": 12938
+    },
+    {
+      "epoch": 0.9514670196337967,
+      "grad_norm": 0.46402454376220703,
+      "learning_rate": 1.2365740538668124e-06,
+      "loss": 0.2977,
+      "step": 12939
+    },
+    {
+      "epoch": 0.9515405544525333,
+      "grad_norm": 0.3886529505252838,
+      "learning_rate": 1.232843060019706e-06,
+      "loss": 0.3017,
+      "step": 12940
+    },
+    {
+      "epoch": 0.9516140892712699,
+      "grad_norm": 0.36521318554878235,
+      "learning_rate": 1.2291176683499772e-06,
+      "loss": 0.2992,
+      "step": 12941
+    },
+    {
+      "epoch": 0.9516876240900066,
+      "grad_norm": 0.38199716806411743,
+      "learning_rate": 1.2253978790689235e-06,
+      "loss": 0.2921,
+      "step": 12942
+    },
+    {
+      "epoch": 0.9517611589087432,
+      "grad_norm": 0.39995309710502625,
+      "learning_rate": 1.221683692387554e-06,
+      "loss": 0.3317,
+      "step": 12943
+    },
+    {
+      "epoch": 0.95183469372748,
+      "grad_norm": 0.4354487955570221,
+      "learning_rate": 1.2179751085165335e-06,
+      "loss": 0.3151,
+      "step": 12944
+    },
+    {
+      "epoch": 0.9519082285462166,
+      "grad_norm": 0.381443053483963,
+      "learning_rate": 1.214272127666205e-06,
+      "loss": 0.3182,
+      "step": 12945
+    },
+    {
+      "epoch": 0.9519817633649533,
+      "grad_norm": 0.377545565366745,
+      "learning_rate": 1.2105747500466224e-06,
+      "loss": 0.2944,
+      "step": 12946
+    },
+    {
+      "epoch": 0.95205529818369,
+      "grad_norm": 0.5138674378395081,
+      "learning_rate": 1.206882975867485e-06,
+      "loss": 0.3228,
+      "step": 12947
+    },
+    {
+      "epoch": 0.9521288330024267,
+      "grad_norm": 0.37166157364845276,
+      "learning_rate": 1.2031968053382026e-06,
+      "loss": 0.3854,
+      "step": 12948
+    },
+    {
+      "epoch": 0.9522023678211633,
+      "grad_norm": 0.3837851881980896,
+      "learning_rate": 1.1995162386678748e-06,
+      "loss": 0.3339,
+      "step": 12949
+    },
+    {
+      "epoch": 0.9522759026399,
+      "grad_norm": 0.38290858268737793,
+      "learning_rate": 1.195841276065246e-06,
+      "loss": 0.3279,
+      "step": 12950
+    },
+    {
+      "epoch": 0.9523494374586366,
+      "grad_norm": 0.39146557450294495,
+      "learning_rate": 1.1921719177387824e-06,
+      "loss": 0.3352,
+      "step": 12951
+    },
+    {
+      "epoch": 0.9524229722773734,
+      "grad_norm": 0.3698999881744385,
+      "learning_rate": 1.1885081638965956e-06,
+      "loss": 0.2937,
+      "step": 12952
+    },
+    {
+      "epoch": 0.95249650709611,
+      "grad_norm": 0.4226614534854889,
+      "learning_rate": 1.1848500147465192e-06,
+      "loss": 0.3485,
+      "step": 12953
+    },
+    {
+      "epoch": 0.9525700419148467,
+      "grad_norm": 0.4593416750431061,
+      "learning_rate": 1.1811974704960205e-06,
+      "loss": 0.32,
+      "step": 12954
+    },
+    {
+      "epoch": 0.9526435767335834,
+      "grad_norm": 0.4618067145347595,
+      "learning_rate": 1.1775505313522784e-06,
+      "loss": 0.3525,
+      "step": 12955
+    },
+    {
+      "epoch": 0.95271711155232,
+      "grad_norm": 0.46640387177467346,
+      "learning_rate": 1.1739091975221828e-06,
+      "loss": 0.3653,
+      "step": 12956
+    },
+    {
+      "epoch": 0.9527906463710567,
+      "grad_norm": 0.4263966977596283,
+      "learning_rate": 1.1702734692122352e-06,
+      "loss": 0.3354,
+      "step": 12957
+    },
+    {
+      "epoch": 0.9528641811897933,
+      "grad_norm": 0.39221468567848206,
+      "learning_rate": 1.166643346628682e-06,
+      "loss": 0.3348,
+      "step": 12958
+    },
+    {
+      "epoch": 0.9529377160085301,
+      "grad_norm": 0.3738037049770355,
+      "learning_rate": 1.163018829977436e-06,
+      "loss": 0.3244,
+      "step": 12959
+    },
+    {
+      "epoch": 0.9530112508272667,
+      "grad_norm": 0.43831580877304077,
+      "learning_rate": 1.1593999194640547e-06,
+      "loss": 0.3413,
+      "step": 12960
+    },
+    {
+      "epoch": 0.9530847856460034,
+      "grad_norm": 0.43620747327804565,
+      "learning_rate": 1.155786615293819e-06,
+      "loss": 0.3377,
+      "step": 12961
+    },
+    {
+      "epoch": 0.95315832046474,
+      "grad_norm": 0.4250563085079193,
+      "learning_rate": 1.1521789176716868e-06,
+      "loss": 0.3282,
+      "step": 12962
+    },
+    {
+      "epoch": 0.9532318552834768,
+      "grad_norm": 0.39500829577445984,
+      "learning_rate": 1.1485768268022722e-06,
+      "loss": 0.3463,
+      "step": 12963
+    },
+    {
+      "epoch": 0.9533053901022134,
+      "grad_norm": 0.3806917369365692,
+      "learning_rate": 1.144980342889923e-06,
+      "loss": 0.3535,
+      "step": 12964
+    },
+    {
+      "epoch": 0.9533789249209501,
+      "grad_norm": 0.5220401883125305,
+      "learning_rate": 1.1413894661385982e-06,
+      "loss": 0.3248,
+      "step": 12965
+    },
+    {
+      "epoch": 0.9534524597396867,
+      "grad_norm": 0.3576585650444031,
+      "learning_rate": 1.1378041967520015e-06,
+      "loss": 0.3053,
+      "step": 12966
+    },
+    {
+      "epoch": 0.9535259945584235,
+      "grad_norm": 0.5042484402656555,
+      "learning_rate": 1.1342245349334813e-06,
+      "loss": 0.3248,
+      "step": 12967
+    },
+    {
+      "epoch": 0.9535995293771601,
+      "grad_norm": 0.48786741495132446,
+      "learning_rate": 1.1306504808860752e-06,
+      "loss": 0.3565,
+      "step": 12968
+    },
+    {
+      "epoch": 0.9536730641958967,
+      "grad_norm": 0.4172051250934601,
+      "learning_rate": 1.127082034812521e-06,
+      "loss": 0.3558,
+      "step": 12969
+    },
+    {
+      "epoch": 0.9537465990146334,
+      "grad_norm": 0.42939871549606323,
+      "learning_rate": 1.1235191969152236e-06,
+      "loss": 0.3315,
+      "step": 12970
+    },
+    {
+      "epoch": 0.95382013383337,
+      "grad_norm": 0.4190232455730438,
+      "learning_rate": 1.1199619673962548e-06,
+      "loss": 0.3307,
+      "step": 12971
+    },
+    {
+      "epoch": 0.9538936686521068,
+      "grad_norm": 0.3903329074382782,
+      "learning_rate": 1.1164103464574082e-06,
+      "loss": 0.3344,
+      "step": 12972
+    },
+    {
+      "epoch": 0.9539672034708434,
+      "grad_norm": 0.4005580246448517,
+      "learning_rate": 1.1128643343001122e-06,
+      "loss": 0.3556,
+      "step": 12973
+    },
+    {
+      "epoch": 0.9540407382895801,
+      "grad_norm": 0.4083109498023987,
+      "learning_rate": 1.1093239311255165e-06,
+      "loss": 0.3898,
+      "step": 12974
+    },
+    {
+      "epoch": 0.9541142731083168,
+      "grad_norm": 0.46803712844848633,
+      "learning_rate": 1.1057891371344275e-06,
+      "loss": 0.3095,
+      "step": 12975
+    },
+    {
+      "epoch": 0.9541878079270535,
+      "grad_norm": 0.556422233581543,
+      "learning_rate": 1.1022599525273513e-06,
+      "loss": 0.3368,
+      "step": 12976
+    },
+    {
+      "epoch": 0.9542613427457901,
+      "grad_norm": 0.5589870810508728,
+      "learning_rate": 1.0987363775044612e-06,
+      "loss": 0.3673,
+      "step": 12977
+    },
+    {
+      "epoch": 0.9543348775645268,
+      "grad_norm": 0.38041335344314575,
+      "learning_rate": 1.0952184122656083e-06,
+      "loss": 0.3586,
+      "step": 12978
+    },
+    {
+      "epoch": 0.9544084123832635,
+      "grad_norm": 0.4587945342063904,
+      "learning_rate": 1.0917060570103555e-06,
+      "loss": 0.3553,
+      "step": 12979
+    },
+    {
+      "epoch": 0.9544819472020002,
+      "grad_norm": 0.5060770511627197,
+      "learning_rate": 1.0881993119379098e-06,
+      "loss": 0.3669,
+      "step": 12980
+    },
+    {
+      "epoch": 0.9545554820207368,
+      "grad_norm": 0.40996047854423523,
+      "learning_rate": 1.0846981772471899e-06,
+      "loss": 0.363,
+      "step": 12981
+    },
+    {
+      "epoch": 0.9546290168394735,
+      "grad_norm": 0.3692571520805359,
+      "learning_rate": 1.0812026531367813e-06,
+      "loss": 0.3261,
+      "step": 12982
+    },
+    {
+      "epoch": 0.9547025516582102,
+      "grad_norm": 0.4183506965637207,
+      "learning_rate": 1.077712739804948e-06,
+      "loss": 0.3389,
+      "step": 12983
+    },
+    {
+      "epoch": 0.9547760864769468,
+      "grad_norm": 0.5018148422241211,
+      "learning_rate": 1.0742284374496425e-06,
+      "loss": 0.3739,
+      "step": 12984
+    },
+    {
+      "epoch": 0.9548496212956835,
+      "grad_norm": 0.39466753602027893,
+      "learning_rate": 1.0707497462685067e-06,
+      "loss": 0.3724,
+      "step": 12985
+    },
+    {
+      "epoch": 0.9549231561144201,
+      "grad_norm": 0.3510737121105194,
+      "learning_rate": 1.0672766664588386e-06,
+      "loss": 0.3104,
+      "step": 12986
+    },
+    {
+      "epoch": 0.9549966909331569,
+      "grad_norm": 0.40252119302749634,
+      "learning_rate": 1.0638091982176578e-06,
+      "loss": 0.3423,
+      "step": 12987
+    },
+    {
+      "epoch": 0.9550702257518935,
+      "grad_norm": 0.45956459641456604,
+      "learning_rate": 1.0603473417416298e-06,
+      "loss": 0.335,
+      "step": 12988
+    },
+    {
+      "epoch": 0.9551437605706302,
+      "grad_norm": 0.4408643841743469,
+      "learning_rate": 1.0568910972271084e-06,
+      "loss": 0.3368,
+      "step": 12989
+    },
+    {
+      "epoch": 0.9552172953893668,
+      "grad_norm": 0.3888570964336395,
+      "learning_rate": 1.0534404648701478e-06,
+      "loss": 0.3663,
+      "step": 12990
+    },
+    {
+      "epoch": 0.9552908302081036,
+      "grad_norm": 0.3980591893196106,
+      "learning_rate": 1.0499954448664585e-06,
+      "loss": 0.353,
+      "step": 12991
+    },
+    {
+      "epoch": 0.9553643650268402,
+      "grad_norm": 0.40783074498176575,
+      "learning_rate": 1.0465560374114725e-06,
+      "loss": 0.3558,
+      "step": 12992
+    },
+    {
+      "epoch": 0.9554378998455769,
+      "grad_norm": 0.4384308457374573,
+      "learning_rate": 1.0431222427002341e-06,
+      "loss": 0.3239,
+      "step": 12993
+    },
+    {
+      "epoch": 0.9555114346643135,
+      "grad_norm": 0.39836937189102173,
+      "learning_rate": 1.039694060927554e-06,
+      "loss": 0.3537,
+      "step": 12994
+    },
+    {
+      "epoch": 0.9555849694830503,
+      "grad_norm": 0.48710885643959045,
+      "learning_rate": 1.0362714922878547e-06,
+      "loss": 0.3506,
+      "step": 12995
+    },
+    {
+      "epoch": 0.9556585043017869,
+      "grad_norm": 0.4431850016117096,
+      "learning_rate": 1.0328545369752806e-06,
+      "loss": 0.3563,
+      "step": 12996
+    },
+    {
+      "epoch": 0.9557320391205235,
+      "grad_norm": 0.42028936743736267,
+      "learning_rate": 1.0294431951836324e-06,
+      "loss": 0.3339,
+      "step": 12997
+    },
+    {
+      "epoch": 0.9558055739392602,
+      "grad_norm": 0.4235572814941406,
+      "learning_rate": 1.0260374671064221e-06,
+      "loss": 0.364,
+      "step": 12998
+    },
+    {
+      "epoch": 0.9558791087579969,
+      "grad_norm": 0.3962479829788208,
+      "learning_rate": 1.0226373529368283e-06,
+      "loss": 0.3351,
+      "step": 12999
+    },
+    {
+      "epoch": 0.9559526435767336,
+      "grad_norm": 0.40031373500823975,
+      "learning_rate": 1.0192428528676857e-06,
+      "loss": 0.3325,
+      "step": 13000
+    },
+    {
+      "epoch": 0.9560261783954702,
+      "grad_norm": 0.4008161127567291,
+      "learning_rate": 1.0158539670915402e-06,
+      "loss": 0.3094,
+      "step": 13001
+    },
+    {
+      "epoch": 0.956099713214207,
+      "grad_norm": 0.4448365271091461,
+      "learning_rate": 1.012470695800638e-06,
+      "loss": 0.3371,
+      "step": 13002
+    },
+    {
+      "epoch": 0.9561732480329436,
+      "grad_norm": 0.42194992303848267,
+      "learning_rate": 1.0090930391868481e-06,
+      "loss": 0.3317,
+      "step": 13003
+    },
+    {
+      "epoch": 0.9562467828516803,
+      "grad_norm": 0.4108468294143677,
+      "learning_rate": 1.0057209974417726e-06,
+      "loss": 0.347,
+      "step": 13004
+    },
+    {
+      "epoch": 0.9563203176704169,
+      "grad_norm": 0.4013306796550751,
+      "learning_rate": 1.0023545707566807e-06,
+      "loss": 0.3213,
+      "step": 13005
+    },
+    {
+      "epoch": 0.9563938524891537,
+      "grad_norm": 0.39153265953063965,
+      "learning_rate": 9.989937593225084e-07,
+      "loss": 0.373,
+      "step": 13006
+    },
+    {
+      "epoch": 0.9564673873078903,
+      "grad_norm": 0.4224957525730133,
+      "learning_rate": 9.956385633298925e-07,
+      "loss": 0.32,
+      "step": 13007
+    },
+    {
+      "epoch": 0.956540922126627,
+      "grad_norm": 0.44373416900634766,
+      "learning_rate": 9.922889829691472e-07,
+      "loss": 0.3423,
+      "step": 13008
+    },
+    {
+      "epoch": 0.9566144569453636,
+      "grad_norm": 0.47351035475730896,
+      "learning_rate": 9.88945018430254e-07,
+      "loss": 0.3478,
+      "step": 13009
+    },
+    {
+      "epoch": 0.9566879917641004,
+      "grad_norm": 0.41912996768951416,
+      "learning_rate": 9.856066699028942e-07,
+      "loss": 0.3758,
+      "step": 13010
+    },
+    {
+      "epoch": 0.956761526582837,
+      "grad_norm": 0.36411887407302856,
+      "learning_rate": 9.822739375764056e-07,
+      "loss": 0.3423,
+      "step": 13011
+    },
+    {
+      "epoch": 0.9568350614015736,
+      "grad_norm": 0.4283328652381897,
+      "learning_rate": 9.789468216398478e-07,
+      "loss": 0.3318,
+      "step": 13012
+    },
+    {
+      "epoch": 0.9569085962203103,
+      "grad_norm": 0.3973466753959656,
+      "learning_rate": 9.756253222819146e-07,
+      "loss": 0.3504,
+      "step": 13013
+    },
+    {
+      "epoch": 0.956982131039047,
+      "grad_norm": 0.3670685291290283,
+      "learning_rate": 9.723094396910326e-07,
+      "loss": 0.3597,
+      "step": 13014
+    },
+    {
+      "epoch": 0.9570556658577837,
+      "grad_norm": 0.40678757429122925,
+      "learning_rate": 9.689991740552629e-07,
+      "loss": 0.3402,
+      "step": 13015
+    },
+    {
+      "epoch": 0.9571292006765203,
+      "grad_norm": 0.5150980353355408,
+      "learning_rate": 9.656945255623662e-07,
+      "loss": 0.3495,
+      "step": 13016
+    },
+    {
+      "epoch": 0.957202735495257,
+      "grad_norm": 0.41802024841308594,
+      "learning_rate": 9.623954943997926e-07,
+      "loss": 0.3597,
+      "step": 13017
+    },
+    {
+      "epoch": 0.9572762703139936,
+      "grad_norm": 0.3794597387313843,
+      "learning_rate": 9.591020807546701e-07,
+      "loss": 0.3455,
+      "step": 13018
+    },
+    {
+      "epoch": 0.9573498051327304,
+      "grad_norm": 0.39611420035362244,
+      "learning_rate": 9.55814284813794e-07,
+      "loss": 0.3538,
+      "step": 13019
+    },
+    {
+      "epoch": 0.957423339951467,
+      "grad_norm": 0.4060439467430115,
+      "learning_rate": 9.525321067636595e-07,
+      "loss": 0.3452,
+      "step": 13020
+    },
+    {
+      "epoch": 0.9574968747702037,
+      "grad_norm": 0.4092616140842438,
+      "learning_rate": 9.492555467904285e-07,
+      "loss": 0.3688,
+      "step": 13021
+    },
+    {
+      "epoch": 0.9575704095889404,
+      "grad_norm": 0.3783702254295349,
+      "learning_rate": 9.459846050799637e-07,
+      "loss": 0.3709,
+      "step": 13022
+    },
+    {
+      "epoch": 0.9576439444076771,
+      "grad_norm": 0.3875822424888611,
+      "learning_rate": 9.427192818177833e-07,
+      "loss": 0.3482,
+      "step": 13023
+    },
+    {
+      "epoch": 0.9577174792264137,
+      "grad_norm": 0.3868676424026489,
+      "learning_rate": 9.394595771891057e-07,
+      "loss": 0.3342,
+      "step": 13024
+    },
+    {
+      "epoch": 0.9577910140451503,
+      "grad_norm": 0.43205690383911133,
+      "learning_rate": 9.362054913788276e-07,
+      "loss": 0.3519,
+      "step": 13025
+    },
+    {
+      "epoch": 0.957864548863887,
+      "grad_norm": 0.42997458577156067,
+      "learning_rate": 9.329570245715125e-07,
+      "loss": 0.3446,
+      "step": 13026
+    },
+    {
+      "epoch": 0.9579380836826237,
+      "grad_norm": 0.402998149394989,
+      "learning_rate": 9.29714176951424e-07,
+      "loss": 0.3384,
+      "step": 13027
+    },
+    {
+      "epoch": 0.9580116185013604,
+      "grad_norm": 0.4259917438030243,
+      "learning_rate": 9.264769487025038e-07,
+      "loss": 0.3442,
+      "step": 13028
+    },
+    {
+      "epoch": 0.958085153320097,
+      "grad_norm": 0.3896722197532654,
+      "learning_rate": 9.232453400083718e-07,
+      "loss": 0.3147,
+      "step": 13029
+    },
+    {
+      "epoch": 0.9581586881388338,
+      "grad_norm": 0.3992428183555603,
+      "learning_rate": 9.200193510523147e-07,
+      "loss": 0.3402,
+      "step": 13030
+    },
+    {
+      "epoch": 0.9582322229575704,
+      "grad_norm": 0.42397427558898926,
+      "learning_rate": 9.167989820173306e-07,
+      "loss": 0.3377,
+      "step": 13031
+    },
+    {
+      "epoch": 0.9583057577763071,
+      "grad_norm": 0.3794506788253784,
+      "learning_rate": 9.135842330860734e-07,
+      "loss": 0.357,
+      "step": 13032
+    },
+    {
+      "epoch": 0.9583792925950437,
+      "grad_norm": 0.40130355954170227,
+      "learning_rate": 9.10375104440886e-07,
+      "loss": 0.3414,
+      "step": 13033
+    },
+    {
+      "epoch": 0.9584528274137805,
+      "grad_norm": 0.45720329880714417,
+      "learning_rate": 9.071715962638005e-07,
+      "loss": 0.35,
+      "step": 13034
+    },
+    {
+      "epoch": 0.9585263622325171,
+      "grad_norm": 0.3600628674030304,
+      "learning_rate": 9.039737087365163e-07,
+      "loss": 0.3187,
+      "step": 13035
+    },
+    {
+      "epoch": 0.9585998970512538,
+      "grad_norm": 0.5059010982513428,
+      "learning_rate": 9.007814420404215e-07,
+      "loss": 0.3467,
+      "step": 13036
+    },
+    {
+      "epoch": 0.9586734318699904,
+      "grad_norm": 0.39880675077438354,
+      "learning_rate": 8.975947963565934e-07,
+      "loss": 0.3367,
+      "step": 13037
+    },
+    {
+      "epoch": 0.9587469666887272,
+      "grad_norm": 0.37966403365135193,
+      "learning_rate": 8.944137718657652e-07,
+      "loss": 0.3145,
+      "step": 13038
+    },
+    {
+      "epoch": 0.9588205015074638,
+      "grad_norm": 0.424885630607605,
+      "learning_rate": 8.912383687483817e-07,
+      "loss": 0.3683,
+      "step": 13039
+    },
+    {
+      "epoch": 0.9588940363262004,
+      "grad_norm": 0.37781253457069397,
+      "learning_rate": 8.88068587184554e-07,
+      "loss": 0.3357,
+      "step": 13040
+    },
+    {
+      "epoch": 0.9589675711449371,
+      "grad_norm": 0.520606517791748,
+      "learning_rate": 8.849044273540829e-07,
+      "loss": 0.3226,
+      "step": 13041
+    },
+    {
+      "epoch": 0.9590411059636738,
+      "grad_norm": 0.414310485124588,
+      "learning_rate": 8.817458894364139e-07,
+      "loss": 0.317,
+      "step": 13042
+    },
+    {
+      "epoch": 0.9591146407824105,
+      "grad_norm": 0.38472780585289,
+      "learning_rate": 8.785929736107257e-07,
+      "loss": 0.3339,
+      "step": 13043
+    },
+    {
+      "epoch": 0.9591881756011471,
+      "grad_norm": 0.4139746129512787,
+      "learning_rate": 8.754456800558642e-07,
+      "loss": 0.3416,
+      "step": 13044
+    },
+    {
+      "epoch": 0.9592617104198838,
+      "grad_norm": 0.46741199493408203,
+      "learning_rate": 8.723040089503199e-07,
+      "loss": 0.3292,
+      "step": 13045
+    },
+    {
+      "epoch": 0.9593352452386205,
+      "grad_norm": 0.4980202317237854,
+      "learning_rate": 8.691679604723169e-07,
+      "loss": 0.3426,
+      "step": 13046
+    },
+    {
+      "epoch": 0.9594087800573572,
+      "grad_norm": 0.419131338596344,
+      "learning_rate": 8.660375347997241e-07,
+      "loss": 0.3537,
+      "step": 13047
+    },
+    {
+      "epoch": 0.9594823148760938,
+      "grad_norm": 0.36145487427711487,
+      "learning_rate": 8.629127321100994e-07,
+      "loss": 0.3367,
+      "step": 13048
+    },
+    {
+      "epoch": 0.9595558496948305,
+      "grad_norm": 0.4007362425327301,
+      "learning_rate": 8.597935525806788e-07,
+      "loss": 0.3262,
+      "step": 13049
+    },
+    {
+      "epoch": 0.9596293845135672,
+      "grad_norm": 0.40373504161834717,
+      "learning_rate": 8.566799963883987e-07,
+      "loss": 0.316,
+      "step": 13050
+    },
+    {
+      "epoch": 0.9597029193323039,
+      "grad_norm": 0.3784237504005432,
+      "learning_rate": 8.535720637098732e-07,
+      "loss": 0.3542,
+      "step": 13051
+    },
+    {
+      "epoch": 0.9597764541510405,
+      "grad_norm": 0.38241589069366455,
+      "learning_rate": 8.504697547213613e-07,
+      "loss": 0.3217,
+      "step": 13052
+    },
+    {
+      "epoch": 0.9598499889697771,
+      "grad_norm": 0.41511350870132446,
+      "learning_rate": 8.473730695988335e-07,
+      "loss": 0.3171,
+      "step": 13053
+    },
+    {
+      "epoch": 0.9599235237885139,
+      "grad_norm": 0.3757646977901459,
+      "learning_rate": 8.442820085179604e-07,
+      "loss": 0.3348,
+      "step": 13054
+    },
+    {
+      "epoch": 0.9599970586072505,
+      "grad_norm": 0.3896996080875397,
+      "learning_rate": 8.41196571654046e-07,
+      "loss": 0.3296,
+      "step": 13055
+    },
+    {
+      "epoch": 0.9600705934259872,
+      "grad_norm": 0.42579343914985657,
+      "learning_rate": 8.38116759182117e-07,
+      "loss": 0.3413,
+      "step": 13056
+    },
+    {
+      "epoch": 0.9601441282447238,
+      "grad_norm": 0.37679994106292725,
+      "learning_rate": 8.35042571276845e-07,
+      "loss": 0.3618,
+      "step": 13057
+    },
+    {
+      "epoch": 0.9602176630634606,
+      "grad_norm": 0.4462145268917084,
+      "learning_rate": 8.319740081126237e-07,
+      "loss": 0.3255,
+      "step": 13058
+    },
+    {
+      "epoch": 0.9602911978821972,
+      "grad_norm": 0.45632073283195496,
+      "learning_rate": 8.289110698634806e-07,
+      "loss": 0.3392,
+      "step": 13059
+    },
+    {
+      "epoch": 0.9603647327009339,
+      "grad_norm": 0.396056592464447,
+      "learning_rate": 8.258537567031654e-07,
+      "loss": 0.3405,
+      "step": 13060
+    },
+    {
+      "epoch": 0.9604382675196705,
+      "grad_norm": 0.40049612522125244,
+      "learning_rate": 8.228020688050953e-07,
+      "loss": 0.3637,
+      "step": 13061
+    },
+    {
+      "epoch": 0.9605118023384073,
+      "grad_norm": 0.426172137260437,
+      "learning_rate": 8.197560063423426e-07,
+      "loss": 0.3442,
+      "step": 13062
+    },
+    {
+      "epoch": 0.9605853371571439,
+      "grad_norm": 0.38611412048339844,
+      "learning_rate": 8.167155694877027e-07,
+      "loss": 0.3564,
+      "step": 13063
+    },
+    {
+      "epoch": 0.9606588719758806,
+      "grad_norm": 0.43075254559516907,
+      "learning_rate": 8.136807584136374e-07,
+      "loss": 0.3457,
+      "step": 13064
+    },
+    {
+      "epoch": 0.9607324067946172,
+      "grad_norm": 0.39291930198669434,
+      "learning_rate": 8.106515732922648e-07,
+      "loss": 0.3485,
+      "step": 13065
+    },
+    {
+      "epoch": 0.960805941613354,
+      "grad_norm": 0.4317275881767273,
+      "learning_rate": 8.076280142954252e-07,
+      "loss": 0.3134,
+      "step": 13066
+    },
+    {
+      "epoch": 0.9608794764320906,
+      "grad_norm": 0.43894070386886597,
+      "learning_rate": 8.046100815946034e-07,
+      "loss": 0.3293,
+      "step": 13067
+    },
+    {
+      "epoch": 0.9609530112508272,
+      "grad_norm": 0.3626691401004791,
+      "learning_rate": 8.015977753609738e-07,
+      "loss": 0.3594,
+      "step": 13068
+    },
+    {
+      "epoch": 0.961026546069564,
+      "grad_norm": 0.4021105468273163,
+      "learning_rate": 7.985910957654108e-07,
+      "loss": 0.3511,
+      "step": 13069
+    },
+    {
+      "epoch": 0.9611000808883006,
+      "grad_norm": 0.40065810084342957,
+      "learning_rate": 7.955900429784558e-07,
+      "loss": 0.3471,
+      "step": 13070
+    },
+    {
+      "epoch": 0.9611736157070373,
+      "grad_norm": 0.3705725073814392,
+      "learning_rate": 7.925946171703281e-07,
+      "loss": 0.3285,
+      "step": 13071
+    },
+    {
+      "epoch": 0.9612471505257739,
+      "grad_norm": 0.4500194489955902,
+      "learning_rate": 7.89604818510925e-07,
+      "loss": 0.3348,
+      "step": 13072
+    },
+    {
+      "epoch": 0.9613206853445107,
+      "grad_norm": 0.41561266779899597,
+      "learning_rate": 7.866206471698446e-07,
+      "loss": 0.3244,
+      "step": 13073
+    },
+    {
+      "epoch": 0.9613942201632473,
+      "grad_norm": 0.3764963150024414,
+      "learning_rate": 7.8364210331634e-07,
+      "loss": 0.3489,
+      "step": 13074
+    },
+    {
+      "epoch": 0.961467754981984,
+      "grad_norm": 0.3907318115234375,
+      "learning_rate": 7.80669187119365e-07,
+      "loss": 0.3604,
+      "step": 13075
+    },
+    {
+      "epoch": 0.9615412898007206,
+      "grad_norm": 0.42085665464401245,
+      "learning_rate": 7.777018987475404e-07,
+      "loss": 0.3242,
+      "step": 13076
+    },
+    {
+      "epoch": 0.9616148246194574,
+      "grad_norm": 0.3907133638858795,
+      "learning_rate": 7.747402383691871e-07,
+      "loss": 0.3616,
+      "step": 13077
+    },
+    {
+      "epoch": 0.961688359438194,
+      "grad_norm": 0.3978095054626465,
+      "learning_rate": 7.717842061522706e-07,
+      "loss": 0.3469,
+      "step": 13078
+    },
+    {
+      "epoch": 0.9617618942569307,
+      "grad_norm": 0.37387076020240784,
+      "learning_rate": 7.688338022644792e-07,
+      "loss": 0.3414,
+      "step": 13079
+    },
+    {
+      "epoch": 0.9618354290756673,
+      "grad_norm": 0.39175209403038025,
+      "learning_rate": 7.658890268731567e-07,
+      "loss": 0.3294,
+      "step": 13080
+    },
+    {
+      "epoch": 0.961908963894404,
+      "grad_norm": 0.41644948720932007,
+      "learning_rate": 7.629498801453472e-07,
+      "loss": 0.3285,
+      "step": 13081
+    },
+    {
+      "epoch": 0.9619824987131407,
+      "grad_norm": 0.3835298717021942,
+      "learning_rate": 7.600163622477285e-07,
+      "loss": 0.3158,
+      "step": 13082
+    },
+    {
+      "epoch": 0.9620560335318773,
+      "grad_norm": 0.3850473165512085,
+      "learning_rate": 7.570884733467231e-07,
+      "loss": 0.3253,
+      "step": 13083
+    },
+    {
+      "epoch": 0.962129568350614,
+      "grad_norm": 0.37342050671577454,
+      "learning_rate": 7.541662136083982e-07,
+      "loss": 0.321,
+      "step": 13084
+    },
+    {
+      "epoch": 0.9622031031693506,
+      "grad_norm": 0.4025447964668274,
+      "learning_rate": 7.5124958319851e-07,
+      "loss": 0.3173,
+      "step": 13085
+    },
+    {
+      "epoch": 0.9622766379880874,
+      "grad_norm": 0.3649718761444092,
+      "learning_rate": 7.483385822824706e-07,
+      "loss": 0.3277,
+      "step": 13086
+    },
+    {
+      "epoch": 0.962350172806824,
+      "grad_norm": 0.4818938076496124,
+      "learning_rate": 7.454332110254259e-07,
+      "loss": 0.3313,
+      "step": 13087
+    },
+    {
+      "epoch": 0.9624237076255607,
+      "grad_norm": 0.40357211232185364,
+      "learning_rate": 7.425334695921437e-07,
+      "loss": 0.3363,
+      "step": 13088
+    },
+    {
+      "epoch": 0.9624972424442974,
+      "grad_norm": 0.4859367907047272,
+      "learning_rate": 7.39639358147115e-07,
+      "loss": 0.3298,
+      "step": 13089
+    },
+    {
+      "epoch": 0.9625707772630341,
+      "grad_norm": 0.3601112365722656,
+      "learning_rate": 7.367508768545084e-07,
+      "loss": 0.3148,
+      "step": 13090
+    },
+    {
+      "epoch": 0.9626443120817707,
+      "grad_norm": 0.4033958911895752,
+      "learning_rate": 7.338680258781372e-07,
+      "loss": 0.3465,
+      "step": 13091
+    },
+    {
+      "epoch": 0.9627178469005074,
+      "grad_norm": 0.3634166121482849,
+      "learning_rate": 7.309908053815373e-07,
+      "loss": 0.3107,
+      "step": 13092
+    },
+    {
+      "epoch": 0.962791381719244,
+      "grad_norm": 0.38710328936576843,
+      "learning_rate": 7.281192155279004e-07,
+      "loss": 0.3369,
+      "step": 13093
+    },
+    {
+      "epoch": 0.9628649165379808,
+      "grad_norm": 0.37955018877983093,
+      "learning_rate": 7.252532564801073e-07,
+      "loss": 0.3197,
+      "step": 13094
+    },
+    {
+      "epoch": 0.9629384513567174,
+      "grad_norm": 0.39843955636024475,
+      "learning_rate": 7.223929284007169e-07,
+      "loss": 0.327,
+      "step": 13095
+    },
+    {
+      "epoch": 0.963011986175454,
+      "grad_norm": 0.3890419006347656,
+      "learning_rate": 7.195382314519661e-07,
+      "loss": 0.329,
+      "step": 13096
+    },
+    {
+      "epoch": 0.9630855209941908,
+      "grad_norm": 0.44457098841667175,
+      "learning_rate": 7.166891657957919e-07,
+      "loss": 0.3138,
+      "step": 13097
+    },
+    {
+      "epoch": 0.9631590558129274,
+      "grad_norm": 0.4046995937824249,
+      "learning_rate": 7.138457315937764e-07,
+      "loss": 0.331,
+      "step": 13098
+    },
+    {
+      "epoch": 0.9632325906316641,
+      "grad_norm": 0.4092302918434143,
+      "learning_rate": 7.110079290072236e-07,
+      "loss": 0.3177,
+      "step": 13099
+    },
+    {
+      "epoch": 0.9633061254504007,
+      "grad_norm": 0.354497492313385,
+      "learning_rate": 7.081757581970828e-07,
+      "loss": 0.2962,
+      "step": 13100
+    },
+    {
+      "epoch": 0.9633796602691375,
+      "grad_norm": 0.43179383873939514,
+      "learning_rate": 7.053492193239919e-07,
+      "loss": 0.3323,
+      "step": 13101
+    },
+    {
+      "epoch": 0.9634531950878741,
+      "grad_norm": 0.381283700466156,
+      "learning_rate": 7.025283125482784e-07,
+      "loss": 0.3041,
+      "step": 13102
+    },
+    {
+      "epoch": 0.9635267299066108,
+      "grad_norm": 0.4441286325454712,
+      "learning_rate": 6.9971303802997e-07,
+      "loss": 0.3386,
+      "step": 13103
+    },
+    {
+      "epoch": 0.9636002647253474,
+      "grad_norm": 0.4268989562988281,
+      "learning_rate": 6.969033959287275e-07,
+      "loss": 0.3563,
+      "step": 13104
+    },
+    {
+      "epoch": 0.9636737995440842,
+      "grad_norm": 0.4739101231098175,
+      "learning_rate": 6.940993864039125e-07,
+      "loss": 0.361,
+      "step": 13105
+    },
+    {
+      "epoch": 0.9637473343628208,
+      "grad_norm": 0.3808363378047943,
+      "learning_rate": 6.913010096145867e-07,
+      "loss": 0.3122,
+      "step": 13106
+    },
+    {
+      "epoch": 0.9638208691815575,
+      "grad_norm": 0.3784494400024414,
+      "learning_rate": 6.885082657194785e-07,
+      "loss": 0.324,
+      "step": 13107
+    },
+    {
+      "epoch": 0.9638944040002941,
+      "grad_norm": 0.39745333790779114,
+      "learning_rate": 6.857211548769837e-07,
+      "loss": 0.3078,
+      "step": 13108
+    },
+    {
+      "epoch": 0.9639679388190308,
+      "grad_norm": 0.41966554522514343,
+      "learning_rate": 6.829396772451979e-07,
+      "loss": 0.3114,
+      "step": 13109
+    },
+    {
+      "epoch": 0.9640414736377675,
+      "grad_norm": 0.41263312101364136,
+      "learning_rate": 6.801638329818838e-07,
+      "loss": 0.301,
+      "step": 13110
+    },
+    {
+      "epoch": 0.9641150084565041,
+      "grad_norm": 0.4113951325416565,
+      "learning_rate": 6.773936222444932e-07,
+      "loss": 0.3152,
+      "step": 13111
+    },
+    {
+      "epoch": 0.9641885432752408,
+      "grad_norm": 0.417944997549057,
+      "learning_rate": 6.74629045190156e-07,
+      "loss": 0.3363,
+      "step": 13112
+    },
+    {
+      "epoch": 0.9642620780939775,
+      "grad_norm": 0.4070371091365814,
+      "learning_rate": 6.718701019756913e-07,
+      "loss": 0.327,
+      "step": 13113
+    },
+    {
+      "epoch": 0.9643356129127142,
+      "grad_norm": 0.38413962721824646,
+      "learning_rate": 6.691167927575626e-07,
+      "loss": 0.2966,
+      "step": 13114
+    },
+    {
+      "epoch": 0.9644091477314508,
+      "grad_norm": 0.4421108365058899,
+      "learning_rate": 6.663691176919673e-07,
+      "loss": 0.3229,
+      "step": 13115
+    },
+    {
+      "epoch": 0.9644826825501875,
+      "grad_norm": 0.40369877219200134,
+      "learning_rate": 6.636270769347475e-07,
+      "loss": 0.3343,
+      "step": 13116
+    },
+    {
+      "epoch": 0.9645562173689242,
+      "grad_norm": 0.3757677674293518,
+      "learning_rate": 6.608906706414342e-07,
+      "loss": 0.3283,
+      "step": 13117
+    },
+    {
+      "epoch": 0.9646297521876609,
+      "grad_norm": 0.41671022772789,
+      "learning_rate": 6.581598989672366e-07,
+      "loss": 0.3163,
+      "step": 13118
+    },
+    {
+      "epoch": 0.9647032870063975,
+      "grad_norm": 0.41891029477119446,
+      "learning_rate": 6.554347620670531e-07,
+      "loss": 0.3139,
+      "step": 13119
+    },
+    {
+      "epoch": 0.9647768218251342,
+      "grad_norm": 0.37070345878601074,
+      "learning_rate": 6.52715260095449e-07,
+      "loss": 0.301,
+      "step": 13120
+    },
+    {
+      "epoch": 0.9648503566438709,
+      "grad_norm": 0.37676987051963806,
+      "learning_rate": 6.500013932066784e-07,
+      "loss": 0.3392,
+      "step": 13121
+    },
+    {
+      "epoch": 0.9649238914626076,
+      "grad_norm": 0.3678625822067261,
+      "learning_rate": 6.47293161554674e-07,
+      "loss": 0.3252,
+      "step": 13122
+    },
+    {
+      "epoch": 0.9649974262813442,
+      "grad_norm": 0.3397931456565857,
+      "learning_rate": 6.445905652930683e-07,
+      "loss": 0.304,
+      "step": 13123
+    },
+    {
+      "epoch": 0.9650709611000808,
+      "grad_norm": 0.37787899374961853,
+      "learning_rate": 6.418936045751278e-07,
+      "loss": 0.316,
+      "step": 13124
+    },
+    {
+      "epoch": 0.9651444959188176,
+      "grad_norm": 0.6305922865867615,
+      "learning_rate": 6.392022795538299e-07,
+      "loss": 0.3217,
+      "step": 13125
+    },
+    {
+      "epoch": 0.9652180307375542,
+      "grad_norm": 0.4060790240764618,
+      "learning_rate": 6.365165903818526e-07,
+      "loss": 0.3339,
+      "step": 13126
+    },
+    {
+      "epoch": 0.9652915655562909,
+      "grad_norm": 0.40128982067108154,
+      "learning_rate": 6.338365372115185e-07,
+      "loss": 0.3142,
+      "step": 13127
+    },
+    {
+      "epoch": 0.9653651003750275,
+      "grad_norm": 0.5002443790435791,
+      "learning_rate": 6.311621201948281e-07,
+      "loss": 0.3345,
+      "step": 13128
+    },
+    {
+      "epoch": 0.9654386351937643,
+      "grad_norm": 0.3958386480808258,
+      "learning_rate": 6.284933394835046e-07,
+      "loss": 0.336,
+      "step": 13129
+    },
+    {
+      "epoch": 0.9655121700125009,
+      "grad_norm": 0.4154229760169983,
+      "learning_rate": 6.258301952289048e-07,
+      "loss": 0.3124,
+      "step": 13130
+    },
+    {
+      "epoch": 0.9655857048312376,
+      "grad_norm": 0.3390556573867798,
+      "learning_rate": 6.231726875820853e-07,
+      "loss": 0.3161,
+      "step": 13131
+    },
+    {
+      "epoch": 0.9656592396499742,
+      "grad_norm": 0.4429776966571808,
+      "learning_rate": 6.205208166937814e-07,
+      "loss": 0.3209,
+      "step": 13132
+    },
+    {
+      "epoch": 0.965732774468711,
+      "grad_norm": 0.3781125247478485,
+      "learning_rate": 6.178745827144283e-07,
+      "loss": 0.3162,
+      "step": 13133
+    },
+    {
+      "epoch": 0.9658063092874476,
+      "grad_norm": 0.4031316936016083,
+      "learning_rate": 6.152339857941059e-07,
+      "loss": 0.3678,
+      "step": 13134
+    },
+    {
+      "epoch": 0.9658798441061843,
+      "grad_norm": 0.4275887608528137,
+      "learning_rate": 6.125990260826053e-07,
+      "loss": 0.324,
+      "step": 13135
+    },
+    {
+      "epoch": 0.965953378924921,
+      "grad_norm": 0.4078723192214966,
+      "learning_rate": 6.099697037293628e-07,
+      "loss": 0.3131,
+      "step": 13136
+    },
+    {
+      "epoch": 0.9660269137436576,
+      "grad_norm": 0.3881116509437561,
+      "learning_rate": 6.073460188835367e-07,
+      "loss": 0.3331,
+      "step": 13137
+    },
+    {
+      "epoch": 0.9661004485623943,
+      "grad_norm": 0.37818577885627747,
+      "learning_rate": 6.047279716939413e-07,
+      "loss": 0.3294,
+      "step": 13138
+    },
+    {
+      "epoch": 0.9661739833811309,
+      "grad_norm": 0.4047588109970093,
+      "learning_rate": 6.02115562309069e-07,
+      "loss": 0.3571,
+      "step": 13139
+    },
+    {
+      "epoch": 0.9662475181998676,
+      "grad_norm": 0.4344087541103363,
+      "learning_rate": 5.995087908771013e-07,
+      "loss": 0.3455,
+      "step": 13140
+    },
+    {
+      "epoch": 0.9663210530186043,
+      "grad_norm": 0.4576096534729004,
+      "learning_rate": 5.969076575458865e-07,
+      "loss": 0.328,
+      "step": 13141
+    },
+    {
+      "epoch": 0.966394587837341,
+      "grad_norm": 0.3872050940990448,
+      "learning_rate": 5.943121624629733e-07,
+      "loss": 0.309,
+      "step": 13142
+    },
+    {
+      "epoch": 0.9664681226560776,
+      "grad_norm": 0.4250621497631073,
+      "learning_rate": 5.917223057755994e-07,
+      "loss": 0.3254,
+      "step": 13143
+    },
+    {
+      "epoch": 0.9665416574748144,
+      "grad_norm": 0.42163780331611633,
+      "learning_rate": 5.891380876306251e-07,
+      "loss": 0.3535,
+      "step": 13144
+    },
+    {
+      "epoch": 0.966615192293551,
+      "grad_norm": 0.36661025881767273,
+      "learning_rate": 5.865595081746667e-07,
+      "loss": 0.3545,
+      "step": 13145
+    },
+    {
+      "epoch": 0.9666887271122877,
+      "grad_norm": 0.43692800402641296,
+      "learning_rate": 5.839865675539624e-07,
+      "loss": 0.366,
+      "step": 13146
+    },
+    {
+      "epoch": 0.9667622619310243,
+      "grad_norm": 0.41102132201194763,
+      "learning_rate": 5.814192659144513e-07,
+      "loss": 0.3086,
+      "step": 13147
+    },
+    {
+      "epoch": 0.9668357967497611,
+      "grad_norm": 0.3847244381904602,
+      "learning_rate": 5.788576034017723e-07,
+      "loss": 0.314,
+      "step": 13148
+    },
+    {
+      "epoch": 0.9669093315684977,
+      "grad_norm": 0.39922085404396057,
+      "learning_rate": 5.763015801612093e-07,
+      "loss": 0.3291,
+      "step": 13149
+    },
+    {
+      "epoch": 0.9669828663872344,
+      "grad_norm": 0.4076964855194092,
+      "learning_rate": 5.737511963377351e-07,
+      "loss": 0.3408,
+      "step": 13150
+    },
+    {
+      "epoch": 0.967056401205971,
+      "grad_norm": 0.37911367416381836,
+      "learning_rate": 5.712064520760341e-07,
+      "loss": 0.3209,
+      "step": 13151
+    },
+    {
+      "epoch": 0.9671299360247076,
+      "grad_norm": 0.39863070845603943,
+      "learning_rate": 5.686673475204351e-07,
+      "loss": 0.315,
+      "step": 13152
+    },
+    {
+      "epoch": 0.9672034708434444,
+      "grad_norm": 0.3586221933364868,
+      "learning_rate": 5.661338828149565e-07,
+      "loss": 0.2964,
+      "step": 13153
+    },
+    {
+      "epoch": 0.967277005662181,
+      "grad_norm": 0.3641381859779358,
+      "learning_rate": 5.636060581032942e-07,
+      "loss": 0.3092,
+      "step": 13154
+    },
+    {
+      "epoch": 0.9673505404809177,
+      "grad_norm": 0.46381819248199463,
+      "learning_rate": 5.610838735288449e-07,
+      "loss": 0.3214,
+      "step": 13155
+    },
+    {
+      "epoch": 0.9674240752996544,
+      "grad_norm": 0.4251638352870941,
+      "learning_rate": 5.585673292346605e-07,
+      "loss": 0.364,
+      "step": 13156
+    },
+    {
+      "epoch": 0.9674976101183911,
+      "grad_norm": 0.3947910964488983,
+      "learning_rate": 5.560564253634714e-07,
+      "loss": 0.326,
+      "step": 13157
+    },
+    {
+      "epoch": 0.9675711449371277,
+      "grad_norm": 0.4485805034637451,
+      "learning_rate": 5.535511620577194e-07,
+      "loss": 0.3286,
+      "step": 13158
+    },
+    {
+      "epoch": 0.9676446797558644,
+      "grad_norm": 0.36305731534957886,
+      "learning_rate": 5.510515394594906e-07,
+      "loss": 0.3044,
+      "step": 13159
+    },
+    {
+      "epoch": 0.967718214574601,
+      "grad_norm": 0.3861240744590759,
+      "learning_rate": 5.485575577105717e-07,
+      "loss": 0.3396,
+      "step": 13160
+    },
+    {
+      "epoch": 0.9677917493933378,
+      "grad_norm": 0.3499179184436798,
+      "learning_rate": 5.460692169524162e-07,
+      "loss": 0.2982,
+      "step": 13161
+    },
+    {
+      "epoch": 0.9678652842120744,
+      "grad_norm": 0.47073304653167725,
+      "learning_rate": 5.43586517326189e-07,
+      "loss": 0.3322,
+      "step": 13162
+    },
+    {
+      "epoch": 0.9679388190308111,
+      "grad_norm": 0.44633564352989197,
+      "learning_rate": 5.411094589726884e-07,
+      "loss": 0.3183,
+      "step": 13163
+    },
+    {
+      "epoch": 0.9680123538495478,
+      "grad_norm": 0.4277419447898865,
+      "learning_rate": 5.386380420324133e-07,
+      "loss": 0.3195,
+      "step": 13164
+    },
+    {
+      "epoch": 0.9680858886682844,
+      "grad_norm": 0.41118207573890686,
+      "learning_rate": 5.361722666455626e-07,
+      "loss": 0.331,
+      "step": 13165
+    },
+    {
+      "epoch": 0.9681594234870211,
+      "grad_norm": 0.43470320105552673,
+      "learning_rate": 5.337121329519912e-07,
+      "loss": 0.3384,
+      "step": 13166
+    },
+    {
+      "epoch": 0.9682329583057577,
+      "grad_norm": 0.3760433495044708,
+      "learning_rate": 5.312576410912318e-07,
+      "loss": 0.3439,
+      "step": 13167
+    },
+    {
+      "epoch": 0.9683064931244945,
+      "grad_norm": 0.4721154570579529,
+      "learning_rate": 5.288087912025175e-07,
+      "loss": 0.3507,
+      "step": 13168
+    },
+    {
+      "epoch": 0.9683800279432311,
+      "grad_norm": 0.3899043798446655,
+      "learning_rate": 5.263655834247483e-07,
+      "loss": 0.3236,
+      "step": 13169
+    },
+    {
+      "epoch": 0.9684535627619678,
+      "grad_norm": 0.42476987838745117,
+      "learning_rate": 5.239280178964912e-07,
+      "loss": 0.3435,
+      "step": 13170
+    },
+    {
+      "epoch": 0.9685270975807044,
+      "grad_norm": 0.38242584466934204,
+      "learning_rate": 5.214960947560244e-07,
+      "loss": 0.3225,
+      "step": 13171
+    },
+    {
+      "epoch": 0.9686006323994412,
+      "grad_norm": 0.38265299797058105,
+      "learning_rate": 5.190698141412819e-07,
+      "loss": 0.3642,
+      "step": 13172
+    },
+    {
+      "epoch": 0.9686741672181778,
+      "grad_norm": 0.4149736166000366,
+      "learning_rate": 5.166491761898873e-07,
+      "loss": 0.3125,
+      "step": 13173
+    },
+    {
+      "epoch": 0.9687477020369145,
+      "grad_norm": 0.4311352074146271,
+      "learning_rate": 5.142341810391416e-07,
+      "loss": 0.31,
+      "step": 13174
+    },
+    {
+      "epoch": 0.9688212368556511,
+      "grad_norm": 0.3693532645702362,
+      "learning_rate": 5.118248288260241e-07,
+      "loss": 0.3497,
+      "step": 13175
+    },
+    {
+      "epoch": 0.9688947716743879,
+      "grad_norm": 0.4074828028678894,
+      "learning_rate": 5.094211196871923e-07,
+      "loss": 0.3234,
+      "step": 13176
+    },
+    {
+      "epoch": 0.9689683064931245,
+      "grad_norm": 0.5208801627159119,
+      "learning_rate": 5.070230537589926e-07,
+      "loss": 0.3503,
+      "step": 13177
+    },
+    {
+      "epoch": 0.9690418413118612,
+      "grad_norm": 0.3946942985057831,
+      "learning_rate": 5.046306311774385e-07,
+      "loss": 0.3225,
+      "step": 13178
+    },
+    {
+      "epoch": 0.9691153761305978,
+      "grad_norm": 0.3711612820625305,
+      "learning_rate": 5.022438520782546e-07,
+      "loss": 0.3058,
+      "step": 13179
+    },
+    {
+      "epoch": 0.9691889109493345,
+      "grad_norm": 0.37817880511283875,
+      "learning_rate": 4.998627165967884e-07,
+      "loss": 0.3321,
+      "step": 13180
+    },
+    {
+      "epoch": 0.9692624457680712,
+      "grad_norm": 0.38563504815101624,
+      "learning_rate": 4.974872248681206e-07,
+      "loss": 0.3337,
+      "step": 13181
+    },
+    {
+      "epoch": 0.9693359805868078,
+      "grad_norm": 0.4169960021972656,
+      "learning_rate": 4.951173770269879e-07,
+      "loss": 0.3148,
+      "step": 13182
+    },
+    {
+      "epoch": 0.9694095154055445,
+      "grad_norm": 0.45209163427352905,
+      "learning_rate": 4.92753173207805e-07,
+      "loss": 0.3118,
+      "step": 13183
+    },
+    {
+      "epoch": 0.9694830502242812,
+      "grad_norm": 0.36850547790527344,
+      "learning_rate": 4.903946135446758e-07,
+      "loss": 0.3133,
+      "step": 13184
+    },
+    {
+      "epoch": 0.9695565850430179,
+      "grad_norm": 0.43797066807746887,
+      "learning_rate": 4.88041698171382e-07,
+      "loss": 0.3098,
+      "step": 13185
+    },
+    {
+      "epoch": 0.9696301198617545,
+      "grad_norm": 0.37009280920028687,
+      "learning_rate": 4.856944272213837e-07,
+      "loss": 0.3518,
+      "step": 13186
+    },
+    {
+      "epoch": 0.9697036546804912,
+      "grad_norm": 0.3593955636024475,
+      "learning_rate": 4.833528008278187e-07,
+      "loss": 0.2913,
+      "step": 13187
+    },
+    {
+      "epoch": 0.9697771894992279,
+      "grad_norm": 0.4284631311893463,
+      "learning_rate": 4.810168191235032e-07,
+      "loss": 0.3107,
+      "step": 13188
+    },
+    {
+      "epoch": 0.9698507243179646,
+      "grad_norm": 0.3963834047317505,
+      "learning_rate": 4.786864822409532e-07,
+      "loss": 0.2982,
+      "step": 13189
+    },
+    {
+      "epoch": 0.9699242591367012,
+      "grad_norm": 0.3764474093914032,
+      "learning_rate": 4.763617903123185e-07,
+      "loss": 0.3412,
+      "step": 13190
+    },
+    {
+      "epoch": 0.969997793955438,
+      "grad_norm": 0.3781980276107788,
+      "learning_rate": 4.7404274346948275e-07,
+      "loss": 0.3323,
+      "step": 13191
+    },
+    {
+      "epoch": 0.9700713287741746,
+      "grad_norm": 0.442199170589447,
+      "learning_rate": 4.71729341843985e-07,
+      "loss": 0.3039,
+      "step": 13192
+    },
+    {
+      "epoch": 0.9701448635929112,
+      "grad_norm": 0.40059927105903625,
+      "learning_rate": 4.6942158556702033e-07,
+      "loss": 0.3109,
+      "step": 13193
+    },
+    {
+      "epoch": 0.9702183984116479,
+      "grad_norm": 0.3626280725002289,
+      "learning_rate": 4.6711947476950626e-07,
+      "loss": 0.301,
+      "step": 13194
+    },
+    {
+      "epoch": 0.9702919332303845,
+      "grad_norm": 0.3873019516468048,
+      "learning_rate": 4.648230095820161e-07,
+      "loss": 0.3076,
+      "step": 13195
+    },
+    {
+      "epoch": 0.9703654680491213,
+      "grad_norm": 0.42981594800949097,
+      "learning_rate": 4.625321901348123e-07,
+      "loss": 0.3157,
+      "step": 13196
+    },
+    {
+      "epoch": 0.9704390028678579,
+      "grad_norm": 0.38062602281570435,
+      "learning_rate": 4.602470165578354e-07,
+      "loss": 0.3081,
+      "step": 13197
+    },
+    {
+      "epoch": 0.9705125376865946,
+      "grad_norm": 0.3912736475467682,
+      "learning_rate": 4.5796748898068173e-07,
+      "loss": 0.31,
+      "step": 13198
+    },
+    {
+      "epoch": 0.9705860725053312,
+      "grad_norm": 0.39628034830093384,
+      "learning_rate": 4.556936075326701e-07,
+      "loss": 0.2964,
+      "step": 13199
+    },
+    {
+      "epoch": 0.970659607324068,
+      "grad_norm": 0.40941041707992554,
+      "learning_rate": 4.5342537234277507e-07,
+      "loss": 0.3342,
+      "step": 13200
+    },
+    {
+      "epoch": 0.9707331421428046,
+      "grad_norm": 0.3891136348247528,
+      "learning_rate": 4.5116278353963813e-07,
+      "loss": 0.3289,
+      "step": 13201
+    },
+    {
+      "epoch": 0.9708066769615413,
+      "grad_norm": 0.3800886869430542,
+      "learning_rate": 4.4890584125161226e-07,
+      "loss": 0.3212,
+      "step": 13202
+    },
+    {
+      "epoch": 0.970880211780278,
+      "grad_norm": 0.4157721996307373,
+      "learning_rate": 4.4665454560670616e-07,
+      "loss": 0.3156,
+      "step": 13203
+    },
+    {
+      "epoch": 0.9709537465990147,
+      "grad_norm": 0.40718352794647217,
+      "learning_rate": 4.4440889673261765e-07,
+      "loss": 0.3266,
+      "step": 13204
+    },
+    {
+      "epoch": 0.9710272814177513,
+      "grad_norm": 0.35676732659339905,
+      "learning_rate": 4.421688947567226e-07,
+      "loss": 0.2904,
+      "step": 13205
+    },
+    {
+      "epoch": 0.971100816236488,
+      "grad_norm": 0.4543990194797516,
+      "learning_rate": 4.3993453980607504e-07,
+      "loss": 0.344,
+      "step": 13206
+    },
+    {
+      "epoch": 0.9711743510552246,
+      "grad_norm": 0.4223882853984833,
+      "learning_rate": 4.3770583200740676e-07,
+      "loss": 0.332,
+      "step": 13207
+    },
+    {
+      "epoch": 0.9712478858739613,
+      "grad_norm": 0.45391640067100525,
+      "learning_rate": 4.35482771487139e-07,
+      "loss": 0.3231,
+      "step": 13208
+    },
+    {
+      "epoch": 0.971321420692698,
+      "grad_norm": 0.4324050545692444,
+      "learning_rate": 4.3326535837135974e-07,
+      "loss": 0.3249,
+      "step": 13209
+    },
+    {
+      "epoch": 0.9713949555114346,
+      "grad_norm": 0.41551968455314636,
+      "learning_rate": 4.3105359278584614e-07,
+      "loss": 0.3031,
+      "step": 13210
+    },
+    {
+      "epoch": 0.9714684903301714,
+      "grad_norm": 0.8754096031188965,
+      "learning_rate": 4.2884747485605335e-07,
+      "loss": 0.3191,
+      "step": 13211
+    },
+    {
+      "epoch": 0.971542025148908,
+      "grad_norm": 0.3566398620605469,
+      "learning_rate": 4.266470047071258e-07,
+      "loss": 0.2957,
+      "step": 13212
+    },
+    {
+      "epoch": 0.9716155599676447,
+      "grad_norm": 0.40444210171699524,
+      "learning_rate": 4.2445218246385255e-07,
+      "loss": 0.338,
+      "step": 13213
+    },
+    {
+      "epoch": 0.9716890947863813,
+      "grad_norm": 0.4574260711669922,
+      "learning_rate": 4.2226300825073396e-07,
+      "loss": 0.3182,
+      "step": 13214
+    },
+    {
+      "epoch": 0.9717626296051181,
+      "grad_norm": 0.4421963691711426,
+      "learning_rate": 4.200794821919485e-07,
+      "loss": 0.3352,
+      "step": 13215
+    },
+    {
+      "epoch": 0.9718361644238547,
+      "grad_norm": 0.3370788097381592,
+      "learning_rate": 4.179016044113526e-07,
+      "loss": 0.3193,
+      "step": 13216
+    },
+    {
+      "epoch": 0.9719096992425914,
+      "grad_norm": 0.5109232664108276,
+      "learning_rate": 4.1572937503245866e-07,
+      "loss": 0.3093,
+      "step": 13217
+    },
+    {
+      "epoch": 0.971983234061328,
+      "grad_norm": 0.46007540822029114,
+      "learning_rate": 4.135627941785014e-07,
+      "loss": 0.3231,
+      "step": 13218
+    },
+    {
+      "epoch": 0.9720567688800648,
+      "grad_norm": 0.4246895909309387,
+      "learning_rate": 4.114018619723603e-07,
+      "loss": 0.3225,
+      "step": 13219
+    },
+    {
+      "epoch": 0.9721303036988014,
+      "grad_norm": 0.4317570924758911,
+      "learning_rate": 4.0924657853659285e-07,
+      "loss": 0.3306,
+      "step": 13220
+    },
+    {
+      "epoch": 0.972203838517538,
+      "grad_norm": 0.44649964570999146,
+      "learning_rate": 4.070969439934791e-07,
+      "loss": 0.3145,
+      "step": 13221
+    },
+    {
+      "epoch": 0.9722773733362747,
+      "grad_norm": 0.4299013912677765,
+      "learning_rate": 4.049529584649214e-07,
+      "loss": 0.3234,
+      "step": 13222
+    },
+    {
+      "epoch": 0.9723509081550114,
+      "grad_norm": 0.39917707443237305,
+      "learning_rate": 4.0281462207254485e-07,
+      "loss": 0.3264,
+      "step": 13223
+    },
+    {
+      "epoch": 0.9724244429737481,
+      "grad_norm": 0.3597029745578766,
+      "learning_rate": 4.00681934937619e-07,
+      "loss": 0.3253,
+      "step": 13224
+    },
+    {
+      "epoch": 0.9724979777924847,
+      "grad_norm": 0.4187258780002594,
+      "learning_rate": 3.9855489718113594e-07,
+      "loss": 0.3289,
+      "step": 13225
+    },
+    {
+      "epoch": 0.9725715126112214,
+      "grad_norm": 0.37095803022384644,
+      "learning_rate": 3.9643350892373255e-07,
+      "loss": 0.3158,
+      "step": 13226
+    },
+    {
+      "epoch": 0.972645047429958,
+      "grad_norm": 0.3526410460472107,
+      "learning_rate": 3.943177702857348e-07,
+      "loss": 0.2919,
+      "step": 13227
+    },
+    {
+      "epoch": 0.9727185822486948,
+      "grad_norm": 0.39344608783721924,
+      "learning_rate": 3.922076813871467e-07,
+      "loss": 0.313,
+      "step": 13228
+    },
+    {
+      "epoch": 0.9727921170674314,
+      "grad_norm": 0.35310715436935425,
+      "learning_rate": 3.901032423476614e-07,
+      "loss": 0.3264,
+      "step": 13229
+    },
+    {
+      "epoch": 0.9728656518861681,
+      "grad_norm": 0.3790527582168579,
+      "learning_rate": 3.880044532866389e-07,
+      "loss": 0.2975,
+      "step": 13230
+    },
+    {
+      "epoch": 0.9729391867049048,
+      "grad_norm": 0.39867085218429565,
+      "learning_rate": 3.859113143231397e-07,
+      "loss": 0.3008,
+      "step": 13231
+    },
+    {
+      "epoch": 0.9730127215236415,
+      "grad_norm": 0.42192062735557556,
+      "learning_rate": 3.838238255758797e-07,
+      "loss": 0.3074,
+      "step": 13232
+    },
+    {
+      "epoch": 0.9730862563423781,
+      "grad_norm": 0.451744943857193,
+      "learning_rate": 3.8174198716325325e-07,
+      "loss": 0.3296,
+      "step": 13233
+    },
+    {
+      "epoch": 0.9731597911611148,
+      "grad_norm": 0.3810254633426666,
+      "learning_rate": 3.7966579920334367e-07,
+      "loss": 0.3067,
+      "step": 13234
+    },
+    {
+      "epoch": 0.9732333259798515,
+      "grad_norm": 0.39745402336120605,
+      "learning_rate": 3.7759526181393445e-07,
+      "loss": 0.294,
+      "step": 13235
+    },
+    {
+      "epoch": 0.9733068607985881,
+      "grad_norm": 0.4675830602645874,
+      "learning_rate": 3.755303751124539e-07,
+      "loss": 0.3135,
+      "step": 13236
+    },
+    {
+      "epoch": 0.9733803956173248,
+      "grad_norm": 0.40875813364982605,
+      "learning_rate": 3.734711392160306e-07,
+      "loss": 0.2851,
+      "step": 13237
+    },
+    {
+      "epoch": 0.9734539304360614,
+      "grad_norm": 0.4639033079147339,
+      "learning_rate": 3.714175542414711e-07,
+      "loss": 0.3129,
+      "step": 13238
+    },
+    {
+      "epoch": 0.9735274652547982,
+      "grad_norm": 0.387595534324646,
+      "learning_rate": 3.693696203052377e-07,
+      "loss": 0.3209,
+      "step": 13239
+    },
+    {
+      "epoch": 0.9736010000735348,
+      "grad_norm": 0.37073585391044617,
+      "learning_rate": 3.6732733752350424e-07,
+      "loss": 0.3013,
+      "step": 13240
+    },
+    {
+      "epoch": 0.9736745348922715,
+      "grad_norm": 0.39574486017227173,
+      "learning_rate": 3.6529070601212245e-07,
+      "loss": 0.3104,
+      "step": 13241
+    },
+    {
+      "epoch": 0.9737480697110081,
+      "grad_norm": 0.45205792784690857,
+      "learning_rate": 3.6325972588658885e-07,
+      "loss": 0.3011,
+      "step": 13242
+    },
+    {
+      "epoch": 0.9738216045297449,
+      "grad_norm": 0.4067242741584778,
+      "learning_rate": 3.6123439726212237e-07,
+      "loss": 0.3345,
+      "step": 13243
+    },
+    {
+      "epoch": 0.9738951393484815,
+      "grad_norm": 0.4172176420688629,
+      "learning_rate": 3.592147202535867e-07,
+      "loss": 0.3017,
+      "step": 13244
+    },
+    {
+      "epoch": 0.9739686741672182,
+      "grad_norm": 0.4057348072528839,
+      "learning_rate": 3.572006949755569e-07,
+      "loss": 0.3254,
+      "step": 13245
+    },
+    {
+      "epoch": 0.9740422089859548,
+      "grad_norm": 0.39415615797042847,
+      "learning_rate": 3.551923215422526e-07,
+      "loss": 0.335,
+      "step": 13246
+    },
+    {
+      "epoch": 0.9741157438046916,
+      "grad_norm": 0.3723656237125397,
+      "learning_rate": 3.531896000676049e-07,
+      "loss": 0.3333,
+      "step": 13247
+    },
+    {
+      "epoch": 0.9741892786234282,
+      "grad_norm": 0.38290372490882874,
+      "learning_rate": 3.5119253066520084e-07,
+      "loss": 0.3484,
+      "step": 13248
+    },
+    {
+      "epoch": 0.9742628134421648,
+      "grad_norm": 0.4378306269645691,
+      "learning_rate": 3.492011134483164e-07,
+      "loss": 0.3025,
+      "step": 13249
+    },
+    {
+      "epoch": 0.9743363482609015,
+      "grad_norm": 0.40498703718185425,
+      "learning_rate": 3.4721534852991675e-07,
+      "loss": 0.298,
+      "step": 13250
+    },
+    {
+      "epoch": 0.9744098830796382,
+      "grad_norm": 0.4019177556037903,
+      "learning_rate": 3.4523523602263406e-07,
+      "loss": 0.307,
+      "step": 13251
+    },
+    {
+      "epoch": 0.9744834178983749,
+      "grad_norm": 0.3625493347644806,
+      "learning_rate": 3.432607760387674e-07,
+      "loss": 0.2984,
+      "step": 13252
+    },
+    {
+      "epoch": 0.9745569527171115,
+      "grad_norm": 0.42665204405784607,
+      "learning_rate": 3.412919686903271e-07,
+      "loss": 0.3285,
+      "step": 13253
+    },
+    {
+      "epoch": 0.9746304875358482,
+      "grad_norm": 0.37903639674186707,
+      "learning_rate": 3.3932881408897943e-07,
+      "loss": 0.2977,
+      "step": 13254
+    },
+    {
+      "epoch": 0.9747040223545849,
+      "grad_norm": 0.3726866543292999,
+      "learning_rate": 3.3737131234607977e-07,
+      "loss": 0.3061,
+      "step": 13255
+    },
+    {
+      "epoch": 0.9747775571733216,
+      "grad_norm": 0.3483209013938904,
+      "learning_rate": 3.3541946357266154e-07,
+      "loss": 0.3011,
+      "step": 13256
+    },
+    {
+      "epoch": 0.9748510919920582,
+      "grad_norm": 0.3744899034500122,
+      "learning_rate": 3.33473267879425e-07,
+      "loss": 0.2878,
+      "step": 13257
+    },
+    {
+      "epoch": 0.974924626810795,
+      "grad_norm": 0.35839542746543884,
+      "learning_rate": 3.3153272537677083e-07,
+      "loss": 0.3228,
+      "step": 13258
+    },
+    {
+      "epoch": 0.9749981616295316,
+      "grad_norm": 0.3603803515434265,
+      "learning_rate": 3.2959783617475536e-07,
+      "loss": 0.2932,
+      "step": 13259
+    },
+    {
+      "epoch": 0.9750716964482683,
+      "grad_norm": 0.4705696403980255,
+      "learning_rate": 3.276686003831464e-07,
+      "loss": 0.2922,
+      "step": 13260
+    },
+    {
+      "epoch": 0.9751452312670049,
+      "grad_norm": 0.4184643030166626,
+      "learning_rate": 3.257450181113564e-07,
+      "loss": 0.3171,
+      "step": 13261
+    },
+    {
+      "epoch": 0.9752187660857417,
+      "grad_norm": 0.3669772446155548,
+      "learning_rate": 3.2382708946850915e-07,
+      "loss": 0.3001,
+      "step": 13262
+    },
+    {
+      "epoch": 0.9752923009044783,
+      "grad_norm": 0.4430077075958252,
+      "learning_rate": 3.2191481456337325e-07,
+      "loss": 0.3236,
+      "step": 13263
+    },
+    {
+      "epoch": 0.9753658357232149,
+      "grad_norm": 0.4027765691280365,
+      "learning_rate": 3.200081935044286e-07,
+      "loss": 0.3155,
+      "step": 13264
+    },
+    {
+      "epoch": 0.9754393705419516,
+      "grad_norm": 0.4594901204109192,
+      "learning_rate": 3.1810722639981086e-07,
+      "loss": 0.3042,
+      "step": 13265
+    },
+    {
+      "epoch": 0.9755129053606882,
+      "grad_norm": 0.403331458568573,
+      "learning_rate": 3.162119133573449e-07,
+      "loss": 0.3144,
+      "step": 13266
+    },
+    {
+      "epoch": 0.975586440179425,
+      "grad_norm": 0.4548698961734772,
+      "learning_rate": 3.143222544845559e-07,
+      "loss": 0.3129,
+      "step": 13267
+    },
+    {
+      "epoch": 0.9756599749981616,
+      "grad_norm": 0.40541350841522217,
+      "learning_rate": 3.124382498886025e-07,
+      "loss": 0.3211,
+      "step": 13268
+    },
+    {
+      "epoch": 0.9757335098168983,
+      "grad_norm": 0.4357725977897644,
+      "learning_rate": 3.105598996763548e-07,
+      "loss": 0.3307,
+      "step": 13269
+    },
+    {
+      "epoch": 0.975807044635635,
+      "grad_norm": 0.3814126253128052,
+      "learning_rate": 3.086872039543498e-07,
+      "loss": 0.331,
+      "step": 13270
+    },
+    {
+      "epoch": 0.9758805794543717,
+      "grad_norm": 0.41433975100517273,
+      "learning_rate": 3.0682016282882475e-07,
+      "loss": 0.3319,
+      "step": 13271
+    },
+    {
+      "epoch": 0.9759541142731083,
+      "grad_norm": 0.40694090723991394,
+      "learning_rate": 3.049587764056727e-07,
+      "loss": 0.3256,
+      "step": 13272
+    },
+    {
+      "epoch": 0.976027649091845,
+      "grad_norm": 0.4249478578567505,
+      "learning_rate": 3.031030447904537e-07,
+      "loss": 0.3177,
+      "step": 13273
+    },
+    {
+      "epoch": 0.9761011839105816,
+      "grad_norm": 0.4258928894996643,
+      "learning_rate": 3.0125296808846126e-07,
+      "loss": 0.3163,
+      "step": 13274
+    },
+    {
+      "epoch": 0.9761747187293184,
+      "grad_norm": 0.3792356550693512,
+      "learning_rate": 2.9940854640461146e-07,
+      "loss": 0.3508,
+      "step": 13275
+    },
+    {
+      "epoch": 0.976248253548055,
+      "grad_norm": 0.3959629237651825,
+      "learning_rate": 2.975697798435317e-07,
+      "loss": 0.3093,
+      "step": 13276
+    },
+    {
+      "epoch": 0.9763217883667916,
+      "grad_norm": 0.3891172409057617,
+      "learning_rate": 2.957366685095053e-07,
+      "loss": 0.3351,
+      "step": 13277
+    },
+    {
+      "epoch": 0.9763953231855284,
+      "grad_norm": 0.42954710125923157,
+      "learning_rate": 2.9390921250652683e-07,
+      "loss": 0.284,
+      "step": 13278
+    },
+    {
+      "epoch": 0.976468858004265,
+      "grad_norm": 0.3787837326526642,
+      "learning_rate": 2.920874119382355e-07,
+      "loss": 0.3043,
+      "step": 13279
+    },
+    {
+      "epoch": 0.9765423928230017,
+      "grad_norm": 0.428617924451828,
+      "learning_rate": 2.9027126690797104e-07,
+      "loss": 0.3165,
+      "step": 13280
+    },
+    {
+      "epoch": 0.9766159276417383,
+      "grad_norm": 0.44265443086624146,
+      "learning_rate": 2.88460777518762e-07,
+      "loss": 0.3168,
+      "step": 13281
+    },
+    {
+      "epoch": 0.976689462460475,
+      "grad_norm": 0.3750629127025604,
+      "learning_rate": 2.8665594387327075e-07,
+      "loss": 0.3003,
+      "step": 13282
+    },
+    {
+      "epoch": 0.9767629972792117,
+      "grad_norm": 0.4582144021987915,
+      "learning_rate": 2.8485676607389324e-07,
+      "loss": 0.3087,
+      "step": 13283
+    },
+    {
+      "epoch": 0.9768365320979484,
+      "grad_norm": 0.3429451286792755,
+      "learning_rate": 2.8306324422266994e-07,
+      "loss": 0.3188,
+      "step": 13284
+    },
+    {
+      "epoch": 0.976910066916685,
+      "grad_norm": 0.39618197083473206,
+      "learning_rate": 2.8127537842133066e-07,
+      "loss": 0.2912,
+      "step": 13285
+    },
+    {
+      "epoch": 0.9769836017354218,
+      "grad_norm": 0.556599497795105,
+      "learning_rate": 2.7949316877129425e-07,
+      "loss": 0.2895,
+      "step": 13286
+    },
+    {
+      "epoch": 0.9770571365541584,
+      "grad_norm": 0.40513452887535095,
+      "learning_rate": 2.7771661537365766e-07,
+      "loss": 0.3083,
+      "step": 13287
+    },
+    {
+      "epoch": 0.9771306713728951,
+      "grad_norm": 0.4172392189502716,
+      "learning_rate": 2.759457183291625e-07,
+      "loss": 0.3034,
+      "step": 13288
+    },
+    {
+      "epoch": 0.9772042061916317,
+      "grad_norm": 0.3522256314754486,
+      "learning_rate": 2.741804777382617e-07,
+      "loss": 0.2833,
+      "step": 13289
+    },
+    {
+      "epoch": 0.9772777410103685,
+      "grad_norm": 0.42798054218292236,
+      "learning_rate": 2.7242089370110856e-07,
+      "loss": 0.3243,
+      "step": 13290
+    },
+    {
+      "epoch": 0.9773512758291051,
+      "grad_norm": 0.413610577583313,
+      "learning_rate": 2.706669663174788e-07,
+      "loss": 0.2776,
+      "step": 13291
+    },
+    {
+      "epoch": 0.9774248106478417,
+      "grad_norm": 0.39784663915634155,
+      "learning_rate": 2.6891869568685947e-07,
+      "loss": 0.3109,
+      "step": 13292
+    },
+    {
+      "epoch": 0.9774983454665784,
+      "grad_norm": 0.465675950050354,
+      "learning_rate": 2.671760819084268e-07,
+      "loss": 0.3822,
+      "step": 13293
+    },
+    {
+      "epoch": 0.977571880285315,
+      "grad_norm": 0.36191144585609436,
+      "learning_rate": 2.6543912508102397e-07,
+      "loss": 0.3091,
+      "step": 13294
+    },
+    {
+      "epoch": 0.9776454151040518,
+      "grad_norm": 0.41843825578689575,
+      "learning_rate": 2.6370782530317217e-07,
+      "loss": 0.3234,
+      "step": 13295
+    },
+    {
+      "epoch": 0.9777189499227884,
+      "grad_norm": 0.3765066862106323,
+      "learning_rate": 2.619821826730595e-07,
+      "loss": 0.3359,
+      "step": 13296
+    },
+    {
+      "epoch": 0.9777924847415251,
+      "grad_norm": 0.4386013448238373,
+      "learning_rate": 2.602621972885744e-07,
+      "loss": 0.3063,
+      "step": 13297
+    },
+    {
+      "epoch": 0.9778660195602618,
+      "grad_norm": 0.45064446330070496,
+      "learning_rate": 2.585478692472831e-07,
+      "loss": 0.3161,
+      "step": 13298
+    },
+    {
+      "epoch": 0.9779395543789985,
+      "grad_norm": 0.41554585099220276,
+      "learning_rate": 2.56839198646408e-07,
+      "loss": 0.2895,
+      "step": 13299
+    },
+    {
+      "epoch": 0.9780130891977351,
+      "grad_norm": 0.6622604131698608,
+      "learning_rate": 2.5513618558289374e-07,
+      "loss": 0.3095,
+      "step": 13300
+    },
+    {
+      "epoch": 0.9780866240164718,
+      "grad_norm": 0.6397599577903748,
+      "learning_rate": 2.5343883015330747e-07,
+      "loss": 0.2931,
+      "step": 13301
+    },
+    {
+      "epoch": 0.9781601588352085,
+      "grad_norm": 0.3932174742221832,
+      "learning_rate": 2.5174713245393887e-07,
+      "loss": 0.3029,
+      "step": 13302
+    },
+    {
+      "epoch": 0.9782336936539452,
+      "grad_norm": 0.390855610370636,
+      "learning_rate": 2.500610925807445e-07,
+      "loss": 0.3267,
+      "step": 13303
+    },
+    {
+      "epoch": 0.9783072284726818,
+      "grad_norm": 0.4197686016559601,
+      "learning_rate": 2.4838071062934785e-07,
+      "loss": 0.2961,
+      "step": 13304
+    },
+    {
+      "epoch": 0.9783807632914184,
+      "grad_norm": 0.38372647762298584,
+      "learning_rate": 2.4670598669508383e-07,
+      "loss": 0.3166,
+      "step": 13305
+    },
+    {
+      "epoch": 0.9784542981101552,
+      "grad_norm": 0.4579291343688965,
+      "learning_rate": 2.450369208729097e-07,
+      "loss": 0.3233,
+      "step": 13306
+    },
+    {
+      "epoch": 0.9785278329288918,
+      "grad_norm": 0.4121510982513428,
+      "learning_rate": 2.4337351325753875e-07,
+      "loss": 0.3077,
+      "step": 13307
+    },
+    {
+      "epoch": 0.9786013677476285,
+      "grad_norm": 0.39824968576431274,
+      "learning_rate": 2.417157639432843e-07,
+      "loss": 0.3247,
+      "step": 13308
+    },
+    {
+      "epoch": 0.9786749025663651,
+      "grad_norm": 0.416445791721344,
+      "learning_rate": 2.400636730241934e-07,
+      "loss": 0.2966,
+      "step": 13309
+    },
+    {
+      "epoch": 0.9787484373851019,
+      "grad_norm": 0.39202168583869934,
+      "learning_rate": 2.3841724059397996e-07,
+      "loss": 0.3605,
+      "step": 13310
+    },
+    {
+      "epoch": 0.9788219722038385,
+      "grad_norm": 0.4331412613391876,
+      "learning_rate": 2.3677646674601373e-07,
+      "loss": 0.3512,
+      "step": 13311
+    },
+    {
+      "epoch": 0.9788955070225752,
+      "grad_norm": 0.33475956320762634,
+      "learning_rate": 2.351413515733758e-07,
+      "loss": 0.3268,
+      "step": 13312
+    },
+    {
+      "epoch": 0.9789690418413118,
+      "grad_norm": 0.3826883137226105,
+      "learning_rate": 2.3351189516880313e-07,
+      "loss": 0.3181,
+      "step": 13313
+    },
+    {
+      "epoch": 0.9790425766600486,
+      "grad_norm": 0.3530716001987457,
+      "learning_rate": 2.3188809762473286e-07,
+      "loss": 0.3085,
+      "step": 13314
+    },
+    {
+      "epoch": 0.9791161114787852,
+      "grad_norm": 0.354184091091156,
+      "learning_rate": 2.3026995903324688e-07,
+      "loss": 0.3369,
+      "step": 13315
+    },
+    {
+      "epoch": 0.9791896462975219,
+      "grad_norm": 0.3948519229888916,
+      "learning_rate": 2.2865747948613846e-07,
+      "loss": 0.3107,
+      "step": 13316
+    },
+    {
+      "epoch": 0.9792631811162585,
+      "grad_norm": 0.35915955901145935,
+      "learning_rate": 2.2705065907487887e-07,
+      "loss": 0.3203,
+      "step": 13317
+    },
+    {
+      "epoch": 0.9793367159349953,
+      "grad_norm": 0.49189135432243347,
+      "learning_rate": 2.2544949789060633e-07,
+      "loss": 0.3352,
+      "step": 13318
+    },
+    {
+      "epoch": 0.9794102507537319,
+      "grad_norm": 0.42854177951812744,
+      "learning_rate": 2.2385399602412593e-07,
+      "loss": 0.3131,
+      "step": 13319
+    },
+    {
+      "epoch": 0.9794837855724685,
+      "grad_norm": 0.4045504331588745,
+      "learning_rate": 2.222641535659431e-07,
+      "loss": 0.2842,
+      "step": 13320
+    },
+    {
+      "epoch": 0.9795573203912052,
+      "grad_norm": 0.43975159525871277,
+      "learning_rate": 2.2067997060624124e-07,
+      "loss": 0.3181,
+      "step": 13321
+    },
+    {
+      "epoch": 0.9796308552099419,
+      "grad_norm": 0.35792359709739685,
+      "learning_rate": 2.191014472348707e-07,
+      "loss": 0.2978,
+      "step": 13322
+    },
+    {
+      "epoch": 0.9797043900286786,
+      "grad_norm": 0.4445738196372986,
+      "learning_rate": 2.1752858354137096e-07,
+      "loss": 0.287,
+      "step": 13323
+    },
+    {
+      "epoch": 0.9797779248474152,
+      "grad_norm": 0.4489285349845886,
+      "learning_rate": 2.1596137961494844e-07,
+      "loss": 0.3295,
+      "step": 13324
+    },
+    {
+      "epoch": 0.979851459666152,
+      "grad_norm": 0.4682174324989319,
+      "learning_rate": 2.143998355445098e-07,
+      "loss": 0.2948,
+      "step": 13325
+    },
+    {
+      "epoch": 0.9799249944848886,
+      "grad_norm": 0.4307081401348114,
+      "learning_rate": 2.1284395141861758e-07,
+      "loss": 0.3068,
+      "step": 13326
+    },
+    {
+      "epoch": 0.9799985293036253,
+      "grad_norm": 0.3717089593410492,
+      "learning_rate": 2.1129372732552334e-07,
+      "loss": 0.3039,
+      "step": 13327
+    },
+    {
+      "epoch": 0.9800720641223619,
+      "grad_norm": 0.41279175877571106,
+      "learning_rate": 2.0974916335316786e-07,
+      "loss": 0.3196,
+      "step": 13328
+    },
+    {
+      "epoch": 0.9801455989410987,
+      "grad_norm": 0.3622387945652008,
+      "learning_rate": 2.0821025958914775e-07,
+      "loss": 0.2929,
+      "step": 13329
+    },
+    {
+      "epoch": 0.9802191337598353,
+      "grad_norm": 0.40959998965263367,
+      "learning_rate": 2.0667701612075984e-07,
+      "loss": 0.319,
+      "step": 13330
+    },
+    {
+      "epoch": 0.980292668578572,
+      "grad_norm": 0.3871116638183594,
+      "learning_rate": 2.0514943303495683e-07,
+      "loss": 0.3364,
+      "step": 13331
+    },
+    {
+      "epoch": 0.9803662033973086,
+      "grad_norm": 0.42784228920936584,
+      "learning_rate": 2.0362751041840267e-07,
+      "loss": 0.3051,
+      "step": 13332
+    },
+    {
+      "epoch": 0.9804397382160452,
+      "grad_norm": 0.38907700777053833,
+      "learning_rate": 2.0211124835741723e-07,
+      "loss": 0.3169,
+      "step": 13333
+    },
+    {
+      "epoch": 0.980513273034782,
+      "grad_norm": 0.3545536994934082,
+      "learning_rate": 2.006006469379984e-07,
+      "loss": 0.2906,
+      "step": 13334
+    },
+    {
+      "epoch": 0.9805868078535186,
+      "grad_norm": 0.4177667200565338,
+      "learning_rate": 1.990957062458443e-07,
+      "loss": 0.3179,
+      "step": 13335
+    },
+    {
+      "epoch": 0.9806603426722553,
+      "grad_norm": 0.4153200387954712,
+      "learning_rate": 1.975964263662977e-07,
+      "loss": 0.3224,
+      "step": 13336
+    },
+    {
+      "epoch": 0.980733877490992,
+      "grad_norm": 0.49843767285346985,
+      "learning_rate": 1.961028073844018e-07,
+      "loss": 0.3152,
+      "step": 13337
+    },
+    {
+      "epoch": 0.9808074123097287,
+      "grad_norm": 0.46412214636802673,
+      "learning_rate": 1.946148493848887e-07,
+      "loss": 0.315,
+      "step": 13338
+    },
+    {
+      "epoch": 0.9808809471284653,
+      "grad_norm": 0.3971943259239197,
+      "learning_rate": 1.9313255245214656e-07,
+      "loss": 0.32,
+      "step": 13339
+    },
+    {
+      "epoch": 0.980954481947202,
+      "grad_norm": 0.4493296146392822,
+      "learning_rate": 1.9165591667026362e-07,
+      "loss": 0.3181,
+      "step": 13340
+    },
+    {
+      "epoch": 0.9810280167659386,
+      "grad_norm": 0.41779401898384094,
+      "learning_rate": 1.9018494212298398e-07,
+      "loss": 0.3094,
+      "step": 13341
+    },
+    {
+      "epoch": 0.9811015515846754,
+      "grad_norm": 0.3818504214286804,
+      "learning_rate": 1.8871962889375204e-07,
+      "loss": 0.3,
+      "step": 13342
+    },
+    {
+      "epoch": 0.981175086403412,
+      "grad_norm": 0.38807055354118347,
+      "learning_rate": 1.87259977065668e-07,
+      "loss": 0.3151,
+      "step": 13343
+    },
+    {
+      "epoch": 0.9812486212221487,
+      "grad_norm": 0.4557124972343445,
+      "learning_rate": 1.8580598672154337e-07,
+      "loss": 0.3367,
+      "step": 13344
+    },
+    {
+      "epoch": 0.9813221560408854,
+      "grad_norm": 0.362336665391922,
+      "learning_rate": 1.8435765794383443e-07,
+      "loss": 0.3098,
+      "step": 13345
+    },
+    {
+      "epoch": 0.9813956908596221,
+      "grad_norm": 0.3898058533668518,
+      "learning_rate": 1.8291499081470875e-07,
+      "loss": 0.3104,
+      "step": 13346
+    },
+    {
+      "epoch": 0.9814692256783587,
+      "grad_norm": 0.39874267578125,
+      "learning_rate": 1.814779854159787e-07,
+      "loss": 0.3321,
+      "step": 13347
+    },
+    {
+      "epoch": 0.9815427604970953,
+      "grad_norm": 0.4128715693950653,
+      "learning_rate": 1.8004664182915687e-07,
+      "loss": 0.3111,
+      "step": 13348
+    },
+    {
+      "epoch": 0.981616295315832,
+      "grad_norm": 0.40371403098106384,
+      "learning_rate": 1.7862096013544494e-07,
+      "loss": 0.2905,
+      "step": 13349
+    },
+    {
+      "epoch": 0.9816898301345687,
+      "grad_norm": 0.41202789545059204,
+      "learning_rate": 1.7720094041568935e-07,
+      "loss": 0.2872,
+      "step": 13350
+    },
+    {
+      "epoch": 0.9817633649533054,
+      "grad_norm": 0.42573025822639465,
+      "learning_rate": 1.757865827504479e-07,
+      "loss": 0.3161,
+      "step": 13351
+    },
+    {
+      "epoch": 0.981836899772042,
+      "grad_norm": 0.3777178227901459,
+      "learning_rate": 1.743778872199342e-07,
+      "loss": 0.3097,
+      "step": 13352
+    },
+    {
+      "epoch": 0.9819104345907788,
+      "grad_norm": 0.3748626112937927,
+      "learning_rate": 1.7297485390406208e-07,
+      "loss": 0.307,
+      "step": 13353
+    },
+    {
+      "epoch": 0.9819839694095154,
+      "grad_norm": 0.3479054868221283,
+      "learning_rate": 1.7157748288241238e-07,
+      "loss": 0.3551,
+      "step": 13354
+    },
+    {
+      "epoch": 0.9820575042282521,
+      "grad_norm": 0.42010635137557983,
+      "learning_rate": 1.7018577423423276e-07,
+      "loss": 0.3103,
+      "step": 13355
+    },
+    {
+      "epoch": 0.9821310390469887,
+      "grad_norm": 0.3308393657207489,
+      "learning_rate": 1.6879972803847123e-07,
+      "loss": 0.2749,
+      "step": 13356
+    },
+    {
+      "epoch": 0.9822045738657255,
+      "grad_norm": 0.505810022354126,
+      "learning_rate": 1.6741934437374262e-07,
+      "loss": 0.3038,
+      "step": 13357
+    },
+    {
+      "epoch": 0.9822781086844621,
+      "grad_norm": 0.4328267276287079,
+      "learning_rate": 1.6604462331835102e-07,
+      "loss": 0.2825,
+      "step": 13358
+    },
+    {
+      "epoch": 0.9823516435031988,
+      "grad_norm": 0.38691145181655884,
+      "learning_rate": 1.6467556495026738e-07,
+      "loss": 0.3228,
+      "step": 13359
+    },
+    {
+      "epoch": 0.9824251783219354,
+      "grad_norm": 0.4005323350429535,
+      "learning_rate": 1.6331216934714066e-07,
+      "loss": 0.3098,
+      "step": 13360
+    },
+    {
+      "epoch": 0.982498713140672,
+      "grad_norm": 0.40327513217926025,
+      "learning_rate": 1.6195443658632014e-07,
+      "loss": 0.3212,
+      "step": 13361
+    },
+    {
+      "epoch": 0.9825722479594088,
+      "grad_norm": 0.44518306851387024,
+      "learning_rate": 1.6060236674479978e-07,
+      "loss": 0.3336,
+      "step": 13362
+    },
+    {
+      "epoch": 0.9826457827781454,
+      "grad_norm": 0.3766849637031555,
+      "learning_rate": 1.592559598992849e-07,
+      "loss": 0.3264,
+      "step": 13363
+    },
+    {
+      "epoch": 0.9827193175968821,
+      "grad_norm": 0.38330772519111633,
+      "learning_rate": 1.5791521612612547e-07,
+      "loss": 0.3175,
+      "step": 13364
+    },
+    {
+      "epoch": 0.9827928524156188,
+      "grad_norm": 0.42017868161201477,
+      "learning_rate": 1.5658013550139406e-07,
+      "loss": 0.3083,
+      "step": 13365
+    },
+    {
+      "epoch": 0.9828663872343555,
+      "grad_norm": 0.40254634618759155,
+      "learning_rate": 1.5525071810080783e-07,
+      "loss": 0.2841,
+      "step": 13366
+    },
+    {
+      "epoch": 0.9829399220530921,
+      "grad_norm": 0.37005314230918884,
+      "learning_rate": 1.5392696399976204e-07,
+      "loss": 0.3168,
+      "step": 13367
+    },
+    {
+      "epoch": 0.9830134568718288,
+      "grad_norm": 0.3812549114227295,
+      "learning_rate": 1.5260887327335217e-07,
+      "loss": 0.3102,
+      "step": 13368
+    },
+    {
+      "epoch": 0.9830869916905655,
+      "grad_norm": 0.3847423791885376,
+      "learning_rate": 1.5129644599634064e-07,
+      "loss": 0.3146,
+      "step": 13369
+    },
+    {
+      "epoch": 0.9831605265093022,
+      "grad_norm": 0.38785871863365173,
+      "learning_rate": 1.499896822431679e-07,
+      "loss": 0.2926,
+      "step": 13370
+    },
+    {
+      "epoch": 0.9832340613280388,
+      "grad_norm": 0.3420933485031128,
+      "learning_rate": 1.4868858208796356e-07,
+      "loss": 0.2947,
+      "step": 13371
+    },
+    {
+      "epoch": 0.9833075961467755,
+      "grad_norm": 0.5454822182655334,
+      "learning_rate": 1.47393145604513e-07,
+      "loss": 0.3392,
+      "step": 13372
+    },
+    {
+      "epoch": 0.9833811309655122,
+      "grad_norm": 0.3614618480205536,
+      "learning_rate": 1.4610337286630194e-07,
+      "loss": 0.3108,
+      "step": 13373
+    },
+    {
+      "epoch": 0.9834546657842489,
+      "grad_norm": 0.3942137658596039,
+      "learning_rate": 1.4481926394649404e-07,
+      "loss": 0.3097,
+      "step": 13374
+    },
+    {
+      "epoch": 0.9835282006029855,
+      "grad_norm": 0.40737199783325195,
+      "learning_rate": 1.4354081891790883e-07,
+      "loss": 0.3199,
+      "step": 13375
+    },
+    {
+      "epoch": 0.9836017354217221,
+      "grad_norm": 0.47670987248420715,
+      "learning_rate": 1.422680378530883e-07,
+      "loss": 0.3344,
+      "step": 13376
+    },
+    {
+      "epoch": 0.9836752702404589,
+      "grad_norm": 0.38012346625328064,
+      "learning_rate": 1.4100092082419692e-07,
+      "loss": 0.3048,
+      "step": 13377
+    },
+    {
+      "epoch": 0.9837488050591955,
+      "grad_norm": 0.3848435580730438,
+      "learning_rate": 1.3973946790311056e-07,
+      "loss": 0.3106,
+      "step": 13378
+    },
+    {
+      "epoch": 0.9838223398779322,
+      "grad_norm": 0.5096490979194641,
+      "learning_rate": 1.3848367916140526e-07,
+      "loss": 0.2939,
+      "step": 13379
+    },
+    {
+      "epoch": 0.9838958746966688,
+      "grad_norm": 0.37088748812675476,
+      "learning_rate": 1.3723355467027966e-07,
+      "loss": 0.2846,
+      "step": 13380
+    },
+    {
+      "epoch": 0.9839694095154056,
+      "grad_norm": 0.3866788446903229,
+      "learning_rate": 1.3598909450066589e-07,
+      "loss": 0.3163,
+      "step": 13381
+    },
+    {
+      "epoch": 0.9840429443341422,
+      "grad_norm": 0.38097336888313293,
+      "learning_rate": 1.3475029872312972e-07,
+      "loss": 0.3079,
+      "step": 13382
+    },
+    {
+      "epoch": 0.9841164791528789,
+      "grad_norm": 0.39945322275161743,
+      "learning_rate": 1.3351716740795938e-07,
+      "loss": 0.3279,
+      "step": 13383
+    },
+    {
+      "epoch": 0.9841900139716155,
+      "grad_norm": 0.37602782249450684,
+      "learning_rate": 1.3228970062508784e-07,
+      "loss": 0.3377,
+      "step": 13384
+    },
+    {
+      "epoch": 0.9842635487903523,
+      "grad_norm": 0.3723997175693512,
+      "learning_rate": 1.3106789844413714e-07,
+      "loss": 0.327,
+      "step": 13385
+    },
+    {
+      "epoch": 0.9843370836090889,
+      "grad_norm": 0.39177432656288147,
+      "learning_rate": 1.2985176093440743e-07,
+      "loss": 0.3522,
+      "step": 13386
+    },
+    {
+      "epoch": 0.9844106184278256,
+      "grad_norm": 0.3805603086948395,
+      "learning_rate": 1.2864128816488797e-07,
+      "loss": 0.3025,
+      "step": 13387
+    },
+    {
+      "epoch": 0.9844841532465622,
+      "grad_norm": 0.38845205307006836,
+      "learning_rate": 1.2743648020423494e-07,
+      "loss": 0.3276,
+      "step": 13388
+    },
+    {
+      "epoch": 0.9845576880652989,
+      "grad_norm": 0.39451467990875244,
+      "learning_rate": 1.2623733712078256e-07,
+      "loss": 0.3308,
+      "step": 13389
+    },
+    {
+      "epoch": 0.9846312228840356,
+      "grad_norm": 0.39413589239120483,
+      "learning_rate": 1.250438589825431e-07,
+      "loss": 0.3292,
+      "step": 13390
+    },
+    {
+      "epoch": 0.9847047577027722,
+      "grad_norm": 0.36100077629089355,
+      "learning_rate": 1.23856045857218e-07,
+      "loss": 0.3354,
+      "step": 13391
+    },
+    {
+      "epoch": 0.984778292521509,
+      "grad_norm": 0.35953861474990845,
+      "learning_rate": 1.2267389781218662e-07,
+      "loss": 0.3159,
+      "step": 13392
+    },
+    {
+      "epoch": 0.9848518273402456,
+      "grad_norm": 0.4133627712726593,
+      "learning_rate": 1.214974149144954e-07,
+      "loss": 0.2997,
+      "step": 13393
+    },
+    {
+      "epoch": 0.9849253621589823,
+      "grad_norm": 0.4724199175834656,
+      "learning_rate": 1.203265972308798e-07,
+      "loss": 0.3178,
+      "step": 13394
+    },
+    {
+      "epoch": 0.9849988969777189,
+      "grad_norm": 0.4545726776123047,
+      "learning_rate": 1.1916144482775338e-07,
+      "loss": 0.321,
+      "step": 13395
+    },
+    {
+      "epoch": 0.9850724317964556,
+      "grad_norm": 0.4313446581363678,
+      "learning_rate": 1.180019577711855e-07,
+      "loss": 0.3373,
+      "step": 13396
+    },
+    {
+      "epoch": 0.9851459666151923,
+      "grad_norm": 0.3740943372249603,
+      "learning_rate": 1.1684813612695689e-07,
+      "loss": 0.3128,
+      "step": 13397
+    },
+    {
+      "epoch": 0.985219501433929,
+      "grad_norm": 0.379253625869751,
+      "learning_rate": 1.1569997996051518e-07,
+      "loss": 0.3142,
+      "step": 13398
+    },
+    {
+      "epoch": 0.9852930362526656,
+      "grad_norm": 0.36644595861434937,
+      "learning_rate": 1.1455748933698607e-07,
+      "loss": 0.3028,
+      "step": 13399
+    },
+    {
+      "epoch": 0.9853665710714024,
+      "grad_norm": 0.37736260890960693,
+      "learning_rate": 1.1342066432117326e-07,
+      "loss": 0.3047,
+      "step": 13400
+    },
+    {
+      "epoch": 0.985440105890139,
+      "grad_norm": 0.3841351866722107,
+      "learning_rate": 1.1228950497754742e-07,
+      "loss": 0.3143,
+      "step": 13401
+    },
+    {
+      "epoch": 0.9855136407088757,
+      "grad_norm": 0.3782109320163727,
+      "learning_rate": 1.1116401137027944e-07,
+      "loss": 0.3431,
+      "step": 13402
+    },
+    {
+      "epoch": 0.9855871755276123,
+      "grad_norm": 0.3690936863422394,
+      "learning_rate": 1.1004418356319601e-07,
+      "loss": 0.316,
+      "step": 13403
+    },
+    {
+      "epoch": 0.985660710346349,
+      "grad_norm": 0.39094769954681396,
+      "learning_rate": 1.0893002161983523e-07,
+      "loss": 0.3192,
+      "step": 13404
+    },
+    {
+      "epoch": 0.9857342451650857,
+      "grad_norm": 0.4279990494251251,
+      "learning_rate": 1.0782152560337988e-07,
+      "loss": 0.2965,
+      "step": 13405
+    },
+    {
+      "epoch": 0.9858077799838223,
+      "grad_norm": 0.40228745341300964,
+      "learning_rate": 1.0671869557671299e-07,
+      "loss": 0.3305,
+      "step": 13406
+    },
+    {
+      "epoch": 0.985881314802559,
+      "grad_norm": 0.3755859434604645,
+      "learning_rate": 1.0562153160237342e-07,
+      "loss": 0.3038,
+      "step": 13407
+    },
+    {
+      "epoch": 0.9859548496212956,
+      "grad_norm": 0.4035826623439789,
+      "learning_rate": 1.0453003374261138e-07,
+      "loss": 0.366,
+      "step": 13408
+    },
+    {
+      "epoch": 0.9860283844400324,
+      "grad_norm": 0.4489748775959015,
+      "learning_rate": 1.0344420205932182e-07,
+      "loss": 0.3315,
+      "step": 13409
+    },
+    {
+      "epoch": 0.986101919258769,
+      "grad_norm": 0.38207271695137024,
+      "learning_rate": 1.0236403661411098e-07,
+      "loss": 0.3006,
+      "step": 13410
+    },
+    {
+      "epoch": 0.9861754540775057,
+      "grad_norm": 0.38467422127723694,
+      "learning_rate": 1.0128953746822989e-07,
+      "loss": 0.3202,
+      "step": 13411
+    },
+    {
+      "epoch": 0.9862489888962424,
+      "grad_norm": 0.38417789340019226,
+      "learning_rate": 1.0022070468264089e-07,
+      "loss": 0.3482,
+      "step": 13412
+    },
+    {
+      "epoch": 0.9863225237149791,
+      "grad_norm": 0.38562220335006714,
+      "learning_rate": 9.915753831796215e-08,
+      "loss": 0.2878,
+      "step": 13413
+    },
+    {
+      "epoch": 0.9863960585337157,
+      "grad_norm": 0.40754348039627075,
+      "learning_rate": 9.81000384344899e-08,
+      "loss": 0.3105,
+      "step": 13414
+    },
+    {
+      "epoch": 0.9864695933524524,
+      "grad_norm": 0.43938061594963074,
+      "learning_rate": 9.704820509222057e-08,
+      "loss": 0.3406,
+      "step": 13415
+    },
+    {
+      "epoch": 0.986543128171189,
+      "grad_norm": 0.3898468017578125,
+      "learning_rate": 9.600203835080646e-08,
+      "loss": 0.3193,
+      "step": 13416
+    },
+    {
+      "epoch": 0.9866166629899257,
+      "grad_norm": 0.45510217547416687,
+      "learning_rate": 9.496153826958898e-08,
+      "loss": 0.342,
+      "step": 13417
+    },
+    {
+      "epoch": 0.9866901978086624,
+      "grad_norm": 0.40047600865364075,
+      "learning_rate": 9.392670490758759e-08,
+      "loss": 0.3149,
+      "step": 13418
+    },
+    {
+      "epoch": 0.986763732627399,
+      "grad_norm": 0.38864684104919434,
+      "learning_rate": 9.289753832349978e-08,
+      "loss": 0.3153,
+      "step": 13419
+    },
+    {
+      "epoch": 0.9868372674461358,
+      "grad_norm": 0.38454312086105347,
+      "learning_rate": 9.18740385757011e-08,
+      "loss": 0.3161,
+      "step": 13420
+    },
+    {
+      "epoch": 0.9869108022648724,
+      "grad_norm": 0.40521639585494995,
+      "learning_rate": 9.085620572224507e-08,
+      "loss": 0.321,
+      "step": 13421
+    },
+    {
+      "epoch": 0.9869843370836091,
+      "grad_norm": 0.36743563413619995,
+      "learning_rate": 8.984403982085221e-08,
+      "loss": 0.3398,
+      "step": 13422
+    },
+    {
+      "epoch": 0.9870578719023457,
+      "grad_norm": 0.37179479002952576,
+      "learning_rate": 8.883754092895435e-08,
+      "loss": 0.3123,
+      "step": 13423
+    },
+    {
+      "epoch": 0.9871314067210825,
+      "grad_norm": 0.3965074121952057,
+      "learning_rate": 8.783670910362807e-08,
+      "loss": 0.3012,
+      "step": 13424
+    },
+    {
+      "epoch": 0.9872049415398191,
+      "grad_norm": 0.39169609546661377,
+      "learning_rate": 8.684154440165015e-08,
+      "loss": 0.3262,
+      "step": 13425
+    },
+    {
+      "epoch": 0.9872784763585558,
+      "grad_norm": 0.37492749094963074,
+      "learning_rate": 8.585204687946436e-08,
+      "loss": 0.3122,
+      "step": 13426
+    },
+    {
+      "epoch": 0.9873520111772924,
+      "grad_norm": 0.37568196654319763,
+      "learning_rate": 8.486821659318133e-08,
+      "loss": 0.333,
+      "step": 13427
+    },
+    {
+      "epoch": 0.9874255459960292,
+      "grad_norm": 0.39522868394851685,
+      "learning_rate": 8.389005359863422e-08,
+      "loss": 0.3481,
+      "step": 13428
+    },
+    {
+      "epoch": 0.9874990808147658,
+      "grad_norm": 0.40709805488586426,
+      "learning_rate": 8.291755795127865e-08,
+      "loss": 0.3372,
+      "step": 13429
+    },
+    {
+      "epoch": 0.9875726156335025,
+      "grad_norm": 0.3688649535179138,
+      "learning_rate": 8.195072970628159e-08,
+      "loss": 0.3044,
+      "step": 13430
+    },
+    {
+      "epoch": 0.9876461504522391,
+      "grad_norm": 0.38678544759750366,
+      "learning_rate": 8.098956891849918e-08,
+      "loss": 0.308,
+      "step": 13431
+    },
+    {
+      "epoch": 0.9877196852709758,
+      "grad_norm": 0.3646492063999176,
+      "learning_rate": 8.003407564243227e-08,
+      "loss": 0.3295,
+      "step": 13432
+    },
+    {
+      "epoch": 0.9877932200897125,
+      "grad_norm": 1.8520188331604004,
+      "learning_rate": 7.908424993227081e-08,
+      "loss": 0.3243,
+      "step": 13433
+    },
+    {
+      "epoch": 0.9878667549084491,
+      "grad_norm": 0.4017718434333801,
+      "learning_rate": 7.814009184191618e-08,
+      "loss": 0.3257,
+      "step": 13434
+    },
+    {
+      "epoch": 0.9879402897271858,
+      "grad_norm": 0.3845238983631134,
+      "learning_rate": 7.720160142490329e-08,
+      "loss": 0.3183,
+      "step": 13435
+    },
+    {
+      "epoch": 0.9880138245459225,
+      "grad_norm": 0.40378352999687195,
+      "learning_rate": 7.626877873446736e-08,
+      "loss": 0.323,
+      "step": 13436
+    },
+    {
+      "epoch": 0.9880873593646592,
+      "grad_norm": 0.38670438528060913,
+      "learning_rate": 7.534162382352161e-08,
+      "loss": 0.3179,
+      "step": 13437
+    },
+    {
+      "epoch": 0.9881608941833958,
+      "grad_norm": 0.4048197269439697,
+      "learning_rate": 7.44201367446462e-08,
+      "loss": 0.3166,
+      "step": 13438
+    },
+    {
+      "epoch": 0.9882344290021325,
+      "grad_norm": 0.37783777713775635,
+      "learning_rate": 7.350431755012155e-08,
+      "loss": 0.3264,
+      "step": 13439
+    },
+    {
+      "epoch": 0.9883079638208692,
+      "grad_norm": 0.3996839225292206,
+      "learning_rate": 7.259416629189497e-08,
+      "loss": 0.3364,
+      "step": 13440
+    },
+    {
+      "epoch": 0.9883814986396059,
+      "grad_norm": 0.3749154508113861,
+      "learning_rate": 7.168968302158075e-08,
+      "loss": 0.3195,
+      "step": 13441
+    },
+    {
+      "epoch": 0.9884550334583425,
+      "grad_norm": 0.39702391624450684,
+      "learning_rate": 7.079086779048227e-08,
+      "loss": 0.3216,
+      "step": 13442
+    },
+    {
+      "epoch": 0.9885285682770792,
+      "grad_norm": 0.39695048332214355,
+      "learning_rate": 6.9897720649581e-08,
+      "loss": 0.2919,
+      "step": 13443
+    },
+    {
+      "epoch": 0.9886021030958159,
+      "grad_norm": 0.3931528329849243,
+      "learning_rate": 6.90102416495475e-08,
+      "loss": 0.3269,
+      "step": 13444
+    },
+    {
+      "epoch": 0.9886756379145525,
+      "grad_norm": 0.4535229206085205,
+      "learning_rate": 6.81284308407193e-08,
+      "loss": 0.3248,
+      "step": 13445
+    },
+    {
+      "epoch": 0.9887491727332892,
+      "grad_norm": 0.3861933648586273,
+      "learning_rate": 6.725228827310081e-08,
+      "loss": 0.3334,
+      "step": 13446
+    },
+    {
+      "epoch": 0.9888227075520258,
+      "grad_norm": 0.43108460307121277,
+      "learning_rate": 6.638181399640786e-08,
+      "loss": 0.3537,
+      "step": 13447
+    },
+    {
+      "epoch": 0.9888962423707626,
+      "grad_norm": 0.41773176193237305,
+      "learning_rate": 6.551700805998983e-08,
+      "loss": 0.2853,
+      "step": 13448
+    },
+    {
+      "epoch": 0.9889697771894992,
+      "grad_norm": 0.4270162284374237,
+      "learning_rate": 6.46578705129186e-08,
+      "loss": 0.3266,
+      "step": 13449
+    },
+    {
+      "epoch": 0.9890433120082359,
+      "grad_norm": 0.4955613911151886,
+      "learning_rate": 6.380440140391075e-08,
+      "loss": 0.3051,
+      "step": 13450
+    },
+    {
+      "epoch": 0.9891168468269725,
+      "grad_norm": 0.4400792717933655,
+      "learning_rate": 6.29566007813942e-08,
+      "loss": 0.315,
+      "step": 13451
+    },
+    {
+      "epoch": 0.9891903816457093,
+      "grad_norm": 0.4178203046321869,
+      "learning_rate": 6.211446869344161e-08,
+      "loss": 0.3191,
+      "step": 13452
+    },
+    {
+      "epoch": 0.9892639164644459,
+      "grad_norm": 0.3947291970252991,
+      "learning_rate": 6.12780051878259e-08,
+      "loss": 0.3215,
+      "step": 13453
+    },
+    {
+      "epoch": 0.9893374512831826,
+      "grad_norm": 0.3830108046531677,
+      "learning_rate": 6.044721031199796e-08,
+      "loss": 0.3256,
+      "step": 13454
+    },
+    {
+      "epoch": 0.9894109861019192,
+      "grad_norm": 0.4689227044582367,
+      "learning_rate": 5.962208411306458e-08,
+      "loss": 0.3396,
+      "step": 13455
+    },
+    {
+      "epoch": 0.989484520920656,
+      "grad_norm": 0.4624197483062744,
+      "learning_rate": 5.880262663784386e-08,
+      "loss": 0.3344,
+      "step": 13456
+    },
+    {
+      "epoch": 0.9895580557393926,
+      "grad_norm": 0.4126584827899933,
+      "learning_rate": 5.798883793280974e-08,
+      "loss": 0.3047,
+      "step": 13457
+    },
+    {
+      "epoch": 0.9896315905581293,
+      "grad_norm": 0.40418577194213867,
+      "learning_rate": 5.718071804411418e-08,
+      "loss": 0.3351,
+      "step": 13458
+    },
+    {
+      "epoch": 0.989705125376866,
+      "grad_norm": 0.3703249990940094,
+      "learning_rate": 5.63782670176094e-08,
+      "loss": 0.3385,
+      "step": 13459
+    },
+    {
+      "epoch": 0.9897786601956026,
+      "grad_norm": 0.40308892726898193,
+      "learning_rate": 5.558148489880344e-08,
+      "loss": 0.3192,
+      "step": 13460
+    },
+    {
+      "epoch": 0.9898521950143393,
+      "grad_norm": 0.3811427354812622,
+      "learning_rate": 5.4790371732893476e-08,
+      "loss": 0.306,
+      "step": 13461
+    },
+    {
+      "epoch": 0.9899257298330759,
+      "grad_norm": 0.4218481779098511,
+      "learning_rate": 5.4004927564743626e-08,
+      "loss": 0.3234,
+      "step": 13462
+    },
+    {
+      "epoch": 0.9899992646518126,
+      "grad_norm": 0.3659425377845764,
+      "learning_rate": 5.3225152438918236e-08,
+      "loss": 0.3075,
+      "step": 13463
+    },
+    {
+      "epoch": 0.9900727994705493,
+      "grad_norm": 0.36112290620803833,
+      "learning_rate": 5.2451046399637496e-08,
+      "loss": 0.3072,
+      "step": 13464
+    },
+    {
+      "epoch": 0.990146334289286,
+      "grad_norm": 0.426575630903244,
+      "learning_rate": 5.168260949081072e-08,
+      "loss": 0.3422,
+      "step": 13465
+    },
+    {
+      "epoch": 0.9902198691080226,
+      "grad_norm": 0.3862641453742981,
+      "learning_rate": 5.0919841756025267e-08,
+      "loss": 0.3313,
+      "step": 13466
+    },
+    {
+      "epoch": 0.9902934039267594,
+      "grad_norm": 0.4105272591114044,
+      "learning_rate": 5.0162743238546525e-08,
+      "loss": 0.3293,
+      "step": 13467
+    },
+    {
+      "epoch": 0.990366938745496,
+      "grad_norm": 0.3887922465801239,
+      "learning_rate": 4.941131398131793e-08,
+      "loss": 0.3044,
+      "step": 13468
+    },
+    {
+      "epoch": 0.9904404735642327,
+      "grad_norm": 0.40653538703918457,
+      "learning_rate": 4.866555402697204e-08,
+      "loss": 0.3155,
+      "step": 13469
+    },
+    {
+      "epoch": 0.9905140083829693,
+      "grad_norm": 0.4499475359916687,
+      "learning_rate": 4.792546341778614e-08,
+      "loss": 0.3042,
+      "step": 13470
+    },
+    {
+      "epoch": 0.9905875432017061,
+      "grad_norm": 0.38735559582710266,
+      "learning_rate": 4.7191042195759984e-08,
+      "loss": 0.3048,
+      "step": 13471
+    },
+    {
+      "epoch": 0.9906610780204427,
+      "grad_norm": 0.4033639430999756,
+      "learning_rate": 4.6462290402538024e-08,
+      "loss": 0.3291,
+      "step": 13472
+    },
+    {
+      "epoch": 0.9907346128391793,
+      "grad_norm": 0.39512816071510315,
+      "learning_rate": 4.5739208079453865e-08,
+      "loss": 0.3154,
+      "step": 13473
+    },
+    {
+      "epoch": 0.990808147657916,
+      "grad_norm": 0.38287925720214844,
+      "learning_rate": 4.5021795267541354e-08,
+      "loss": 0.3324,
+      "step": 13474
+    },
+    {
+      "epoch": 0.9908816824766526,
+      "grad_norm": 0.39967066049575806,
+      "learning_rate": 4.431005200746796e-08,
+      "loss": 0.3438,
+      "step": 13475
+    },
+    {
+      "epoch": 0.9909552172953894,
+      "grad_norm": 0.42425981163978577,
+      "learning_rate": 4.3603978339623595e-08,
+      "loss": 0.3171,
+      "step": 13476
+    },
+    {
+      "epoch": 0.991028752114126,
+      "grad_norm": 0.3731802999973297,
+      "learning_rate": 4.29035743040429e-08,
+      "loss": 0.3195,
+      "step": 13477
+    },
+    {
+      "epoch": 0.9911022869328627,
+      "grad_norm": 0.37033265829086304,
+      "learning_rate": 4.2208839940460763e-08,
+      "loss": 0.3348,
+      "step": 13478
+    },
+    {
+      "epoch": 0.9911758217515994,
+      "grad_norm": 0.4596717655658722,
+      "learning_rate": 4.1519775288290094e-08,
+      "loss": 0.3211,
+      "step": 13479
+    },
+    {
+      "epoch": 0.9912493565703361,
+      "grad_norm": 0.37420397996902466,
+      "learning_rate": 4.0836380386599646e-08,
+      "loss": 0.3179,
+      "step": 13480
+    },
+    {
+      "epoch": 0.9913228913890727,
+      "grad_norm": 0.34159940481185913,
+      "learning_rate": 4.015865527416951e-08,
+      "loss": 0.3127,
+      "step": 13481
+    },
+    {
+      "epoch": 0.9913964262078094,
+      "grad_norm": 0.38169148564338684,
+      "learning_rate": 3.94865999894245e-08,
+      "loss": 0.3082,
+      "step": 13482
+    },
+    {
+      "epoch": 0.991469961026546,
+      "grad_norm": 0.40326979756355286,
+      "learning_rate": 3.882021457050078e-08,
+      "loss": 0.3432,
+      "step": 13483
+    },
+    {
+      "epoch": 0.9915434958452828,
+      "grad_norm": 0.4049158990383148,
+      "learning_rate": 3.8159499055179235e-08,
+      "loss": 0.2905,
+      "step": 13484
+    },
+    {
+      "epoch": 0.9916170306640194,
+      "grad_norm": 0.44763219356536865,
+      "learning_rate": 3.750445348095211e-08,
+      "loss": 0.3173,
+      "step": 13485
+    },
+    {
+      "epoch": 0.9916905654827561,
+      "grad_norm": 0.46975305676460266,
+      "learning_rate": 3.685507788495635e-08,
+      "loss": 0.3137,
+      "step": 13486
+    },
+    {
+      "epoch": 0.9917641003014928,
+      "grad_norm": 0.3501894474029541,
+      "learning_rate": 3.621137230404026e-08,
+      "loss": 0.3123,
+      "step": 13487
+    },
+    {
+      "epoch": 0.9918376351202294,
+      "grad_norm": 0.44287195801734924,
+      "learning_rate": 3.557333677471908e-08,
+      "loss": 0.2853,
+      "step": 13488
+    },
+    {
+      "epoch": 0.9919111699389661,
+      "grad_norm": 0.39823055267333984,
+      "learning_rate": 3.4940971333163875e-08,
+      "loss": 0.2867,
+      "step": 13489
+    },
+    {
+      "epoch": 0.9919847047577027,
+      "grad_norm": 0.44509416818618774,
+      "learning_rate": 3.431427601525705e-08,
+      "loss": 0.3317,
+      "step": 13490
+    },
+    {
+      "epoch": 0.9920582395764395,
+      "grad_norm": 0.3792590796947479,
+      "learning_rate": 3.3693250856547956e-08,
+      "loss": 0.3119,
+      "step": 13491
+    },
+    {
+      "epoch": 0.9921317743951761,
+      "grad_norm": 0.5109067559242249,
+      "learning_rate": 3.307789589225285e-08,
+      "loss": 0.3193,
+      "step": 13492
+    },
+    {
+      "epoch": 0.9922053092139128,
+      "grad_norm": 0.37449702620506287,
+      "learning_rate": 3.2468211157277165e-08,
+      "loss": 0.3092,
+      "step": 13493
+    },
+    {
+      "epoch": 0.9922788440326494,
+      "grad_norm": 0.38684412837028503,
+      "learning_rate": 3.1864196686215434e-08,
+      "loss": 0.3368,
+      "step": 13494
+    },
+    {
+      "epoch": 0.9923523788513862,
+      "grad_norm": 0.4126976430416107,
+      "learning_rate": 3.126585251330694e-08,
+      "loss": 0.3363,
+      "step": 13495
+    },
+    {
+      "epoch": 0.9924259136701228,
+      "grad_norm": 0.4741186499595642,
+      "learning_rate": 3.06731786725023e-08,
+      "loss": 0.3556,
+      "step": 13496
+    },
+    {
+      "epoch": 0.9924994484888595,
+      "grad_norm": 0.40471118688583374,
+      "learning_rate": 3.008617519740797e-08,
+      "loss": 0.3279,
+      "step": 13497
+    },
+    {
+      "epoch": 0.9925729833075961,
+      "grad_norm": 0.3804692029953003,
+      "learning_rate": 2.9504842121341746e-08,
+      "loss": 0.3092,
+      "step": 13498
+    },
+    {
+      "epoch": 0.9926465181263329,
+      "grad_norm": 0.39251312613487244,
+      "learning_rate": 2.892917947725504e-08,
+      "loss": 0.3203,
+      "step": 13499
+    },
+    {
+      "epoch": 0.9927200529450695,
+      "grad_norm": 0.42377036809921265,
+      "learning_rate": 2.835918729781062e-08,
+      "loss": 0.2971,
+      "step": 13500
+    },
+    {
+      "epoch": 0.9927935877638061,
+      "grad_norm": 0.4212222099304199,
+      "learning_rate": 2.7794865615338174e-08,
+      "loss": 0.3082,
+      "step": 13501
+    },
+    {
+      "epoch": 0.9928671225825428,
+      "grad_norm": 0.3699338436126709,
+      "learning_rate": 2.7236214461845434e-08,
+      "loss": 0.3228,
+      "step": 13502
+    },
+    {
+      "epoch": 0.9929406574012795,
+      "grad_norm": 0.4136442244052887,
+      "learning_rate": 2.6683233869029268e-08,
+      "loss": 0.3187,
+      "step": 13503
+    },
+    {
+      "epoch": 0.9930141922200162,
+      "grad_norm": 0.42001351714134216,
+      "learning_rate": 2.6135923868231272e-08,
+      "loss": 0.3216,
+      "step": 13504
+    },
+    {
+      "epoch": 0.9930877270387528,
+      "grad_norm": 0.3836801052093506,
+      "learning_rate": 2.5594284490515485e-08,
+      "loss": 0.3044,
+      "step": 13505
+    },
+    {
+      "epoch": 0.9931612618574895,
+      "grad_norm": 0.38666409254074097,
+      "learning_rate": 2.5058315766601782e-08,
+      "loss": 0.3065,
+      "step": 13506
+    },
+    {
+      "epoch": 0.9932347966762262,
+      "grad_norm": 0.3897716701030731,
+      "learning_rate": 2.4528017726888065e-08,
+      "loss": 0.2884,
+      "step": 13507
+    },
+    {
+      "epoch": 0.9933083314949629,
+      "grad_norm": 0.3827457129955292,
+      "learning_rate": 2.4003390401450278e-08,
+      "loss": 0.3185,
+      "step": 13508
+    },
+    {
+      "epoch": 0.9933818663136995,
+      "grad_norm": 0.3972342610359192,
+      "learning_rate": 2.3484433820053498e-08,
+      "loss": 0.3365,
+      "step": 13509
+    },
+    {
+      "epoch": 0.9934554011324362,
+      "grad_norm": 0.374813437461853,
+      "learning_rate": 2.297114801212974e-08,
+      "loss": 0.3195,
+      "step": 13510
+    },
+    {
+      "epoch": 0.9935289359511729,
+      "grad_norm": 0.383487731218338,
+      "learning_rate": 2.246353300678905e-08,
+      "loss": 0.3106,
+      "step": 13511
+    },
+    {
+      "epoch": 0.9936024707699096,
+      "grad_norm": 0.4647825062274933,
+      "learning_rate": 2.1961588832830616e-08,
+      "loss": 0.2955,
+      "step": 13512
+    },
+    {
+      "epoch": 0.9936760055886462,
+      "grad_norm": 0.3471170961856842,
+      "learning_rate": 2.1465315518720553e-08,
+      "loss": 0.3201,
+      "step": 13513
+    },
+    {
+      "epoch": 0.993749540407383,
+      "grad_norm": 0.4170209765434265,
+      "learning_rate": 2.0974713092603014e-08,
+      "loss": 0.3136,
+      "step": 13514
+    },
+    {
+      "epoch": 0.9938230752261196,
+      "grad_norm": 0.47472915053367615,
+      "learning_rate": 2.04897815823224e-08,
+      "loss": 0.32,
+      "step": 13515
+    },
+    {
+      "epoch": 0.9938966100448562,
+      "grad_norm": 0.37100017070770264,
+      "learning_rate": 2.0010521015367824e-08,
+      "loss": 0.315,
+      "step": 13516
+    },
+    {
+      "epoch": 0.9939701448635929,
+      "grad_norm": 0.4099785089492798,
+      "learning_rate": 1.953693141892865e-08,
+      "loss": 0.2934,
+      "step": 13517
+    },
+    {
+      "epoch": 0.9940436796823295,
+      "grad_norm": 0.48907825350761414,
+      "learning_rate": 1.9069012819861177e-08,
+      "loss": 0.3539,
+      "step": 13518
+    },
+    {
+      "epoch": 0.9941172145010663,
+      "grad_norm": 0.35052943229675293,
+      "learning_rate": 1.8606765244721934e-08,
+      "loss": 0.2869,
+      "step": 13519
+    },
+    {
+      "epoch": 0.9941907493198029,
+      "grad_norm": 0.38984376192092896,
+      "learning_rate": 1.8150188719723294e-08,
+      "loss": 0.3124,
+      "step": 13520
+    },
+    {
+      "epoch": 0.9942642841385396,
+      "grad_norm": 0.4241129457950592,
+      "learning_rate": 1.7699283270766753e-08,
+      "loss": 0.2993,
+      "step": 13521
+    },
+    {
+      "epoch": 0.9943378189572762,
+      "grad_norm": 0.39513424038887024,
+      "learning_rate": 1.725404892340965e-08,
+      "loss": 0.3337,
+      "step": 13522
+    },
+    {
+      "epoch": 0.994411353776013,
+      "grad_norm": 0.42530038952827454,
+      "learning_rate": 1.6814485702920657e-08,
+      "loss": 0.3085,
+      "step": 13523
+    },
+    {
+      "epoch": 0.9944848885947496,
+      "grad_norm": 0.44088903069496155,
+      "learning_rate": 1.6380593634235387e-08,
+      "loss": 0.3094,
+      "step": 13524
+    },
+    {
+      "epoch": 0.9945584234134863,
+      "grad_norm": 0.35616907477378845,
+      "learning_rate": 1.5952372741956378e-08,
+      "loss": 0.3099,
+      "step": 13525
+    },
+    {
+      "epoch": 0.994631958232223,
+      "grad_norm": 0.38095149397850037,
+      "learning_rate": 1.5529823050375313e-08,
+      "loss": 0.3016,
+      "step": 13526
+    },
+    {
+      "epoch": 0.9947054930509597,
+      "grad_norm": 0.4793216288089752,
+      "learning_rate": 1.511294458347301e-08,
+      "loss": 0.3044,
+      "step": 13527
+    },
+    {
+      "epoch": 0.9947790278696963,
+      "grad_norm": 0.37746939063072205,
+      "learning_rate": 1.4701737364863911e-08,
+      "loss": 0.3144,
+      "step": 13528
+    },
+    {
+      "epoch": 0.9948525626884329,
+      "grad_norm": 0.5004724264144897,
+      "learning_rate": 1.4296201417907106e-08,
+      "loss": 0.3139,
+      "step": 13529
+    },
+    {
+      "epoch": 0.9949260975071696,
+      "grad_norm": 0.36428210139274597,
+      "learning_rate": 1.3896336765584217e-08,
+      "loss": 0.2822,
+      "step": 13530
+    },
+    {
+      "epoch": 0.9949996323259063,
+      "grad_norm": 0.37922272086143494,
+      "learning_rate": 1.3502143430577097e-08,
+      "loss": 0.3054,
+      "step": 13531
+    },
+    {
+      "epoch": 0.995073167144643,
+      "grad_norm": 0.4167078733444214,
+      "learning_rate": 1.3113621435245637e-08,
+      "loss": 0.3184,
+      "step": 13532
+    },
+    {
+      "epoch": 0.9951467019633796,
+      "grad_norm": 0.40363460779190063,
+      "learning_rate": 1.2730770801627767e-08,
+      "loss": 0.3102,
+      "step": 13533
+    },
+    {
+      "epoch": 0.9952202367821164,
+      "grad_norm": 0.3784312903881073,
+      "learning_rate": 1.235359155145055e-08,
+      "loss": 0.3048,
+      "step": 13534
+    },
+    {
+      "epoch": 0.995293771600853,
+      "grad_norm": 0.4381413459777832,
+      "learning_rate": 1.198208370609688e-08,
+      "loss": 0.3221,
+      "step": 13535
+    },
+    {
+      "epoch": 0.9953673064195897,
+      "grad_norm": 0.3733614981174469,
+      "learning_rate": 1.1616247286638792e-08,
+      "loss": 0.3125,
+      "step": 13536
+    },
+    {
+      "epoch": 0.9954408412383263,
+      "grad_norm": 0.3498537838459015,
+      "learning_rate": 1.1256082313826355e-08,
+      "loss": 0.3077,
+      "step": 13537
+    },
+    {
+      "epoch": 0.995514376057063,
+      "grad_norm": 0.4566275179386139,
+      "learning_rate": 1.0901588808098773e-08,
+      "loss": 0.3114,
+      "step": 13538
+    },
+    {
+      "epoch": 0.9955879108757997,
+      "grad_norm": 0.37820965051651,
+      "learning_rate": 1.0552766789551083e-08,
+      "loss": 0.3246,
+      "step": 13539
+    },
+    {
+      "epoch": 0.9956614456945364,
+      "grad_norm": 0.4071732461452484,
+      "learning_rate": 1.0209616277978561e-08,
+      "loss": 0.3243,
+      "step": 13540
+    },
+    {
+      "epoch": 0.995734980513273,
+      "grad_norm": 0.4355309009552002,
+      "learning_rate": 9.872137292832318e-09,
+      "loss": 0.3456,
+      "step": 13541
+    },
+    {
+      "epoch": 0.9958085153320098,
+      "grad_norm": 0.3738511800765991,
+      "learning_rate": 9.540329853263697e-09,
+      "loss": 0.3056,
+      "step": 13542
+    },
+    {
+      "epoch": 0.9958820501507464,
+      "grad_norm": 0.4524499475955963,
+      "learning_rate": 9.21419397809098e-09,
+      "loss": 0.3089,
+      "step": 13543
+    },
+    {
+      "epoch": 0.995955584969483,
+      "grad_norm": 0.38424646854400635,
+      "learning_rate": 8.893729685821583e-09,
+      "loss": 0.3105,
+      "step": 13544
+    },
+    {
+      "epoch": 0.9960291197882197,
+      "grad_norm": 0.3844616115093231,
+      "learning_rate": 8.578936994618758e-09,
+      "loss": 0.3054,
+      "step": 13545
+    },
+    {
+      "epoch": 0.9961026546069564,
+      "grad_norm": 0.41955727338790894,
+      "learning_rate": 8.269815922345991e-09,
+      "loss": 0.3065,
+      "step": 13546
+    },
+    {
+      "epoch": 0.9961761894256931,
+      "grad_norm": 0.3904637098312378,
+      "learning_rate": 7.966366486522603e-09,
+      "loss": 0.2951,
+      "step": 13547
+    },
+    {
+      "epoch": 0.9962497242444297,
+      "grad_norm": 0.3406813442707062,
+      "learning_rate": 7.668588704390356e-09,
+      "loss": 0.3153,
+      "step": 13548
+    },
+    {
+      "epoch": 0.9963232590631664,
+      "grad_norm": 0.3957519233226776,
+      "learning_rate": 7.3764825928135385e-09,
+      "loss": 0.3016,
+      "step": 13549
+    },
+    {
+      "epoch": 0.996396793881903,
+      "grad_norm": 0.37325143814086914,
+      "learning_rate": 7.09004816836778e-09,
+      "loss": 0.3142,
+      "step": 13550
+    },
+    {
+      "epoch": 0.9964703287006398,
+      "grad_norm": 0.34584668278694153,
+      "learning_rate": 6.809285447306745e-09,
+      "loss": 0.314,
+      "step": 13551
+    },
+    {
+      "epoch": 0.9965438635193764,
+      "grad_norm": 0.4025714695453644,
+      "learning_rate": 6.534194445539931e-09,
+      "loss": 0.3263,
+      "step": 13552
+    },
+    {
+      "epoch": 0.9966173983381131,
+      "grad_norm": 0.4254026710987091,
+      "learning_rate": 6.264775178688176e-09,
+      "loss": 0.3058,
+      "step": 13553
+    },
+    {
+      "epoch": 0.9966909331568498,
+      "grad_norm": 0.3662380278110504,
+      "learning_rate": 6.001027662028147e-09,
+      "loss": 0.2913,
+      "step": 13554
+    },
+    {
+      "epoch": 0.9967644679755865,
+      "grad_norm": 0.362680584192276,
+      "learning_rate": 5.742951910514549e-09,
+      "loss": 0.2789,
+      "step": 13555
+    },
+    {
+      "epoch": 0.9968380027943231,
+      "grad_norm": 0.37382733821868896,
+      "learning_rate": 5.490547938791224e-09,
+      "loss": 0.2788,
+      "step": 13556
+    },
+    {
+      "epoch": 0.9969115376130597,
+      "grad_norm": 0.42981797456741333,
+      "learning_rate": 5.243815761168946e-09,
+      "loss": 0.3126,
+      "step": 13557
+    },
+    {
+      "epoch": 0.9969850724317965,
+      "grad_norm": 0.4066285789012909,
+      "learning_rate": 5.002755391647629e-09,
+      "loss": 0.2897,
+      "step": 13558
+    },
+    {
+      "epoch": 0.9970586072505331,
+      "grad_norm": 0.40725177526474,
+      "learning_rate": 4.767366843905219e-09,
+      "loss": 0.3117,
+      "step": 13559
+    },
+    {
+      "epoch": 0.9971321420692698,
+      "grad_norm": 0.4016869068145752,
+      "learning_rate": 4.537650131275495e-09,
+      "loss": 0.2985,
+      "step": 13560
+    },
+    {
+      "epoch": 0.9972056768880064,
+      "grad_norm": 0.3793587386608124,
+      "learning_rate": 4.3136052668035774e-09,
+      "loss": 0.2982,
+      "step": 13561
+    },
+    {
+      "epoch": 0.9972792117067432,
+      "grad_norm": 0.36968162655830383,
+      "learning_rate": 4.095232263190418e-09,
+      "loss": 0.3157,
+      "step": 13562
+    },
+    {
+      "epoch": 0.9973527465254798,
+      "grad_norm": 0.4093943238258362,
+      "learning_rate": 3.8825311328372085e-09,
+      "loss": 0.3328,
+      "step": 13563
+    },
+    {
+      "epoch": 0.9974262813442165,
+      "grad_norm": 0.4410431385040283,
+      "learning_rate": 3.675501887789867e-09,
+      "loss": 0.2806,
+      "step": 13564
+    },
+    {
+      "epoch": 0.9974998161629531,
+      "grad_norm": 0.3838639557361603,
+      "learning_rate": 3.474144539794555e-09,
+      "loss": 0.2905,
+      "step": 13565
+    },
+    {
+      "epoch": 0.9975733509816899,
+      "grad_norm": 0.45033130049705505,
+      "learning_rate": 3.2784591002865684e-09,
+      "loss": 0.3106,
+      "step": 13566
+    },
+    {
+      "epoch": 0.9976468858004265,
+      "grad_norm": 0.3942461311817169,
+      "learning_rate": 3.0884455803459333e-09,
+      "loss": 0.2837,
+      "step": 13567
+    },
+    {
+      "epoch": 0.9977204206191632,
+      "grad_norm": 0.40582549571990967,
+      "learning_rate": 2.9041039907640176e-09,
+      "loss": 0.3469,
+      "step": 13568
+    },
+    {
+      "epoch": 0.9977939554378998,
+      "grad_norm": 0.41947853565216064,
+      "learning_rate": 2.7254343419880203e-09,
+      "loss": 0.3464,
+      "step": 13569
+    },
+    {
+      "epoch": 0.9978674902566366,
+      "grad_norm": 0.38435348868370056,
+      "learning_rate": 2.5524366441542767e-09,
+      "loss": 0.3541,
+      "step": 13570
+    },
+    {
+      "epoch": 0.9979410250753732,
+      "grad_norm": 0.39836567640304565,
+      "learning_rate": 2.3851109070882618e-09,
+      "loss": 0.3233,
+      "step": 13571
+    },
+    {
+      "epoch": 0.9980145598941098,
+      "grad_norm": 0.4241611659526825,
+      "learning_rate": 2.2234571402601766e-09,
+      "loss": 0.3104,
+      "step": 13572
+    },
+    {
+      "epoch": 0.9980880947128465,
+      "grad_norm": 0.37018710374832153,
+      "learning_rate": 2.0674753528515668e-09,
+      "loss": 0.282,
+      "step": 13573
+    },
+    {
+      "epoch": 0.9981616295315832,
+      "grad_norm": 0.39942678809165955,
+      "learning_rate": 1.917165553710909e-09,
+      "loss": 0.3095,
+      "step": 13574
+    },
+    {
+      "epoch": 0.9982351643503199,
+      "grad_norm": 0.3739287257194519,
+      "learning_rate": 1.7725277513647165e-09,
+      "loss": 0.3022,
+      "step": 13575
+    },
+    {
+      "epoch": 0.9983086991690565,
+      "grad_norm": 0.38510990142822266,
+      "learning_rate": 1.6335619540064351e-09,
+      "loss": 0.2844,
+      "step": 13576
+    },
+    {
+      "epoch": 0.9983822339877932,
+      "grad_norm": 0.3840303421020508,
+      "learning_rate": 1.5002681695297505e-09,
+      "loss": 0.3363,
+      "step": 13577
+    },
+    {
+      "epoch": 0.9984557688065299,
+      "grad_norm": 0.4786124527454376,
+      "learning_rate": 1.3726464054841792e-09,
+      "loss": 0.3065,
+      "step": 13578
+    },
+    {
+      "epoch": 0.9985293036252666,
+      "grad_norm": 0.4030840992927551,
+      "learning_rate": 1.2506966691194777e-09,
+      "loss": 0.3073,
+      "step": 13579
+    },
+    {
+      "epoch": 0.9986028384440032,
+      "grad_norm": 0.3808080852031708,
+      "learning_rate": 1.1344189673523353e-09,
+      "loss": 0.3106,
+      "step": 13580
+    },
+    {
+      "epoch": 0.99867637326274,
+      "grad_norm": 0.3861558139324188,
+      "learning_rate": 1.0238133067663747e-09,
+      "loss": 0.3016,
+      "step": 13581
+    },
+    {
+      "epoch": 0.9987499080814766,
+      "grad_norm": 0.3768991529941559,
+      "learning_rate": 9.18879693645458e-10,
+      "loss": 0.3233,
+      "step": 13582
+    },
+    {
+      "epoch": 0.9988234429002133,
+      "grad_norm": 0.377593070268631,
+      "learning_rate": 8.196181339403808e-10,
+      "loss": 0.3078,
+      "step": 13583
+    },
+    {
+      "epoch": 0.9988969777189499,
+      "grad_norm": 0.40440160036087036,
+      "learning_rate": 7.260286332799737e-10,
+      "loss": 0.2989,
+      "step": 13584
+    },
+    {
+      "epoch": 0.9989705125376865,
+      "grad_norm": 0.39562270045280457,
+      "learning_rate": 6.381111969711029e-10,
+      "loss": 0.2937,
+      "step": 13585
+    },
+    {
+      "epoch": 0.9990440473564233,
+      "grad_norm": 0.36248382925987244,
+      "learning_rate": 5.558658299986697e-10,
+      "loss": 0.2845,
+      "step": 13586
+    },
+    {
+      "epoch": 0.9991175821751599,
+      "grad_norm": 0.4101647734642029,
+      "learning_rate": 4.79292537036713e-10,
+      "loss": 0.3197,
+      "step": 13587
+    },
+    {
+      "epoch": 0.9991911169938966,
+      "grad_norm": 0.4572960436344147,
+      "learning_rate": 4.083913224151026e-10,
+      "loss": 0.2938,
+      "step": 13588
+    },
+    {
+      "epoch": 0.9992646518126332,
+      "grad_norm": 0.39929187297821045,
+      "learning_rate": 3.4316219017505036e-10,
+      "loss": 0.2959,
+      "step": 13589
+    },
+    {
+      "epoch": 0.99933818663137,
+      "grad_norm": 0.38032838702201843,
+      "learning_rate": 2.836051440024967e-10,
+      "loss": 0.316,
+      "step": 13590
+    },
+    {
+      "epoch": 0.9994117214501066,
+      "grad_norm": 0.36629095673561096,
+      "learning_rate": 2.2972018727251966e-10,
+      "loss": 0.2991,
+      "step": 13591
+    },
+    {
+      "epoch": 0.9994852562688433,
+      "grad_norm": 0.3724888265132904,
+      "learning_rate": 1.8150732304933472e-10,
+      "loss": 0.341,
+      "step": 13592
+    },
+    {
+      "epoch": 0.99955879108758,
+      "grad_norm": 0.36640429496765137,
+      "learning_rate": 1.389665540529883e-10,
+      "loss": 0.3275,
+      "step": 13593
+    },
+    {
+      "epoch": 0.9996323259063167,
+      "grad_norm": 0.38490229845046997,
+      "learning_rate": 1.0209788272597109e-10,
+      "loss": 0.2983,
+      "step": 13594
+    },
+    {
+      "epoch": 0.9997058607250533,
+      "grad_norm": 0.39137816429138184,
+      "learning_rate": 7.09013111332979e-11,
+      "loss": 0.285,
+      "step": 13595
+    },
+    {
+      "epoch": 0.99977939554379,
+      "grad_norm": 0.3737274408340454,
+      "learning_rate": 4.5376841051325557e-11,
+      "loss": 0.3109,
+      "step": 13596
+    },
+    {
+      "epoch": 0.9998529303625266,
+      "grad_norm": 0.36507630348205566,
+      "learning_rate": 2.5524473934446234e-11,
+      "loss": 0.3042,
+      "step": 13597
+    },
+    {
+      "epoch": 0.9999264651812634,
+      "grad_norm": 0.36330220103263855,
+      "learning_rate": 1.1344210903985186e-11,
+      "loss": 0.2921,
+      "step": 13598
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.410180002450943,
+      "learning_rate": 2.8360527704052175e-12,
+      "loss": 0.3049,
+      "step": 13599
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 13599,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.862831568852353e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}