diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,80521 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.733371596199222,
+  "eval_steps": 500,
+  "global_step": 11500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.25009894478923006,
+      "learning_rate": 2.1231422505307853e-09,
+      "loss": 0.1048,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.19922636436321503,
+      "learning_rate": 4.246284501061571e-09,
+      "loss": 0.1915,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.3269235595332411,
+      "learning_rate": 6.369426751592357e-09,
+      "loss": 0.2746,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.358077214427226,
+      "learning_rate": 8.492569002123141e-09,
+      "loss": 0.3747,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2540606765888093,
+      "learning_rate": 1.0615711252653927e-08,
+      "loss": 0.2772,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.28658288356106465,
+      "learning_rate": 1.2738853503184714e-08,
+      "loss": 0.349,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.685969966490638,
+      "learning_rate": 1.4861995753715499e-08,
+      "loss": 0.294,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7418569063982221,
+      "learning_rate": 1.6985138004246283e-08,
+      "loss": 0.2128,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.13356416195804294,
+      "learning_rate": 1.910828025477707e-08,
+      "loss": 0.1117,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.10029129219851991,
+      "learning_rate": 2.1231422505307853e-08,
+      "loss": 0.0753,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.19137471392482466,
+      "learning_rate": 2.3354564755838637e-08,
+      "loss": 0.2082,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.22620629379442936,
+      "learning_rate": 2.5477707006369427e-08,
+      "loss": 0.1041,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.24389827304886408,
+      "learning_rate": 2.760084925690021e-08,
+      "loss": 0.181,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.3382479000349999,
+      "learning_rate": 2.9723991507430998e-08,
+      "loss": 0.33,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.18771980115214004,
+      "learning_rate": 3.184713375796178e-08,
+      "loss": 0.0976,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.42516237338923024,
+      "learning_rate": 3.3970276008492565e-08,
+      "loss": 0.319,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.24522305917300966,
+      "learning_rate": 3.609341825902335e-08,
+      "loss": 0.4258,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.24229655895652444,
+      "learning_rate": 3.821656050955414e-08,
+      "loss": 0.2826,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.3229545378825092,
+      "learning_rate": 4.033970276008492e-08,
+      "loss": 0.286,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.16404456968742298,
+      "learning_rate": 4.2462845010615706e-08,
+      "loss": 0.028,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.16530036487321828,
+      "learning_rate": 4.458598726114649e-08,
+      "loss": 0.1198,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2385576349885164,
+      "learning_rate": 4.6709129511677274e-08,
+      "loss": 0.1685,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.19659189540970257,
+      "learning_rate": 4.883227176220807e-08,
+      "loss": 0.1903,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2947538234856561,
+      "learning_rate": 5.0955414012738854e-08,
+      "loss": 0.1341,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.23116107483481232,
+      "learning_rate": 5.307855626326964e-08,
+      "loss": 0.209,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1033019559393957,
+      "learning_rate": 5.520169851380042e-08,
+      "loss": 0.1789,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.08804539463365178,
+      "learning_rate": 5.732484076433121e-08,
+      "loss": 0.0189,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2914851513918007,
+      "learning_rate": 5.9447983014861996e-08,
+      "loss": 0.1725,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.14836108588407937,
+      "learning_rate": 6.157112526539278e-08,
+      "loss": 0.2839,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.1919117139250952,
+      "learning_rate": 6.369426751592356e-08,
+      "loss": 0.1446,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2761565442014638,
+      "learning_rate": 6.581740976645435e-08,
+      "loss": 0.2991,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.22305278681009405,
+      "learning_rate": 6.794055201698513e-08,
+      "loss": 0.0708,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.1610106563978104,
+      "learning_rate": 7.006369426751591e-08,
+      "loss": 0.261,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.14077057547983793,
+      "learning_rate": 7.21868365180467e-08,
+      "loss": 0.1801,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.20065851726476622,
+      "learning_rate": 7.43099787685775e-08,
+      "loss": 0.2476,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.3081035531257673,
+      "learning_rate": 7.643312101910828e-08,
+      "loss": 0.216,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.09010003471735234,
+      "learning_rate": 7.855626326963906e-08,
+      "loss": 0.0802,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.19947396878358303,
+      "learning_rate": 8.067940552016985e-08,
+      "loss": 0.305,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.262252267634813,
+      "learning_rate": 8.280254777070063e-08,
+      "loss": 0.1195,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.17317934071438829,
+      "learning_rate": 8.492569002123141e-08,
+      "loss": 0.0463,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.22859501169296872,
+      "learning_rate": 8.70488322717622e-08,
+      "loss": 0.1598,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.26001327021273624,
+      "learning_rate": 8.917197452229298e-08,
+      "loss": 0.066,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.1198363946982399,
+      "learning_rate": 9.129511677282376e-08,
+      "loss": 0.0997,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2593161588612782,
+      "learning_rate": 9.341825902335455e-08,
+      "loss": 0.0994,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2306799823690304,
+      "learning_rate": 9.554140127388536e-08,
+      "loss": 0.2179,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.17571689838456886,
+      "learning_rate": 9.766454352441614e-08,
+      "loss": 0.1061,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.17914073942300426,
+      "learning_rate": 9.978768577494693e-08,
+      "loss": 0.223,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.24889993931575963,
+      "learning_rate": 1.0191082802547771e-07,
+      "loss": 0.2474,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.22781020002071803,
+      "learning_rate": 1.0403397027600849e-07,
+      "loss": 0.11,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2501869999590839,
+      "learning_rate": 1.0615711252653928e-07,
+      "loss": 0.176,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.36315613525619006,
+      "learning_rate": 1.0828025477707006e-07,
+      "loss": 0.1577,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2658342666766177,
+      "learning_rate": 1.1040339702760084e-07,
+      "loss": 0.1983,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.18959948719710448,
+      "learning_rate": 1.1252653927813163e-07,
+      "loss": 0.1979,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2601087356669149,
+      "learning_rate": 1.1464968152866242e-07,
+      "loss": 0.133,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.31557862443430856,
+      "learning_rate": 1.1677282377919321e-07,
+      "loss": 0.3968,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.19390524170160062,
+      "learning_rate": 1.1889596602972399e-07,
+      "loss": 0.2072,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.1998692909914467,
+      "learning_rate": 1.2101910828025477e-07,
+      "loss": 0.1156,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.1221613667000933,
+      "learning_rate": 1.2314225053078556e-07,
+      "loss": 0.2259,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.23782027171460882,
+      "learning_rate": 1.2526539278131634e-07,
+      "loss": 0.1484,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2328260333787332,
+      "learning_rate": 1.2738853503184713e-07,
+      "loss": 0.288,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.14448131959238655,
+      "learning_rate": 1.295116772823779e-07,
+      "loss": 0.0096,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.332046819939343,
+      "learning_rate": 1.316348195329087e-07,
+      "loss": 0.0806,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2544066675974567,
+      "learning_rate": 1.3375796178343948e-07,
+      "loss": 0.2087,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.10688903983575725,
+      "learning_rate": 1.3588110403397026e-07,
+      "loss": 0.0316,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.15536549771069488,
+      "learning_rate": 1.3800424628450104e-07,
+      "loss": 0.1157,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.18711607110158102,
+      "learning_rate": 1.4012738853503183e-07,
+      "loss": 0.3105,
+      "step": 66
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2615454076682301,
+      "learning_rate": 1.422505307855626e-07,
+      "loss": 0.1635,
+      "step": 67
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.13564631530215016,
+      "learning_rate": 1.443736730360934e-07,
+      "loss": 0.1035,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.21407966082271274,
+      "learning_rate": 1.464968152866242e-07,
+      "loss": 0.119,
+      "step": 69
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.30779767305709155,
+      "learning_rate": 1.48619957537155e-07,
+      "loss": 0.2524,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.33050380029323323,
+      "learning_rate": 1.5074309978768577e-07,
+      "loss": 0.2835,
+      "step": 71
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2502548847332565,
+      "learning_rate": 1.5286624203821656e-07,
+      "loss": 0.1124,
+      "step": 72
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2553321164753365,
+      "learning_rate": 1.5498938428874734e-07,
+      "loss": 0.2557,
+      "step": 73
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.18679106407193102,
+      "learning_rate": 1.5711252653927812e-07,
+      "loss": 0.1674,
+      "step": 74
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.1283137929571681,
+      "learning_rate": 1.592356687898089e-07,
+      "loss": 0.3204,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.24057748518138783,
+      "learning_rate": 1.613588110403397e-07,
+      "loss": 0.1441,
+      "step": 76
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.21625351951617686,
+      "learning_rate": 1.6348195329087047e-07,
+      "loss": 0.1564,
+      "step": 77
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2807641314251535,
+      "learning_rate": 1.6560509554140126e-07,
+      "loss": 0.2258,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.21426607779258156,
+      "learning_rate": 1.6772823779193204e-07,
+      "loss": 0.0787,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.21735166406748668,
+      "learning_rate": 1.6985138004246283e-07,
+      "loss": 0.1448,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.09655181133610861,
+      "learning_rate": 1.719745222929936e-07,
+      "loss": 0.1662,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1774750298108412,
+      "learning_rate": 1.740976645435244e-07,
+      "loss": 0.1068,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9165863046549484,
+      "learning_rate": 1.7622080679405518e-07,
+      "loss": 0.2897,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.09591992055241581,
+      "learning_rate": 1.7834394904458596e-07,
+      "loss": 0.1542,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1871035981427734,
+      "learning_rate": 1.8046709129511674e-07,
+      "loss": 0.0815,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3013920904314661,
+      "learning_rate": 1.8259023354564753e-07,
+      "loss": 0.3896,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.08166990499696496,
+      "learning_rate": 1.847133757961783e-07,
+      "loss": 0.1132,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.24959187128555374,
+      "learning_rate": 1.868365180467091e-07,
+      "loss": 0.1128,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.11530906694869626,
+      "learning_rate": 1.8895966029723993e-07,
+      "loss": 0.1156,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.284104155324781,
+      "learning_rate": 1.9108280254777072e-07,
+      "loss": 0.2533,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2161792934914752,
+      "learning_rate": 1.932059447983015e-07,
+      "loss": 0.0977,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.16412721895417143,
+      "learning_rate": 1.9532908704883228e-07,
+      "loss": 0.2882,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.21110753539717453,
+      "learning_rate": 1.9745222929936307e-07,
+      "loss": 0.0277,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26361369211501084,
+      "learning_rate": 1.9957537154989385e-07,
+      "loss": 0.2768,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2529278381367476,
+      "learning_rate": 2.0169851380042463e-07,
+      "loss": 0.1897,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.16391164203616654,
+      "learning_rate": 2.0382165605095542e-07,
+      "loss": 0.1368,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2820574987511089,
+      "learning_rate": 2.059447983014862e-07,
+      "loss": 0.1703,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.11268193516563275,
+      "learning_rate": 2.0806794055201698e-07,
+      "loss": 0.2199,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.287806588908634,
+      "learning_rate": 2.1019108280254777e-07,
+      "loss": 0.0926,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2822299996909002,
+      "learning_rate": 2.1231422505307855e-07,
+      "loss": 0.1967,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2393906685301478,
+      "learning_rate": 2.1443736730360934e-07,
+      "loss": 0.398,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.21571853982926306,
+      "learning_rate": 2.1656050955414012e-07,
+      "loss": 0.2037,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.19160969775378794,
+      "learning_rate": 2.186836518046709e-07,
+      "loss": 0.3974,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2942322069415299,
+      "learning_rate": 2.208067940552017e-07,
+      "loss": 0.3434,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.25922201553758256,
+      "learning_rate": 2.2292993630573247e-07,
+      "loss": 0.1251,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2071198246502389,
+      "learning_rate": 2.2505307855626325e-07,
+      "loss": 0.1807,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1794504223080536,
+      "learning_rate": 2.2717622080679404e-07,
+      "loss": 0.1192,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3541472935508919,
+      "learning_rate": 2.2929936305732485e-07,
+      "loss": 0.1795,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.42212667279454036,
+      "learning_rate": 2.3142250530785563e-07,
+      "loss": 0.2508,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.15008163868349053,
+      "learning_rate": 2.3354564755838642e-07,
+      "loss": 0.1445,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2883339348097026,
+      "learning_rate": 2.356687898089172e-07,
+      "loss": 0.1372,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2689503988887403,
+      "learning_rate": 2.3779193205944798e-07,
+      "loss": 0.2331,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3632782308638637,
+      "learning_rate": 2.3991507430997877e-07,
+      "loss": 0.3856,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.386197873074015,
+      "learning_rate": 2.4203821656050955e-07,
+      "loss": 0.0899,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.09151484128845049,
+      "learning_rate": 2.4416135881104033e-07,
+      "loss": 0.0109,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.24683497049830289,
+      "learning_rate": 2.462845010615711e-07,
+      "loss": 0.1895,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2510248585389427,
+      "learning_rate": 2.484076433121019e-07,
+      "loss": 0.1953,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.31437307952859195,
+      "learning_rate": 2.505307855626327e-07,
+      "loss": 0.2637,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5792771847840741,
+      "learning_rate": 2.5265392781316347e-07,
+      "loss": 0.2178,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3569547700025303,
+      "learning_rate": 2.5477707006369425e-07,
+      "loss": 0.2827,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3541865319331499,
+      "learning_rate": 2.5690021231422504e-07,
+      "loss": 0.2158,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2803812751973915,
+      "learning_rate": 2.590233545647558e-07,
+      "loss": 0.1308,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3733045506699708,
+      "learning_rate": 2.611464968152866e-07,
+      "loss": 0.1748,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2627044050122649,
+      "learning_rate": 2.632696390658174e-07,
+      "loss": 0.1862,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1801284370762871,
+      "learning_rate": 2.6539278131634817e-07,
+      "loss": 0.2093,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.14356965529788235,
+      "learning_rate": 2.6751592356687895e-07,
+      "loss": 0.1077,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1448176733758346,
+      "learning_rate": 2.6963906581740974e-07,
+      "loss": 0.2931,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4005460509049744,
+      "learning_rate": 2.717622080679405e-07,
+      "loss": 0.1749,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.22251676974314188,
+      "learning_rate": 2.738853503184713e-07,
+      "loss": 0.1389,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3027281145545714,
+      "learning_rate": 2.760084925690021e-07,
+      "loss": 0.1792,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.23846282537302702,
+      "learning_rate": 2.7813163481953287e-07,
+      "loss": 0.3108,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2564118214473458,
+      "learning_rate": 2.8025477707006366e-07,
+      "loss": 0.2239,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.17751117550418236,
+      "learning_rate": 2.8237791932059444e-07,
+      "loss": 0.1463,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.25702667079807706,
+      "learning_rate": 2.845010615711252e-07,
+      "loss": 0.3392,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.18390417299723882,
+      "learning_rate": 2.86624203821656e-07,
+      "loss": 0.101,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2256740628934894,
+      "learning_rate": 2.887473460721868e-07,
+      "loss": 0.2245,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.22750489451154088,
+      "learning_rate": 2.908704883227176e-07,
+      "loss": 0.3181,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2252501766471252,
+      "learning_rate": 2.929936305732484e-07,
+      "loss": 0.1086,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26754928057573385,
+      "learning_rate": 2.951167728237792e-07,
+      "loss": 0.3815,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3578695711287303,
+      "learning_rate": 2.9723991507431e-07,
+      "loss": 0.3195,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.166913223467953,
+      "learning_rate": 2.9936305732484076e-07,
+      "loss": 0.1541,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.22575856821984214,
+      "learning_rate": 3.0148619957537155e-07,
+      "loss": 0.1101,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1816892201464045,
+      "learning_rate": 3.0360934182590233e-07,
+      "loss": 0.243,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.24199343241011556,
+      "learning_rate": 3.057324840764331e-07,
+      "loss": 0.0549,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3745084458743019,
+      "learning_rate": 3.078556263269639e-07,
+      "loss": 0.0462,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1060462353898553,
+      "learning_rate": 3.099787685774947e-07,
+      "loss": 0.1055,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.33017772424792025,
+      "learning_rate": 3.1210191082802546e-07,
+      "loss": 0.3886,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.13086145221413323,
+      "learning_rate": 3.1422505307855625e-07,
+      "loss": 0.1973,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26957199286967665,
+      "learning_rate": 3.1634819532908703e-07,
+      "loss": 0.3495,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.25739770577352045,
+      "learning_rate": 3.184713375796178e-07,
+      "loss": 0.1319,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26789222142663155,
+      "learning_rate": 3.205944798301486e-07,
+      "loss": 0.3428,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.09704698154957962,
+      "learning_rate": 3.227176220806794e-07,
+      "loss": 0.1104,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.29446869545909066,
+      "learning_rate": 3.2484076433121017e-07,
+      "loss": 0.2044,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.24583974076593526,
+      "learning_rate": 3.2696390658174095e-07,
+      "loss": 0.143,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.21672372184958014,
+      "learning_rate": 3.2908704883227173e-07,
+      "loss": 0.2422,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2316459136996034,
+      "learning_rate": 3.312101910828025e-07,
+      "loss": 0.0438,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.20623875642113082,
+      "learning_rate": 3.333333333333333e-07,
+      "loss": 0.1578,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4061097538574797,
+      "learning_rate": 3.354564755838641e-07,
+      "loss": 0.3662,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.14448612182415435,
+      "learning_rate": 3.3757961783439487e-07,
+      "loss": 0.0628,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.39492608570239246,
+      "learning_rate": 3.3970276008492565e-07,
+      "loss": 0.2601,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.25801533070450783,
+      "learning_rate": 3.4182590233545644e-07,
+      "loss": 0.1741,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.14966054415149585,
+      "learning_rate": 3.439490445859872e-07,
+      "loss": 0.1854,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.253115839014238,
+      "learning_rate": 3.46072186836518e-07,
+      "loss": 0.2261,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.34562150869603286,
+      "learning_rate": 3.481953290870488e-07,
+      "loss": 0.09,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.23803559936593066,
+      "learning_rate": 3.5031847133757957e-07,
+      "loss": 0.1875,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3221505541878738,
+      "learning_rate": 3.5244161358811035e-07,
+      "loss": 0.1857,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.12490966501456342,
+      "learning_rate": 3.5456475583864114e-07,
+      "loss": 0.0353,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.18338891854904169,
+      "learning_rate": 3.566878980891719e-07,
+      "loss": 0.0618,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.17791003089439517,
+      "learning_rate": 3.588110403397027e-07,
+      "loss": 0.3212,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.21835973917262472,
+      "learning_rate": 3.609341825902335e-07,
+      "loss": 0.2773,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.21765605514455302,
+      "learning_rate": 3.6305732484076427e-07,
+      "loss": 0.1221,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.21150022657865714,
+      "learning_rate": 3.6518046709129506e-07,
+      "loss": 0.3392,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.17460086087779267,
+      "learning_rate": 3.6730360934182584e-07,
+      "loss": 0.1001,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2999670269163138,
+      "learning_rate": 3.694267515923566e-07,
+      "loss": 0.0441,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.25135495215228293,
+      "learning_rate": 3.715498938428874e-07,
+      "loss": 0.2128,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6471046329195711,
+      "learning_rate": 3.736730360934182e-07,
+      "loss": 0.1948,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2278373196163679,
+      "learning_rate": 3.757961783439491e-07,
+      "loss": 0.2304,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2758925917903719,
+      "learning_rate": 3.7791932059447986e-07,
+      "loss": 0.4609,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.21079555432541505,
+      "learning_rate": 3.8004246284501065e-07,
+      "loss": 0.2711,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.22056801853242466,
+      "learning_rate": 3.8216560509554143e-07,
+      "loss": 0.1321,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.934771789969799,
+      "learning_rate": 3.842887473460722e-07,
+      "loss": 0.3233,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.21372971984552552,
+      "learning_rate": 3.86411889596603e-07,
+      "loss": 0.3842,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.18177493199129321,
+      "learning_rate": 3.885350318471338e-07,
+      "loss": 0.1263,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.20539564694344178,
+      "learning_rate": 3.9065817409766457e-07,
+      "loss": 0.099,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.24517169926202315,
+      "learning_rate": 3.9278131634819535e-07,
+      "loss": 0.1368,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1843016081672305,
+      "learning_rate": 3.9490445859872613e-07,
+      "loss": 0.0505,
+      "step": 186
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.30375760425814136,
+      "learning_rate": 3.970276008492569e-07,
+      "loss": 0.2151,
+      "step": 187
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.23789039380870222,
+      "learning_rate": 3.991507430997877e-07,
+      "loss": 0.0694,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.24787341950436864,
+      "learning_rate": 4.012738853503185e-07,
+      "loss": 0.3139,
+      "step": 189
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.24142286451697284,
+      "learning_rate": 4.0339702760084927e-07,
+      "loss": 0.1738,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.20276550233245788,
+      "learning_rate": 4.0552016985138005e-07,
+      "loss": 0.0852,
+      "step": 191
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4582239097053706,
+      "learning_rate": 4.0764331210191083e-07,
+      "loss": 0.219,
+      "step": 192
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3893721484305137,
+      "learning_rate": 4.097664543524416e-07,
+      "loss": 0.2916,
+      "step": 193
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.46880224835496076,
+      "learning_rate": 4.118895966029724e-07,
+      "loss": 0.2952,
+      "step": 194
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.13977758365510837,
+      "learning_rate": 4.140127388535032e-07,
+      "loss": 0.1819,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.13470332419651734,
+      "learning_rate": 4.1613588110403397e-07,
+      "loss": 0.1215,
+      "step": 196
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2307751266090855,
+      "learning_rate": 4.1825902335456475e-07,
+      "loss": 0.2767,
+      "step": 197
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.271595431624329,
+      "learning_rate": 4.2038216560509554e-07,
+      "loss": 0.1746,
+      "step": 198
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.23079560235515081,
+      "learning_rate": 4.225053078556263e-07,
+      "loss": 0.1271,
+      "step": 199
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.21300416406820002,
+      "learning_rate": 4.246284501061571e-07,
+      "loss": 0.2239,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.11774526827543365,
+      "learning_rate": 4.267515923566879e-07,
+      "loss": 0.1041,
+      "step": 201
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.14021883811615443,
+      "learning_rate": 4.2887473460721867e-07,
+      "loss": 0.3552,
+      "step": 202
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.15598643532331874,
+      "learning_rate": 4.3099787685774946e-07,
+      "loss": 0.0227,
+      "step": 203
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.19572744171804726,
+      "learning_rate": 4.3312101910828024e-07,
+      "loss": 0.1274,
+      "step": 204
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.27843730774946024,
+      "learning_rate": 4.35244161358811e-07,
+      "loss": 0.3357,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2768138090599354,
+      "learning_rate": 4.373673036093418e-07,
+      "loss": 0.2559,
+      "step": 206
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.30303560402599317,
+      "learning_rate": 4.394904458598726e-07,
+      "loss": 0.2036,
+      "step": 207
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3574035759761096,
+      "learning_rate": 4.416135881104034e-07,
+      "loss": 0.0257,
+      "step": 208
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.513602085503996,
+      "learning_rate": 4.4373673036093416e-07,
+      "loss": 0.1839,
+      "step": 209
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.23574962632529362,
+      "learning_rate": 4.4585987261146494e-07,
+      "loss": 0.131,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.301502611286239,
+      "learning_rate": 4.479830148619957e-07,
+      "loss": 0.1754,
+      "step": 211
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.12831294475009764,
+      "learning_rate": 4.501061571125265e-07,
+      "loss": 0.0413,
+      "step": 212
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.44448268764246396,
+      "learning_rate": 4.522292993630573e-07,
+      "loss": 0.0187,
+      "step": 213
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.16541563293031547,
+      "learning_rate": 4.543524416135881e-07,
+      "loss": 0.0696,
+      "step": 214
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4156837966875057,
+      "learning_rate": 4.5647558386411886e-07,
+      "loss": 0.1157,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.27956594396333684,
+      "learning_rate": 4.585987261146497e-07,
+      "loss": 0.0714,
+      "step": 216
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1141071927218783,
+      "learning_rate": 4.607218683651805e-07,
+      "loss": 0.0591,
+      "step": 217
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.168783943119781,
+      "learning_rate": 4.6284501061571126e-07,
+      "loss": 0.2204,
+      "step": 218
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1561287939942871,
+      "learning_rate": 4.6496815286624205e-07,
+      "loss": 0.0735,
+      "step": 219
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.30322314822119223,
+      "learning_rate": 4.6709129511677283e-07,
+      "loss": 0.1685,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3697363694305286,
+      "learning_rate": 4.692144373673036e-07,
+      "loss": 0.3518,
+      "step": 221
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2906261422640796,
+      "learning_rate": 4.713375796178344e-07,
+      "loss": 0.1692,
+      "step": 222
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26954453234957454,
+      "learning_rate": 4.734607218683652e-07,
+      "loss": 0.1832,
+      "step": 223
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.27235398428229896,
+      "learning_rate": 4.7558386411889597e-07,
+      "loss": 0.3283,
+      "step": 224
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.280661228991786,
+      "learning_rate": 4.777070063694267e-07,
+      "loss": 0.314,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2218287050986227,
+      "learning_rate": 4.798301486199575e-07,
+      "loss": 0.1771,
+      "step": 226
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.12591559039362718,
+      "learning_rate": 4.819532908704883e-07,
+      "loss": 0.0138,
+      "step": 227
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.13838304163340706,
+      "learning_rate": 4.840764331210191e-07,
+      "loss": 0.0261,
+      "step": 228
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.11523289156706043,
+      "learning_rate": 4.861995753715499e-07,
+      "loss": 0.1658,
+      "step": 229
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.24297355446371952,
+      "learning_rate": 4.883227176220807e-07,
+      "loss": 0.3567,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.19210771351944797,
+      "learning_rate": 4.904458598726115e-07,
+      "loss": 0.1158,
+      "step": 231
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.17117268316319734,
+      "learning_rate": 4.925690021231422e-07,
+      "loss": 0.0205,
+      "step": 232
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5389672372369377,
+      "learning_rate": 4.94692144373673e-07,
+      "loss": 0.1087,
+      "step": 233
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.10165201346621026,
+      "learning_rate": 4.968152866242038e-07,
+      "loss": 0.021,
+      "step": 234
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.15794862575618773,
+      "learning_rate": 4.989384288747346e-07,
+      "loss": 0.1067,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0857455512752192,
+      "learning_rate": 5.010615711252654e-07,
+      "loss": 0.0126,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.15674156646349519,
+      "learning_rate": 5.031847133757962e-07,
+      "loss": 0.17,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.23624012189202026,
+      "learning_rate": 5.053078556263269e-07,
+      "loss": 0.0348,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.20950870199232627,
+      "learning_rate": 5.074309978768577e-07,
+      "loss": 0.1308,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.182355080636987,
+      "learning_rate": 5.095541401273885e-07,
+      "loss": 0.2538,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.19906201837253645,
+      "learning_rate": 5.116772823779193e-07,
+      "loss": 0.1013,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.31748922822807213,
+      "learning_rate": 5.138004246284501e-07,
+      "loss": 0.1983,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2879583102322477,
+      "learning_rate": 5.159235668789809e-07,
+      "loss": 0.3305,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2848482889330333,
+      "learning_rate": 5.180467091295116e-07,
+      "loss": 0.1683,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.38641844976427114,
+      "learning_rate": 5.201698513800424e-07,
+      "loss": 0.1918,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29791922743032145,
+      "learning_rate": 5.222929936305732e-07,
+      "loss": 0.2141,
+      "step": 246
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.37328860142096176,
+      "learning_rate": 5.24416135881104e-07,
+      "loss": 0.324,
+      "step": 247
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11778988848411559,
+      "learning_rate": 5.265392781316348e-07,
+      "loss": 0.197,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.33803319607529914,
+      "learning_rate": 5.286624203821656e-07,
+      "loss": 0.1999,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.09994307309166663,
+      "learning_rate": 5.307855626326963e-07,
+      "loss": 0.0133,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.18546932643609904,
+      "learning_rate": 5.329087048832271e-07,
+      "loss": 0.1493,
+      "step": 251
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.24972406037114434,
+      "learning_rate": 5.350318471337579e-07,
+      "loss": 0.2281,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.22815217212037803,
+      "learning_rate": 5.371549893842887e-07,
+      "loss": 0.3029,
+      "step": 253
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.35664572216679813,
+      "learning_rate": 5.392781316348195e-07,
+      "loss": 0.2978,
+      "step": 254
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.36878615874476434,
+      "learning_rate": 5.414012738853503e-07,
+      "loss": 0.1935,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.12433161555769073,
+      "learning_rate": 5.43524416135881e-07,
+      "loss": 0.0079,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7621643835905139,
+      "learning_rate": 5.456475583864118e-07,
+      "loss": 0.2617,
+      "step": 257
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.19432234383374675,
+      "learning_rate": 5.477707006369426e-07,
+      "loss": 0.1661,
+      "step": 258
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.15478225006265445,
+      "learning_rate": 5.498938428874734e-07,
+      "loss": 0.1457,
+      "step": 259
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4026440623845048,
+      "learning_rate": 5.520169851380042e-07,
+      "loss": 0.2138,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.14132590741784368,
+      "learning_rate": 5.54140127388535e-07,
+      "loss": 0.3274,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2744286830896772,
+      "learning_rate": 5.562632696390657e-07,
+      "loss": 0.319,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.32236468776382265,
+      "learning_rate": 5.583864118895965e-07,
+      "loss": 0.256,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.17485595559343078,
+      "learning_rate": 5.605095541401273e-07,
+      "loss": 0.2133,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29452707985083865,
+      "learning_rate": 5.626326963906581e-07,
+      "loss": 0.0416,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3424552020931376,
+      "learning_rate": 5.647558386411889e-07,
+      "loss": 0.1472,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26338706700603143,
+      "learning_rate": 5.668789808917197e-07,
+      "loss": 0.296,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.14705260009709054,
+      "learning_rate": 5.690021231422504e-07,
+      "loss": 0.1678,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7860159787237884,
+      "learning_rate": 5.711252653927812e-07,
+      "loss": 0.1939,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.13624004223090053,
+      "learning_rate": 5.73248407643312e-07,
+      "loss": 0.1554,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.09898779142099313,
+      "learning_rate": 5.753715498938428e-07,
+      "loss": 0.192,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.18706483719575634,
+      "learning_rate": 5.774946921443736e-07,
+      "loss": 0.2032,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29195550701482514,
+      "learning_rate": 5.796178343949044e-07,
+      "loss": 0.2077,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3392739493646915,
+      "learning_rate": 5.817409766454351e-07,
+      "loss": 0.2532,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3041085515552397,
+      "learning_rate": 5.838641188959659e-07,
+      "loss": 0.0918,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4770793774693724,
+      "learning_rate": 5.859872611464968e-07,
+      "loss": 0.1106,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.19562567545163273,
+      "learning_rate": 5.881104033970276e-07,
+      "loss": 0.1333,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2707278306013383,
+      "learning_rate": 5.902335456475584e-07,
+      "loss": 0.1306,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.16296943278549977,
+      "learning_rate": 5.923566878980892e-07,
+      "loss": 0.0364,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5386648312096373,
+      "learning_rate": 5.9447983014862e-07,
+      "loss": 0.0808,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2942127275073548,
+      "learning_rate": 5.966029723991507e-07,
+      "loss": 0.2626,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3119903160409143,
+      "learning_rate": 5.987261146496815e-07,
+      "loss": 0.1895,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2737013844339624,
+      "learning_rate": 6.008492569002123e-07,
+      "loss": 0.1385,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.16855653755832362,
+      "learning_rate": 6.029723991507431e-07,
+      "loss": 0.1529,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.49718223528016353,
+      "learning_rate": 6.050955414012739e-07,
+      "loss": 0.4262,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2627780618572779,
+      "learning_rate": 6.072186836518047e-07,
+      "loss": 0.3654,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2270233033606091,
+      "learning_rate": 6.093418259023354e-07,
+      "loss": 0.1655,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.32556073415138226,
+      "learning_rate": 6.114649681528662e-07,
+      "loss": 0.1,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0688499303724151,
+      "learning_rate": 6.13588110403397e-07,
+      "loss": 0.0131,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.19737684668244823,
+      "learning_rate": 6.157112526539278e-07,
+      "loss": 0.2689,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.22678346062623986,
+      "learning_rate": 6.178343949044586e-07,
+      "loss": 0.2251,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.21751630427029792,
+      "learning_rate": 6.199575371549894e-07,
+      "loss": 0.2584,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2141418417559859,
+      "learning_rate": 6.220806794055201e-07,
+      "loss": 0.0784,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1430074561073039,
+      "learning_rate": 6.242038216560509e-07,
+      "loss": 0.1433,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2786748881959617,
+      "learning_rate": 6.263269639065817e-07,
+      "loss": 0.2284,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2433691630645959,
+      "learning_rate": 6.284501061571125e-07,
+      "loss": 0.1846,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7923395523484135,
+      "learning_rate": 6.305732484076433e-07,
+      "loss": 0.3029,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1963409180306421,
+      "learning_rate": 6.326963906581741e-07,
+      "loss": 0.1897,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.28965013571514175,
+      "learning_rate": 6.348195329087048e-07,
+      "loss": 0.0649,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.10288304663737333,
+      "learning_rate": 6.369426751592356e-07,
+      "loss": 0.0935,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2386140993298097,
+      "learning_rate": 6.390658174097664e-07,
+      "loss": 0.1989,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2001752439033513,
+      "learning_rate": 6.411889596602972e-07,
+      "loss": 0.2559,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.14419532926564962,
+      "learning_rate": 6.43312101910828e-07,
+      "loss": 0.1163,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.20905738114675854,
+      "learning_rate": 6.454352441613588e-07,
+      "loss": 0.015,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.24380431007046924,
+      "learning_rate": 6.475583864118895e-07,
+      "loss": 0.112,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3023639200080016,
+      "learning_rate": 6.496815286624203e-07,
+      "loss": 0.2968,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2606866540139771,
+      "learning_rate": 6.518046709129511e-07,
+      "loss": 0.1304,
+      "step": 307
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2656805524127184,
+      "learning_rate": 6.539278131634819e-07,
+      "loss": 0.2697,
+      "step": 308
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3193198432455117,
+      "learning_rate": 6.560509554140127e-07,
+      "loss": 0.1832,
+      "step": 309
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2874561835870219,
+      "learning_rate": 6.581740976645435e-07,
+      "loss": 0.0569,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3174031017726217,
+      "learning_rate": 6.602972399150743e-07,
+      "loss": 0.1652,
+      "step": 311
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.20165045009761387,
+      "learning_rate": 6.62420382165605e-07,
+      "loss": 0.0533,
+      "step": 312
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26014267892389276,
+      "learning_rate": 6.645435244161358e-07,
+      "loss": 0.3167,
+      "step": 313
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1251119388682503,
+      "learning_rate": 6.666666666666666e-07,
+      "loss": 0.1834,
+      "step": 314
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3945081001141345,
+      "learning_rate": 6.687898089171974e-07,
+      "loss": 0.2264,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3474798985267554,
+      "learning_rate": 6.709129511677282e-07,
+      "loss": 0.2383,
+      "step": 316
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1947402716044741,
+      "learning_rate": 6.73036093418259e-07,
+      "loss": 0.2622,
+      "step": 317
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.22362536364169935,
+      "learning_rate": 6.751592356687897e-07,
+      "loss": 0.0834,
+      "step": 318
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11661468027254816,
+      "learning_rate": 6.772823779193205e-07,
+      "loss": 0.0487,
+      "step": 319
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.17541144954253332,
+      "learning_rate": 6.794055201698513e-07,
+      "loss": 0.1103,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.13308947738335472,
+      "learning_rate": 6.815286624203821e-07,
+      "loss": 0.0908,
+      "step": 321
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.28308382340263966,
+      "learning_rate": 6.836518046709129e-07,
+      "loss": 0.2402,
+      "step": 322
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.16304676460641138,
+      "learning_rate": 6.857749469214437e-07,
+      "loss": 0.1215,
+      "step": 323
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2690315548677506,
+      "learning_rate": 6.878980891719744e-07,
+      "loss": 0.1385,
+      "step": 324
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.28782689849209325,
+      "learning_rate": 6.900212314225052e-07,
+      "loss": 0.1083,
+      "step": 325
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.32365798582335054,
+      "learning_rate": 6.92144373673036e-07,
+      "loss": 0.1046,
+      "step": 326
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3017609449576566,
+      "learning_rate": 6.942675159235668e-07,
+      "loss": 0.2847,
+      "step": 327
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26305840526601176,
+      "learning_rate": 6.963906581740976e-07,
+      "loss": 0.132,
+      "step": 328
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.527745134158186,
+      "learning_rate": 6.985138004246284e-07,
+      "loss": 0.3415,
+      "step": 329
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2400082999037914,
+      "learning_rate": 7.006369426751591e-07,
+      "loss": 0.224,
+      "step": 330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.31498865831609935,
+      "learning_rate": 7.027600849256899e-07,
+      "loss": 0.2224,
+      "step": 331
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.23024024876382654,
+      "learning_rate": 7.048832271762207e-07,
+      "loss": 0.1899,
+      "step": 332
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4186895240598085,
+      "learning_rate": 7.070063694267515e-07,
+      "loss": 0.2308,
+      "step": 333
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.13259159510712912,
+      "learning_rate": 7.091295116772823e-07,
+      "loss": 0.1469,
+      "step": 334
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2373762968766546,
+      "learning_rate": 7.112526539278131e-07,
+      "loss": 0.1811,
+      "step": 335
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2953297267584623,
+      "learning_rate": 7.133757961783438e-07,
+      "loss": 0.2956,
+      "step": 336
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2963767478220723,
+      "learning_rate": 7.154989384288746e-07,
+      "loss": 0.27,
+      "step": 337
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.22335625807264956,
+      "learning_rate": 7.176220806794054e-07,
+      "loss": 0.1269,
+      "step": 338
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.32139447372756724,
+      "learning_rate": 7.197452229299362e-07,
+      "loss": 0.2346,
+      "step": 339
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.23705093106999992,
+      "learning_rate": 7.21868365180467e-07,
+      "loss": 0.2248,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.22767202512578014,
+      "learning_rate": 7.239915074309978e-07,
+      "loss": 0.2066,
+      "step": 341
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.18397356381653115,
+      "learning_rate": 7.261146496815285e-07,
+      "loss": 0.1799,
+      "step": 342
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.20673006822064957,
+      "learning_rate": 7.282377919320593e-07,
+      "loss": 0.4484,
+      "step": 343
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0557207926549818,
+      "learning_rate": 7.303609341825901e-07,
+      "loss": 0.0687,
+      "step": 344
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3081173807962164,
+      "learning_rate": 7.324840764331209e-07,
+      "loss": 0.2799,
+      "step": 345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.18776167618511286,
+      "learning_rate": 7.346072186836517e-07,
+      "loss": 0.2008,
+      "step": 346
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3296962816092498,
+      "learning_rate": 7.367303609341825e-07,
+      "loss": 0.4462,
+      "step": 347
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1485189730693178,
+      "learning_rate": 7.388535031847132e-07,
+      "loss": 0.0807,
+      "step": 348
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6852474737176507,
+      "learning_rate": 7.40976645435244e-07,
+      "loss": 0.2873,
+      "step": 349
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3262034889242401,
+      "learning_rate": 7.430997876857748e-07,
+      "loss": 0.1459,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.15031259529226046,
+      "learning_rate": 7.452229299363056e-07,
+      "loss": 0.2377,
+      "step": 351
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.16892482959317795,
+      "learning_rate": 7.473460721868364e-07,
+      "loss": 0.2246,
+      "step": 352
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.25425496343673953,
+      "learning_rate": 7.494692144373672e-07,
+      "loss": 0.0964,
+      "step": 353
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.13793263456492666,
+      "learning_rate": 7.515923566878982e-07,
+      "loss": 0.1007,
+      "step": 354
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2097891423154737,
+      "learning_rate": 7.537154989384289e-07,
+      "loss": 0.2488,
+      "step": 355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.08684897361559107,
+      "learning_rate": 7.558386411889597e-07,
+      "loss": 0.1011,
+      "step": 356
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.14510536271973992,
+      "learning_rate": 7.579617834394905e-07,
+      "loss": 0.0371,
+      "step": 357
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2233939359582414,
+      "learning_rate": 7.600849256900213e-07,
+      "loss": 0.2513,
+      "step": 358
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1778509190137439,
+      "learning_rate": 7.622080679405521e-07,
+      "loss": 0.1493,
+      "step": 359
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29882084943077136,
+      "learning_rate": 7.643312101910829e-07,
+      "loss": 0.0833,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.27539509978181104,
+      "learning_rate": 7.664543524416136e-07,
+      "loss": 0.4315,
+      "step": 361
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.43024944450808494,
+      "learning_rate": 7.685774946921444e-07,
+      "loss": 0.2586,
+      "step": 362
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29794875496527123,
+      "learning_rate": 7.707006369426752e-07,
+      "loss": 0.1013,
+      "step": 363
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.11911472604816405,
+      "learning_rate": 7.72823779193206e-07,
+      "loss": 0.1041,
+      "step": 364
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.16868684281299706,
+      "learning_rate": 7.749469214437368e-07,
+      "loss": 0.2312,
+      "step": 365
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1845638602526959,
+      "learning_rate": 7.770700636942676e-07,
+      "loss": 0.322,
+      "step": 366
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26984042118586926,
+      "learning_rate": 7.791932059447983e-07,
+      "loss": 0.2608,
+      "step": 367
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.16810527194085234,
+      "learning_rate": 7.813163481953291e-07,
+      "loss": 0.2321,
+      "step": 368
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.06283977236084466,
+      "learning_rate": 7.834394904458599e-07,
+      "loss": 0.0854,
+      "step": 369
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3218832078829887,
+      "learning_rate": 7.855626326963907e-07,
+      "loss": 0.2296,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1727750015726721,
+      "learning_rate": 7.876857749469215e-07,
+      "loss": 0.0982,
+      "step": 371
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2994955471352695,
+      "learning_rate": 7.898089171974523e-07,
+      "loss": 0.0651,
+      "step": 372
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3306255736205435,
+      "learning_rate": 7.91932059447983e-07,
+      "loss": 0.2255,
+      "step": 373
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3446522069251835,
+      "learning_rate": 7.940552016985138e-07,
+      "loss": 0.2778,
+      "step": 374
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.15815947749538553,
+      "learning_rate": 7.961783439490446e-07,
+      "loss": 0.0535,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.10367095348687529,
+      "learning_rate": 7.983014861995754e-07,
+      "loss": 0.049,
+      "step": 376
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.249563603920521,
+      "learning_rate": 8.004246284501062e-07,
+      "loss": 0.2211,
+      "step": 377
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3102621403983259,
+      "learning_rate": 8.02547770700637e-07,
+      "loss": 0.158,
+      "step": 378
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.45287223375211105,
+      "learning_rate": 8.046709129511678e-07,
+      "loss": 0.1806,
+      "step": 379
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.24097223015886662,
+      "learning_rate": 8.067940552016985e-07,
+      "loss": 0.171,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2050006159478445,
+      "learning_rate": 8.089171974522293e-07,
+      "loss": 0.0992,
+      "step": 381
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.20793762739204616,
+      "learning_rate": 8.110403397027601e-07,
+      "loss": 0.1235,
+      "step": 382
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.19965927601019257,
+      "learning_rate": 8.131634819532909e-07,
+      "loss": 0.5675,
+      "step": 383
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4643045643537523,
+      "learning_rate": 8.152866242038217e-07,
+      "loss": 0.1717,
+      "step": 384
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.12032396043051378,
+      "learning_rate": 8.174097664543525e-07,
+      "loss": 0.1565,
+      "step": 385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3713133529651683,
+      "learning_rate": 8.195329087048832e-07,
+      "loss": 0.0515,
+      "step": 386
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3202619840308792,
+      "learning_rate": 8.21656050955414e-07,
+      "loss": 0.291,
+      "step": 387
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.19443370285639025,
+      "learning_rate": 8.237791932059448e-07,
+      "loss": 0.1902,
+      "step": 388
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.19849023334196433,
+      "learning_rate": 8.259023354564756e-07,
+      "loss": 0.1461,
+      "step": 389
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.214838462555769,
+      "learning_rate": 8.280254777070064e-07,
+      "loss": 0.1556,
+      "step": 390
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.13309551953092594,
+      "learning_rate": 8.301486199575372e-07,
+      "loss": 0.0695,
+      "step": 391
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.47221558686874704,
+      "learning_rate": 8.322717622080679e-07,
+      "loss": 0.3792,
+      "step": 392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3251883958033443,
+      "learning_rate": 8.343949044585987e-07,
+      "loss": 0.1702,
+      "step": 393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13748152191871976,
+      "learning_rate": 8.365180467091295e-07,
+      "loss": 0.0414,
+      "step": 394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2577512517607871,
+      "learning_rate": 8.386411889596603e-07,
+      "loss": 0.2511,
+      "step": 395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.18287998914848866,
+      "learning_rate": 8.407643312101911e-07,
+      "loss": 0.1164,
+      "step": 396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17709339849932243,
+      "learning_rate": 8.428874734607219e-07,
+      "loss": 0.0207,
+      "step": 397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2298605814218488,
+      "learning_rate": 8.450106157112526e-07,
+      "loss": 0.2123,
+      "step": 398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3962396303985176,
+      "learning_rate": 8.471337579617834e-07,
+      "loss": 0.4138,
+      "step": 399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2808567887236285,
+      "learning_rate": 8.492569002123142e-07,
+      "loss": 0.4247,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.23215587936104334,
+      "learning_rate": 8.51380042462845e-07,
+      "loss": 0.1985,
+      "step": 401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2939695074098435,
+      "learning_rate": 8.535031847133758e-07,
+      "loss": 0.1631,
+      "step": 402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.23766967230175637,
+      "learning_rate": 8.556263269639066e-07,
+      "loss": 0.1726,
+      "step": 403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.30850612413413675,
+      "learning_rate": 8.577494692144373e-07,
+      "loss": 0.2422,
+      "step": 404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2413195428537783,
+      "learning_rate": 8.598726114649681e-07,
+      "loss": 0.5659,
+      "step": 405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13921330227742174,
+      "learning_rate": 8.619957537154989e-07,
+      "loss": 0.0168,
+      "step": 406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2163843683745216,
+      "learning_rate": 8.641188959660297e-07,
+      "loss": 0.3101,
+      "step": 407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2196577154163935,
+      "learning_rate": 8.662420382165605e-07,
+      "loss": 0.3257,
+      "step": 408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6236373549005418,
+      "learning_rate": 8.683651804670913e-07,
+      "loss": 0.3986,
+      "step": 409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.12046982288675889,
+      "learning_rate": 8.70488322717622e-07,
+      "loss": 0.125,
+      "step": 410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2581329329149649,
+      "learning_rate": 8.726114649681528e-07,
+      "loss": 0.4828,
+      "step": 411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2851014489197922,
+      "learning_rate": 8.747346072186836e-07,
+      "loss": 0.2913,
+      "step": 412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2410155492065017,
+      "learning_rate": 8.768577494692144e-07,
+      "loss": 0.1909,
+      "step": 413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17536848451664563,
+      "learning_rate": 8.789808917197452e-07,
+      "loss": 0.1295,
+      "step": 414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.38772909706184794,
+      "learning_rate": 8.81104033970276e-07,
+      "loss": 0.188,
+      "step": 415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.28915557836813943,
+      "learning_rate": 8.832271762208067e-07,
+      "loss": 0.1589,
+      "step": 416
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.28975293174705274,
+      "learning_rate": 8.853503184713375e-07,
+      "loss": 0.1225,
+      "step": 417
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.22249415674474898,
+      "learning_rate": 8.874734607218683e-07,
+      "loss": 0.2004,
+      "step": 418
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2940454009135285,
+      "learning_rate": 8.895966029723991e-07,
+      "loss": 0.1272,
+      "step": 419
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.42358785034734187,
+      "learning_rate": 8.917197452229299e-07,
+      "loss": 0.0576,
+      "step": 420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.21235244241456386,
+      "learning_rate": 8.938428874734607e-07,
+      "loss": 0.2632,
+      "step": 421
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.37224640019004696,
+      "learning_rate": 8.959660297239914e-07,
+      "loss": 0.2549,
+      "step": 422
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.36378654543030986,
+      "learning_rate": 8.980891719745222e-07,
+      "loss": 0.3476,
+      "step": 423
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2566243356722543,
+      "learning_rate": 9.00212314225053e-07,
+      "loss": 0.2631,
+      "step": 424
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4315065982171834,
+      "learning_rate": 9.023354564755838e-07,
+      "loss": 0.2681,
+      "step": 425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13878811119352702,
+      "learning_rate": 9.044585987261146e-07,
+      "loss": 0.0866,
+      "step": 426
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.098010200776135,
+      "learning_rate": 9.065817409766454e-07,
+      "loss": 0.0576,
+      "step": 427
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.18421455741569376,
+      "learning_rate": 9.087048832271762e-07,
+      "loss": 0.1636,
+      "step": 428
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.22135093664942307,
+      "learning_rate": 9.108280254777069e-07,
+      "loss": 0.1907,
+      "step": 429
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2613270863068073,
+      "learning_rate": 9.129511677282377e-07,
+      "loss": 0.0568,
+      "step": 430
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4885576563006937,
+      "learning_rate": 9.150743099787685e-07,
+      "loss": 0.1314,
+      "step": 431
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3843438674385242,
+      "learning_rate": 9.171974522292994e-07,
+      "loss": 0.2748,
+      "step": 432
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3162268006181552,
+      "learning_rate": 9.193205944798302e-07,
+      "loss": 0.2064,
+      "step": 433
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.18415232052742567,
+      "learning_rate": 9.21443736730361e-07,
+      "loss": 0.1955,
+      "step": 434
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1977546707271865,
+      "learning_rate": 9.235668789808917e-07,
+      "loss": 0.1213,
+      "step": 435
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17277134525309018,
+      "learning_rate": 9.256900212314225e-07,
+      "loss": 0.288,
+      "step": 436
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.25520185550781,
+      "learning_rate": 9.278131634819533e-07,
+      "loss": 0.1035,
+      "step": 437
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17704131808478485,
+      "learning_rate": 9.299363057324841e-07,
+      "loss": 0.0181,
+      "step": 438
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3245504988404467,
+      "learning_rate": 9.320594479830149e-07,
+      "loss": 0.2044,
+      "step": 439
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.18244120441800724,
+      "learning_rate": 9.341825902335457e-07,
+      "loss": 0.2578,
+      "step": 440
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.21902255767587192,
+      "learning_rate": 9.363057324840764e-07,
+      "loss": 0.1456,
+      "step": 441
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1695979665834792,
+      "learning_rate": 9.384288747346072e-07,
+      "loss": 0.1774,
+      "step": 442
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2676615780508529,
+      "learning_rate": 9.40552016985138e-07,
+      "loss": 0.2396,
+      "step": 443
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.34088965282414585,
+      "learning_rate": 9.426751592356688e-07,
+      "loss": 0.1528,
+      "step": 444
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.07925269437411404,
+      "learning_rate": 9.447983014861996e-07,
+      "loss": 0.0918,
+      "step": 445
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13870357480386233,
+      "learning_rate": 9.469214437367304e-07,
+      "loss": 0.0852,
+      "step": 446
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.32870650312044764,
+      "learning_rate": 9.490445859872611e-07,
+      "loss": 0.1525,
+      "step": 447
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.24305219592733074,
+      "learning_rate": 9.511677282377919e-07,
+      "loss": 0.2167,
+      "step": 448
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.31525378059408554,
+      "learning_rate": 9.532908704883227e-07,
+      "loss": 0.364,
+      "step": 449
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4545081071608993,
+      "learning_rate": 9.554140127388535e-07,
+      "loss": 0.0589,
+      "step": 450
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1864295918607807,
+      "learning_rate": 9.575371549893843e-07,
+      "loss": 0.1124,
+      "step": 451
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1497477565593572,
+      "learning_rate": 9.59660297239915e-07,
+      "loss": 0.2063,
+      "step": 452
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.15682831310264794,
+      "learning_rate": 9.617834394904458e-07,
+      "loss": 0.1217,
+      "step": 453
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.23565246524026945,
+      "learning_rate": 9.639065817409766e-07,
+      "loss": 0.2472,
+      "step": 454
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.21338418196035058,
+      "learning_rate": 9.660297239915074e-07,
+      "loss": 0.1409,
+      "step": 455
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.32363018578448816,
+      "learning_rate": 9.681528662420382e-07,
+      "loss": 0.0964,
+      "step": 456
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.36305631286103945,
+      "learning_rate": 9.70276008492569e-07,
+      "loss": 0.2474,
+      "step": 457
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1315636111272709,
+      "learning_rate": 9.723991507430998e-07,
+      "loss": 0.0133,
+      "step": 458
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.36665855866774594,
+      "learning_rate": 9.745222929936306e-07,
+      "loss": 0.0556,
+      "step": 459
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2932467384725678,
+      "learning_rate": 9.766454352441613e-07,
+      "loss": 0.4045,
+      "step": 460
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.19772647917989614,
+      "learning_rate": 9.787685774946921e-07,
+      "loss": 0.1258,
+      "step": 461
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1944562471708232,
+      "learning_rate": 9.80891719745223e-07,
+      "loss": 0.1912,
+      "step": 462
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5129004712182335,
+      "learning_rate": 9.830148619957537e-07,
+      "loss": 0.0369,
+      "step": 463
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.27134534580369873,
+      "learning_rate": 9.851380042462845e-07,
+      "loss": 0.2076,
+      "step": 464
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6036942576112112,
+      "learning_rate": 9.872611464968153e-07,
+      "loss": 0.1239,
+      "step": 465
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4238546988928349,
+      "learning_rate": 9.89384288747346e-07,
+      "loss": 0.1025,
+      "step": 466
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.27541270253213207,
+      "learning_rate": 9.915074309978768e-07,
+      "loss": 0.194,
+      "step": 467
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2546318338738281,
+      "learning_rate": 9.936305732484076e-07,
+      "loss": 0.2443,
+      "step": 468
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.18805689752215418,
+      "learning_rate": 9.957537154989384e-07,
+      "loss": 0.1765,
+      "step": 469
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.29960419963214957,
+      "learning_rate": 9.978768577494692e-07,
+      "loss": 0.2171,
+      "step": 470
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.34422612896806737,
+      "learning_rate": 1e-06,
+      "loss": 0.4547,
+      "step": 471
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.21025726996204436,
+      "learning_rate": 9.999999893344976e-07,
+      "loss": 0.2975,
+      "step": 472
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3184464372094676,
+      "learning_rate": 9.999999573379902e-07,
+      "loss": 0.3635,
+      "step": 473
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.22529194348745743,
+      "learning_rate": 9.999999040104795e-07,
+      "loss": 0.1233,
+      "step": 474
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17920512378839853,
+      "learning_rate": 9.99999829351968e-07,
+      "loss": 0.1818,
+      "step": 475
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.21146143920686314,
+      "learning_rate": 9.999997333624587e-07,
+      "loss": 0.1197,
+      "step": 476
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.21073388057143283,
+      "learning_rate": 9.999996160419555e-07,
+      "loss": 0.1825,
+      "step": 477
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.16078577139204342,
+      "learning_rate": 9.999994773904636e-07,
+      "loss": 0.2614,
+      "step": 478
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2761128508668464,
+      "learning_rate": 9.999993174079888e-07,
+      "loss": 0.1577,
+      "step": 479
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2885503791210814,
+      "learning_rate": 9.999991360945382e-07,
+      "loss": 0.1658,
+      "step": 480
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5056550557135914,
+      "learning_rate": 9.99998933450119e-07,
+      "loss": 0.1356,
+      "step": 481
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2587423302824715,
+      "learning_rate": 9.999987094747404e-07,
+      "loss": 0.1906,
+      "step": 482
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.27175524731807416,
+      "learning_rate": 9.999984641684116e-07,
+      "loss": 0.1844,
+      "step": 483
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.16102534545447217,
+      "learning_rate": 9.999981975311433e-07,
+      "loss": 0.1798,
+      "step": 484
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13692929348976293,
+      "learning_rate": 9.999979095629469e-07,
+      "loss": 0.1092,
+      "step": 485
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.19888672041303426,
+      "learning_rate": 9.999976002638344e-07,
+      "loss": 0.3372,
+      "step": 486
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.26586280153311215,
+      "learning_rate": 9.99997269633819e-07,
+      "loss": 0.1877,
+      "step": 487
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.10917284468967871,
+      "learning_rate": 9.999969176729153e-07,
+      "loss": 0.1221,
+      "step": 488
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.650651973475683,
+      "learning_rate": 9.999965443811376e-07,
+      "loss": 0.1607,
+      "step": 489
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2145075448697716,
+      "learning_rate": 9.999961497585024e-07,
+      "loss": 0.0983,
+      "step": 490
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6452394084802534,
+      "learning_rate": 9.999957338050265e-07,
+      "loss": 0.3028,
+      "step": 491
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.29862832309920007,
+      "learning_rate": 9.999952965207273e-07,
+      "loss": 0.1307,
+      "step": 492
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.32137994202195397,
+      "learning_rate": 9.999948379056235e-07,
+      "loss": 0.1424,
+      "step": 493
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.38034356609412096,
+      "learning_rate": 9.99994357959735e-07,
+      "loss": 0.3911,
+      "step": 494
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.15666426264417455,
+      "learning_rate": 9.99993856683082e-07,
+      "loss": 0.2112,
+      "step": 495
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.12027522270375962,
+      "learning_rate": 9.99993334075686e-07,
+      "loss": 0.0216,
+      "step": 496
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.10923574165943764,
+      "learning_rate": 9.99992790137569e-07,
+      "loss": 0.1166,
+      "step": 497
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.14919427521926085,
+      "learning_rate": 9.999922248687548e-07,
+      "loss": 0.1056,
+      "step": 498
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3044315277531875,
+      "learning_rate": 9.99991638269267e-07,
+      "loss": 0.1135,
+      "step": 499
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6105525827468412,
+      "learning_rate": 9.999910303391308e-07,
+      "loss": 0.1656,
+      "step": 500
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.23900556126543304,
+      "learning_rate": 9.999904010783723e-07,
+      "loss": 0.1068,
+      "step": 501
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.394750778990148,
+      "learning_rate": 9.999897504870182e-07,
+      "loss": 0.2906,
+      "step": 502
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.12818377999218414,
+      "learning_rate": 9.99989078565096e-07,
+      "loss": 0.1879,
+      "step": 503
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3565725372269199,
+      "learning_rate": 9.999883853126348e-07,
+      "loss": 0.134,
+      "step": 504
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.26619545869125333,
+      "learning_rate": 9.999876707296637e-07,
+      "loss": 0.2273,
+      "step": 505
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2532572699129161,
+      "learning_rate": 9.999869348162139e-07,
+      "loss": 0.021,
+      "step": 506
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.16057755469215762,
+      "learning_rate": 9.999861775723161e-07,
+      "loss": 0.038,
+      "step": 507
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.14580176791619404,
+      "learning_rate": 9.999853989980027e-07,
+      "loss": 0.1291,
+      "step": 508
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.23289785982715136,
+      "learning_rate": 9.999845990933073e-07,
+      "loss": 0.1479,
+      "step": 509
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.33869594797132546,
+      "learning_rate": 9.999837778582638e-07,
+      "loss": 0.1374,
+      "step": 510
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1817354272596778,
+      "learning_rate": 9.999829352929074e-07,
+      "loss": 0.0891,
+      "step": 511
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3110651313704324,
+      "learning_rate": 9.999820713972737e-07,
+      "loss": 0.0974,
+      "step": 512
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.23108817473383325,
+      "learning_rate": 9.999811861713998e-07,
+      "loss": 0.2463,
+      "step": 513
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.15727001210456104,
+      "learning_rate": 9.999802796153234e-07,
+      "loss": 0.015,
+      "step": 514
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.26637691734140634,
+      "learning_rate": 9.99979351729083e-07,
+      "loss": 0.3334,
+      "step": 515
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2124085837214984,
+      "learning_rate": 9.999784025127185e-07,
+      "loss": 0.0869,
+      "step": 516
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.199038642426334,
+      "learning_rate": 9.999774319662703e-07,
+      "loss": 0.1575,
+      "step": 517
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.27989246765365144,
+      "learning_rate": 9.999764400897798e-07,
+      "loss": 0.0992,
+      "step": 518
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.24058349608902685,
+      "learning_rate": 9.999754268832893e-07,
+      "loss": 0.4741,
+      "step": 519
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.21542311327582966,
+      "learning_rate": 9.99974392346842e-07,
+      "loss": 0.2214,
+      "step": 520
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3549191787539132,
+      "learning_rate": 9.999733364804818e-07,
+      "loss": 0.0226,
+      "step": 521
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2232365983566046,
+      "learning_rate": 9.999722592842543e-07,
+      "loss": 0.1226,
+      "step": 522
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2583979267337865,
+      "learning_rate": 9.999711607582052e-07,
+      "loss": 0.2159,
+      "step": 523
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2723403267606881,
+      "learning_rate": 9.99970040902381e-07,
+      "loss": 0.3245,
+      "step": 524
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2542485452458295,
+      "learning_rate": 9.999688997168301e-07,
+      "loss": 0.2327,
+      "step": 525
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.41809755135138027,
+      "learning_rate": 9.999677372016007e-07,
+      "loss": 0.1287,
+      "step": 526
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.27112163543158724,
+      "learning_rate": 9.999665533567426e-07,
+      "loss": 0.3427,
+      "step": 527
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.26191221466291176,
+      "learning_rate": 9.999653481823063e-07,
+      "loss": 0.2782,
+      "step": 528
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2102654999678496,
+      "learning_rate": 9.99964121678343e-07,
+      "loss": 0.3097,
+      "step": 529
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.11716270764715653,
+      "learning_rate": 9.999628738449055e-07,
+      "loss": 0.0224,
+      "step": 530
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2342432126524212,
+      "learning_rate": 9.999616046820466e-07,
+      "loss": 0.1761,
+      "step": 531
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.528700125502073,
+      "learning_rate": 9.999603141898207e-07,
+      "loss": 0.3711,
+      "step": 532
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.22774259370215127,
+      "learning_rate": 9.999590023682826e-07,
+      "loss": 0.1242,
+      "step": 533
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.25913121938300426,
+      "learning_rate": 9.999576692174884e-07,
+      "loss": 0.3121,
+      "step": 534
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2695919605172889,
+      "learning_rate": 9.999563147374952e-07,
+      "loss": 0.2923,
+      "step": 535
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17673669493872907,
+      "learning_rate": 9.999549389283606e-07,
+      "loss": 0.135,
+      "step": 536
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.23203256134608177,
+      "learning_rate": 9.99953541790143e-07,
+      "loss": 0.0209,
+      "step": 537
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.39836612884069494,
+      "learning_rate": 9.999521233229025e-07,
+      "loss": 0.351,
+      "step": 538
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.34472204539733897,
+      "learning_rate": 9.999506835266992e-07,
+      "loss": 0.2228,
+      "step": 539
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.264377801633734,
+      "learning_rate": 9.999492224015948e-07,
+      "loss": 0.0628,
+      "step": 540
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.20910132050425237,
+      "learning_rate": 9.999477399476516e-07,
+      "loss": 0.1251,
+      "step": 541
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.23636435994906232,
+      "learning_rate": 9.999462361649327e-07,
+      "loss": 0.2616,
+      "step": 542
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.40297277864631065,
+      "learning_rate": 9.999447110535024e-07,
+      "loss": 0.3826,
+      "step": 543
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.28148856030908415,
+      "learning_rate": 9.999431646134257e-07,
+      "loss": 0.1426,
+      "step": 544
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.26535308731379725,
+      "learning_rate": 9.999415968447687e-07,
+      "loss": 0.4973,
+      "step": 545
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.14706290474883787,
+      "learning_rate": 9.99940007747598e-07,
+      "loss": 0.1553,
+      "step": 546
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17983524372845375,
+      "learning_rate": 9.999383973219816e-07,
+      "loss": 0.2053,
+      "step": 547
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1621418678808146,
+      "learning_rate": 9.999367655679881e-07,
+      "loss": 0.0513,
+      "step": 548
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.06329448454305941,
+      "learning_rate": 9.999351124856873e-07,
+      "loss": 0.0875,
+      "step": 549
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.20894730328341823,
+      "learning_rate": 9.999334380751497e-07,
+      "loss": 0.0524,
+      "step": 550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3801867943102373,
+      "learning_rate": 9.999317423364465e-07,
+      "loss": 0.2819,
+      "step": 551
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.12575459885606005,
+      "learning_rate": 9.9993002526965e-07,
+      "loss": 0.174,
+      "step": 552
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.19373342460912157,
+      "learning_rate": 9.99928286874834e-07,
+      "loss": 0.2381,
+      "step": 553
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.290343647970398,
+      "learning_rate": 9.99926527152072e-07,
+      "loss": 0.2266,
+      "step": 554
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.21789424364573587,
+      "learning_rate": 9.999247461014395e-07,
+      "loss": 0.1309,
+      "step": 555
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3589600866174122,
+      "learning_rate": 9.999229437230123e-07,
+      "loss": 0.1268,
+      "step": 556
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.32039536701754084,
+      "learning_rate": 9.999211200168675e-07,
+      "loss": 0.0894,
+      "step": 557
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.602307055588835,
+      "learning_rate": 9.999192749830827e-07,
+      "loss": 0.2676,
+      "step": 558
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1731340196071816,
+      "learning_rate": 9.999174086217368e-07,
+      "loss": 0.2218,
+      "step": 559
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.21485030799997704,
+      "learning_rate": 9.99915520932909e-07,
+      "loss": 0.197,
+      "step": 560
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.09116657996386676,
+      "learning_rate": 9.999136119166803e-07,
+      "loss": 0.0076,
+      "step": 561
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2595362787199046,
+      "learning_rate": 9.999116815731318e-07,
+      "loss": 0.1199,
+      "step": 562
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2943434903221813,
+      "learning_rate": 9.999097299023463e-07,
+      "loss": 0.2567,
+      "step": 563
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2549772125656584,
+      "learning_rate": 9.999077569044066e-07,
+      "loss": 0.2859,
+      "step": 564
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.11980427571428975,
+      "learning_rate": 9.999057625793969e-07,
+      "loss": 0.0486,
+      "step": 565
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.23365723416624679,
+      "learning_rate": 9.999037469274026e-07,
+      "loss": 0.1725,
+      "step": 566
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.07751232484547327,
+      "learning_rate": 9.999017099485095e-07,
+      "loss": 0.0058,
+      "step": 567
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.47467674056310083,
+      "learning_rate": 9.998996516428045e-07,
+      "loss": 0.2391,
+      "step": 568
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.30761449261074536,
+      "learning_rate": 9.998975720103756e-07,
+      "loss": 0.2047,
+      "step": 569
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.18044750339675913,
+      "learning_rate": 9.99895471051311e-07,
+      "loss": 0.2432,
+      "step": 570
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.28473339587930035,
+      "learning_rate": 9.99893348765701e-07,
+      "loss": 0.1458,
+      "step": 571
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.20667850273765226,
+      "learning_rate": 9.998912051536358e-07,
+      "loss": 0.1582,
+      "step": 572
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3232007620220984,
+      "learning_rate": 9.998890402152067e-07,
+      "loss": 0.2699,
+      "step": 573
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2747885272567359,
+      "learning_rate": 9.998868539505065e-07,
+      "loss": 0.061,
+      "step": 574
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3028579312727913,
+      "learning_rate": 9.99884646359628e-07,
+      "loss": 0.4315,
+      "step": 575
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.26332033672438776,
+      "learning_rate": 9.998824174426656e-07,
+      "loss": 0.0787,
+      "step": 576
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.35321268776480974,
+      "learning_rate": 9.998801671997146e-07,
+      "loss": 0.3332,
+      "step": 577
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.24897210129496727,
+      "learning_rate": 9.998778956308707e-07,
+      "loss": 0.0626,
+      "step": 578
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.47698351985136805,
+      "learning_rate": 9.998756027362308e-07,
+      "loss": 0.3223,
+      "step": 579
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.19134790383960557,
+      "learning_rate": 9.998732885158927e-07,
+      "loss": 0.2695,
+      "step": 580
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4879603420778916,
+      "learning_rate": 9.998709529699555e-07,
+      "loss": 0.1466,
+      "step": 581
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3128280251335007,
+      "learning_rate": 9.998685960985186e-07,
+      "loss": 0.4176,
+      "step": 582
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1838525897173844,
+      "learning_rate": 9.998662179016821e-07,
+      "loss": 0.1679,
+      "step": 583
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.19676301033745802,
+      "learning_rate": 9.998638183795483e-07,
+      "loss": 0.2997,
+      "step": 584
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1403235253701055,
+      "learning_rate": 9.99861397532219e-07,
+      "loss": 0.1538,
+      "step": 585
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2713200681540278,
+      "learning_rate": 9.998589553597976e-07,
+      "loss": 0.2481,
+      "step": 586
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.31210448436474636,
+      "learning_rate": 9.998564918623884e-07,
+      "loss": 0.3303,
+      "step": 587
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.19441767560085377,
+      "learning_rate": 9.998540070400965e-07,
+      "loss": 0.1644,
+      "step": 588
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1307191376988514,
+      "learning_rate": 9.998515008930277e-07,
+      "loss": 0.202,
+      "step": 589
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7254158512474338,
+      "learning_rate": 9.99848973421289e-07,
+      "loss": 0.3441,
+      "step": 590
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.27264410203686634,
+      "learning_rate": 9.998464246249883e-07,
+      "loss": 0.2037,
+      "step": 591
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2911967943345156,
+      "learning_rate": 9.998438545042345e-07,
+      "loss": 0.0963,
+      "step": 592
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.327683277210842,
+      "learning_rate": 9.998412630591369e-07,
+      "loss": 0.1515,
+      "step": 593
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2219327938303094,
+      "learning_rate": 9.998386502898061e-07,
+      "loss": 0.1894,
+      "step": 594
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.18622294921364274,
+      "learning_rate": 9.99836016196354e-07,
+      "loss": 0.1695,
+      "step": 595
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2925134602430663,
+      "learning_rate": 9.998333607788922e-07,
+      "loss": 0.1408,
+      "step": 596
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.40499870275179556,
+      "learning_rate": 9.998306840375349e-07,
+      "loss": 0.1963,
+      "step": 597
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.40439244999887497,
+      "learning_rate": 9.998279859723954e-07,
+      "loss": 0.2666,
+      "step": 598
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.30494083909444997,
+      "learning_rate": 9.998252665835895e-07,
+      "loss": 0.1802,
+      "step": 599
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.29650189969662283,
+      "learning_rate": 9.99822525871233e-07,
+      "loss": 0.1154,
+      "step": 600
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3300319194535553,
+      "learning_rate": 9.998197638354427e-07,
+      "loss": 0.0357,
+      "step": 601
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.29748899925065037,
+      "learning_rate": 9.998169804763365e-07,
+      "loss": 0.3921,
+      "step": 602
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2842270442567387,
+      "learning_rate": 9.99814175794033e-07,
+      "loss": 0.1997,
+      "step": 603
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.17694701192442455,
+      "learning_rate": 9.998113497886522e-07,
+      "loss": 0.0825,
+      "step": 604
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.19775297135044023,
+      "learning_rate": 9.998085024603144e-07,
+      "loss": 0.2962,
+      "step": 605
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4098023299487314,
+      "learning_rate": 9.998056338091413e-07,
+      "loss": 0.1432,
+      "step": 606
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2230885301929253,
+      "learning_rate": 9.99802743835255e-07,
+      "loss": 0.3254,
+      "step": 607
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.12637777237273629,
+      "learning_rate": 9.997998325387788e-07,
+      "loss": 0.0687,
+      "step": 608
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3196469626868912,
+      "learning_rate": 9.997968999198371e-07,
+      "loss": 0.0266,
+      "step": 609
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2310289775253604,
+      "learning_rate": 9.997939459785552e-07,
+      "loss": 0.2164,
+      "step": 610
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.26611036284271894,
+      "learning_rate": 9.997909707150584e-07,
+      "loss": 0.3605,
+      "step": 611
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2614133341008609,
+      "learning_rate": 9.997879741294744e-07,
+      "loss": 0.0632,
+      "step": 612
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.26078759864867007,
+      "learning_rate": 9.997849562219307e-07,
+      "loss": 0.1604,
+      "step": 613
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.44154590338271715,
+      "learning_rate": 9.99781916992556e-07,
+      "loss": 0.2396,
+      "step": 614
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14067566349075272,
+      "learning_rate": 9.9977885644148e-07,
+      "loss": 0.2071,
+      "step": 615
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3575616851456624,
+      "learning_rate": 9.997757745688334e-07,
+      "loss": 0.3583,
+      "step": 616
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.33687583636678,
+      "learning_rate": 9.997726713747475e-07,
+      "loss": 0.1614,
+      "step": 617
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1287137337645716,
+      "learning_rate": 9.997695468593547e-07,
+      "loss": 0.0285,
+      "step": 618
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.22629252964500782,
+      "learning_rate": 9.997664010227885e-07,
+      "loss": 0.1418,
+      "step": 619
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2244281138166755,
+      "learning_rate": 9.997632338651828e-07,
+      "loss": 0.2675,
+      "step": 620
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4123947347059294,
+      "learning_rate": 9.997600453866732e-07,
+      "loss": 0.3591,
+      "step": 621
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.11262626115015796,
+      "learning_rate": 9.997568355873953e-07,
+      "loss": 0.1854,
+      "step": 622
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.19934916889725812,
+      "learning_rate": 9.997536044674862e-07,
+      "loss": 0.0942,
+      "step": 623
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4240857699247332,
+      "learning_rate": 9.997503520270835e-07,
+      "loss": 0.1083,
+      "step": 624
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2192476132653266,
+      "learning_rate": 9.997470782663262e-07,
+      "loss": 0.2008,
+      "step": 625
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.47564784426711987,
+      "learning_rate": 9.99743783185354e-07,
+      "loss": 0.1532,
+      "step": 626
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.33507348750716026,
+      "learning_rate": 9.997404667843074e-07,
+      "loss": 0.0814,
+      "step": 627
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.20428464002313707,
+      "learning_rate": 9.997371290633278e-07,
+      "loss": 0.3139,
+      "step": 628
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3458666228430019,
+      "learning_rate": 9.997337700225578e-07,
+      "loss": 0.1347,
+      "step": 629
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1155058198126477,
+      "learning_rate": 9.997303896621404e-07,
+      "loss": 0.1214,
+      "step": 630
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2868224800278387,
+      "learning_rate": 9.997269879822204e-07,
+      "loss": 0.1442,
+      "step": 631
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.21098114298120535,
+      "learning_rate": 9.99723564982942e-07,
+      "loss": 0.4881,
+      "step": 632
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.28539541990386547,
+      "learning_rate": 9.99720120664452e-07,
+      "loss": 0.212,
+      "step": 633
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.15762416827649048,
+      "learning_rate": 9.997166550268972e-07,
+      "loss": 0.0298,
+      "step": 634
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.17249326656550196,
+      "learning_rate": 9.997131680704251e-07,
+      "loss": 0.1836,
+      "step": 635
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.26370313587358046,
+      "learning_rate": 9.99709659795185e-07,
+      "loss": 0.3679,
+      "step": 636
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.23461575570233473,
+      "learning_rate": 9.99706130201326e-07,
+      "loss": 0.2692,
+      "step": 637
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2979758421620129,
+      "learning_rate": 9.99702579288999e-07,
+      "loss": 0.2391,
+      "step": 638
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1770101807416079,
+      "learning_rate": 9.996990070583555e-07,
+      "loss": 0.2021,
+      "step": 639
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3250207900693622,
+      "learning_rate": 9.996954135095478e-07,
+      "loss": 0.1972,
+      "step": 640
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.29805102006964246,
+      "learning_rate": 9.996917986427293e-07,
+      "loss": 0.1832,
+      "step": 641
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.34048618804477215,
+      "learning_rate": 9.99688162458054e-07,
+      "loss": 0.0834,
+      "step": 642
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.276125773326266,
+      "learning_rate": 9.996845049556775e-07,
+      "loss": 0.0967,
+      "step": 643
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3643296711012014,
+      "learning_rate": 9.996808261357553e-07,
+      "loss": 0.4839,
+      "step": 644
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3594869795982038,
+      "learning_rate": 9.996771259984446e-07,
+      "loss": 0.2085,
+      "step": 645
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.25612813685311064,
+      "learning_rate": 9.996734045439032e-07,
+      "loss": 0.2249,
+      "step": 646
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5294491242173487,
+      "learning_rate": 9.9966966177229e-07,
+      "loss": 0.2232,
+      "step": 647
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.223394109480478,
+      "learning_rate": 9.996658976837644e-07,
+      "loss": 0.2066,
+      "step": 648
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0964490455397499,
+      "learning_rate": 9.996621122784872e-07,
+      "loss": 0.0176,
+      "step": 649
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9916404275370678,
+      "learning_rate": 9.996583055566199e-07,
+      "loss": 0.2847,
+      "step": 650
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4085104775813451,
+      "learning_rate": 9.99654477518325e-07,
+      "loss": 0.1965,
+      "step": 651
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.26930592600432185,
+      "learning_rate": 9.996506281637653e-07,
+      "loss": 0.1166,
+      "step": 652
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.20570651628474856,
+      "learning_rate": 9.996467574931058e-07,
+      "loss": 0.2854,
+      "step": 653
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.23019324170520708,
+      "learning_rate": 9.996428655065108e-07,
+      "loss": 0.238,
+      "step": 654
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.28456364609373247,
+      "learning_rate": 9.99638952204147e-07,
+      "loss": 0.3202,
+      "step": 655
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.33711673215738003,
+      "learning_rate": 9.996350175861809e-07,
+      "loss": 0.2112,
+      "step": 656
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2628481495394333,
+      "learning_rate": 9.996310616527808e-07,
+      "loss": 0.097,
+      "step": 657
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14330502365215603,
+      "learning_rate": 9.99627084404115e-07,
+      "loss": 0.1624,
+      "step": 658
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14064107397795123,
+      "learning_rate": 9.996230858403536e-07,
+      "loss": 0.0913,
+      "step": 659
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3607971179922472,
+      "learning_rate": 9.99619065961667e-07,
+      "loss": 0.0854,
+      "step": 660
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.23005910265204066,
+      "learning_rate": 9.996150247682265e-07,
+      "loss": 0.2756,
+      "step": 661
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.24785734359247824,
+      "learning_rate": 9.996109622602048e-07,
+      "loss": 0.1309,
+      "step": 662
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4946815559380326,
+      "learning_rate": 9.99606878437775e-07,
+      "loss": 0.3128,
+      "step": 663
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.18161354687738152,
+      "learning_rate": 9.996027733011117e-07,
+      "loss": 0.1091,
+      "step": 664
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.40597640467228835,
+      "learning_rate": 9.995986468503894e-07,
+      "loss": 0.3422,
+      "step": 665
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.10124882585850081,
+      "learning_rate": 9.995944990857847e-07,
+      "loss": 0.1157,
+      "step": 666
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.31196022827667125,
+      "learning_rate": 9.995903300074744e-07,
+      "loss": 0.3875,
+      "step": 667
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.31771538807403404,
+      "learning_rate": 9.995861396156362e-07,
+      "loss": 0.21,
+      "step": 668
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3493822437991453,
+      "learning_rate": 9.995819279104491e-07,
+      "loss": 0.0353,
+      "step": 669
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4421172995145604,
+      "learning_rate": 9.995776948920927e-07,
+      "loss": 0.1876,
+      "step": 670
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.19443582410038723,
+      "learning_rate": 9.995734405607474e-07,
+      "loss": 0.0966,
+      "step": 671
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3275481656952158,
+      "learning_rate": 9.99569164916595e-07,
+      "loss": 0.4793,
+      "step": 672
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.22924677843911015,
+      "learning_rate": 9.995648679598176e-07,
+      "loss": 0.2181,
+      "step": 673
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3008393077777119,
+      "learning_rate": 9.99560549690599e-07,
+      "loss": 0.4154,
+      "step": 674
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5001730906408609,
+      "learning_rate": 9.99556210109123e-07,
+      "loss": 0.0971,
+      "step": 675
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.25220347513813307,
+      "learning_rate": 9.995518492155746e-07,
+      "loss": 0.1822,
+      "step": 676
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.21326006771705913,
+      "learning_rate": 9.9954746701014e-07,
+      "loss": 0.0362,
+      "step": 677
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.12982483839799883,
+      "learning_rate": 9.995430634930066e-07,
+      "loss": 0.1691,
+      "step": 678
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.33526482948055236,
+      "learning_rate": 9.99538638664362e-07,
+      "loss": 0.5494,
+      "step": 679
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.20965784744111918,
+      "learning_rate": 9.995341925243944e-07,
+      "loss": 0.1009,
+      "step": 680
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4229254242252686,
+      "learning_rate": 9.99529725073294e-07,
+      "loss": 0.1194,
+      "step": 681
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.08776084595143485,
+      "learning_rate": 9.995252363112518e-07,
+      "loss": 0.065,
+      "step": 682
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.16284348597463982,
+      "learning_rate": 9.995207262384585e-07,
+      "loss": 0.0518,
+      "step": 683
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.20175325495331153,
+      "learning_rate": 9.995161948551069e-07,
+      "loss": 0.0596,
+      "step": 684
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41025397147531095,
+      "learning_rate": 9.995116421613905e-07,
+      "loss": 0.2243,
+      "step": 685
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.27349920467101396,
+      "learning_rate": 9.995070681575032e-07,
+      "loss": 0.3185,
+      "step": 686
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.39644977768940787,
+      "learning_rate": 9.995024728436401e-07,
+      "loss": 0.2763,
+      "step": 687
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1399900794362018,
+      "learning_rate": 9.994978562199973e-07,
+      "loss": 0.0156,
+      "step": 688
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3472303888369003,
+      "learning_rate": 9.99493218286772e-07,
+      "loss": 0.182,
+      "step": 689
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2636485704732172,
+      "learning_rate": 9.99488559044162e-07,
+      "loss": 0.159,
+      "step": 690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.44025629657577875,
+      "learning_rate": 9.994838784923657e-07,
+      "loss": 0.13,
+      "step": 691
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5309811169145624,
+      "learning_rate": 9.994791766315833e-07,
+      "loss": 0.2435,
+      "step": 692
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3815430042476451,
+      "learning_rate": 9.99474453462015e-07,
+      "loss": 0.0291,
+      "step": 693
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6649567885730471,
+      "learning_rate": 9.994697089838626e-07,
+      "loss": 0.068,
+      "step": 694
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3355111021093103,
+      "learning_rate": 9.994649431973283e-07,
+      "loss": 0.0949,
+      "step": 695
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.22658701080249202,
+      "learning_rate": 9.994601561026153e-07,
+      "loss": 0.0596,
+      "step": 696
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.23806279247513487,
+      "learning_rate": 9.994553476999281e-07,
+      "loss": 0.2143,
+      "step": 697
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3346821326673875,
+      "learning_rate": 9.994505179894718e-07,
+      "loss": 0.2204,
+      "step": 698
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3255836963757066,
+      "learning_rate": 9.994456669714523e-07,
+      "loss": 0.117,
+      "step": 699
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.17276392954244343,
+      "learning_rate": 9.994407946460768e-07,
+      "loss": 0.0316,
+      "step": 700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.40489743890177804,
+      "learning_rate": 9.99435901013553e-07,
+      "loss": 0.1418,
+      "step": 701
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5784755364483929,
+      "learning_rate": 9.994309860740894e-07,
+      "loss": 0.2431,
+      "step": 702
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41512771975633733,
+      "learning_rate": 9.994260498278962e-07,
+      "loss": 0.2355,
+      "step": 703
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3187904283921282,
+      "learning_rate": 9.994210922751836e-07,
+      "loss": 0.3454,
+      "step": 704
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.10930687596375602,
+      "learning_rate": 9.994161134161632e-07,
+      "loss": 0.0869,
+      "step": 705
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.21349147939129332,
+      "learning_rate": 9.994111132510477e-07,
+      "loss": 0.0998,
+      "step": 706
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.29174491502664607,
+      "learning_rate": 9.994060917800499e-07,
+      "loss": 0.1836,
+      "step": 707
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.267973293440664,
+      "learning_rate": 9.994010490033843e-07,
+      "loss": 0.0963,
+      "step": 708
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2959997574431314,
+      "learning_rate": 9.993959849212662e-07,
+      "loss": 0.1727,
+      "step": 709
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.37866755142864683,
+      "learning_rate": 9.993908995339114e-07,
+      "loss": 0.3663,
+      "step": 710
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2138710941633362,
+      "learning_rate": 9.993857928415368e-07,
+      "loss": 0.2025,
+      "step": 711
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3630054849873252,
+      "learning_rate": 9.993806648443606e-07,
+      "loss": 0.3173,
+      "step": 712
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.21645649187262386,
+      "learning_rate": 9.993755155426014e-07,
+      "loss": 0.1201,
+      "step": 713
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4298086689210384,
+      "learning_rate": 9.993703449364785e-07,
+      "loss": 0.2854,
+      "step": 714
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6737470060141839,
+      "learning_rate": 9.99365153026213e-07,
+      "loss": 0.2807,
+      "step": 715
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46771804450657134,
+      "learning_rate": 9.993599398120263e-07,
+      "loss": 0.3458,
+      "step": 716
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13088635700932683,
+      "learning_rate": 9.993547052941407e-07,
+      "loss": 0.1335,
+      "step": 717
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.38230493975198065,
+      "learning_rate": 9.993494494727795e-07,
+      "loss": 0.2078,
+      "step": 718
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5303441806830073,
+      "learning_rate": 9.99344172348167e-07,
+      "loss": 0.3254,
+      "step": 719
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.41501870663594365,
+      "learning_rate": 9.993388739205283e-07,
+      "loss": 0.2984,
+      "step": 720
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4466052256901839,
+      "learning_rate": 9.993335541900893e-07,
+      "loss": 0.326,
+      "step": 721
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3664984275313783,
+      "learning_rate": 9.993282131570772e-07,
+      "loss": 0.408,
+      "step": 722
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.42448567094972184,
+      "learning_rate": 9.9932285082172e-07,
+      "loss": 0.2234,
+      "step": 723
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.17745007270794735,
+      "learning_rate": 9.99317467184246e-07,
+      "loss": 0.1807,
+      "step": 724
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44935936650217406,
+      "learning_rate": 9.993120622448849e-07,
+      "loss": 0.2376,
+      "step": 725
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6126905176000934,
+      "learning_rate": 9.993066360038678e-07,
+      "loss": 0.2615,
+      "step": 726
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7901474797504096,
+      "learning_rate": 9.993011884614256e-07,
+      "loss": 0.0211,
+      "step": 727
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.26384046180651166,
+      "learning_rate": 9.99295719617791e-07,
+      "loss": 0.1528,
+      "step": 728
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1624837760476425,
+      "learning_rate": 9.992902294731974e-07,
+      "loss": 0.0162,
+      "step": 729
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3153028508242539,
+      "learning_rate": 9.992847180278792e-07,
+      "loss": 0.3626,
+      "step": 730
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3489224986364525,
+      "learning_rate": 9.992791852820708e-07,
+      "loss": 0.1076,
+      "step": 731
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.38130284266583797,
+      "learning_rate": 9.992736312360089e-07,
+      "loss": 0.3346,
+      "step": 732
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8070432281761301,
+      "learning_rate": 9.992680558899303e-07,
+      "loss": 0.3271,
+      "step": 733
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.43492179476209475,
+      "learning_rate": 9.992624592440725e-07,
+      "loss": 0.2262,
+      "step": 734
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15688682197524684,
+      "learning_rate": 9.992568412986748e-07,
+      "loss": 0.1294,
+      "step": 735
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.32013111372934266,
+      "learning_rate": 9.992512020539765e-07,
+      "loss": 0.2633,
+      "step": 736
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3857147256820132,
+      "learning_rate": 9.992455415102182e-07,
+      "loss": 0.2661,
+      "step": 737
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4282079559143247,
+      "learning_rate": 9.992398596676417e-07,
+      "loss": 0.1998,
+      "step": 738
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.08828878270340566,
+      "learning_rate": 9.99234156526489e-07,
+      "loss": 0.0104,
+      "step": 739
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3516698478793413,
+      "learning_rate": 9.992284320870037e-07,
+      "loss": 0.045,
+      "step": 740
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2377456378438165,
+      "learning_rate": 9.9922268634943e-07,
+      "loss": 0.141,
+      "step": 741
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3916903296743999,
+      "learning_rate": 9.992169193140127e-07,
+      "loss": 0.2449,
+      "step": 742
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2735201883748301,
+      "learning_rate": 9.992111309809982e-07,
+      "loss": 0.35,
+      "step": 743
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3854829590577924,
+      "learning_rate": 9.992053213506333e-07,
+      "loss": 0.1347,
+      "step": 744
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2756410028217851,
+      "learning_rate": 9.99199490423166e-07,
+      "loss": 0.0884,
+      "step": 745
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.27719372195536185,
+      "learning_rate": 9.991936381988447e-07,
+      "loss": 0.1471,
+      "step": 746
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.21292987230744306,
+      "learning_rate": 9.991877646779194e-07,
+      "loss": 0.2303,
+      "step": 747
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.561405462595992,
+      "learning_rate": 9.991818698606404e-07,
+      "loss": 0.1685,
+      "step": 748
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.26414580347221267,
+      "learning_rate": 9.991759537472597e-07,
+      "loss": 0.3096,
+      "step": 749
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.25524290308172876,
+      "learning_rate": 9.99170016338029e-07,
+      "loss": 0.1581,
+      "step": 750
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.148642376017468,
+      "learning_rate": 9.99164057633202e-07,
+      "loss": 0.4003,
+      "step": 751
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3809899101312142,
+      "learning_rate": 9.99158077633033e-07,
+      "loss": 0.3427,
+      "step": 752
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.335677292084092,
+      "learning_rate": 9.99152076337777e-07,
+      "loss": 0.2826,
+      "step": 753
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3201739427658402,
+      "learning_rate": 9.9914605374769e-07,
+      "loss": 0.367,
+      "step": 754
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4314453049894622,
+      "learning_rate": 9.991400098630288e-07,
+      "loss": 0.3241,
+      "step": 755
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.39924962518783,
+      "learning_rate": 9.991339446840515e-07,
+      "loss": 0.1677,
+      "step": 756
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.23400593462214092,
+      "learning_rate": 9.991278582110166e-07,
+      "loss": 0.2681,
+      "step": 757
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8307402139466588,
+      "learning_rate": 9.99121750444184e-07,
+      "loss": 0.1325,
+      "step": 758
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.43037593280946795,
+      "learning_rate": 9.99115621383814e-07,
+      "loss": 0.3028,
+      "step": 759
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2593830448653943,
+      "learning_rate": 9.991094710301686e-07,
+      "loss": 0.0571,
+      "step": 760
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.32735630380268255,
+      "learning_rate": 9.991032993835095e-07,
+      "loss": 0.0852,
+      "step": 761
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2900643412585144,
+      "learning_rate": 9.990971064441004e-07,
+      "loss": 0.2157,
+      "step": 762
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.29488411113941265,
+      "learning_rate": 9.990908922122056e-07,
+      "loss": 0.1244,
+      "step": 763
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0569396764700685,
+      "learning_rate": 9.990846566880899e-07,
+      "loss": 0.1896,
+      "step": 764
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.17865347993534295,
+      "learning_rate": 9.990783998720193e-07,
+      "loss": 0.0746,
+      "step": 765
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5269334730810578,
+      "learning_rate": 9.990721217642612e-07,
+      "loss": 0.0305,
+      "step": 766
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3051566717293546,
+      "learning_rate": 9.99065822365083e-07,
+      "loss": 0.314,
+      "step": 767
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9345007572272831,
+      "learning_rate": 9.990595016747535e-07,
+      "loss": 0.5108,
+      "step": 768
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5998423994556221,
+      "learning_rate": 9.990531596935424e-07,
+      "loss": 0.3033,
+      "step": 769
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45181850336453955,
+      "learning_rate": 9.990467964217204e-07,
+      "loss": 0.2796,
+      "step": 770
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4850613814989678,
+      "learning_rate": 9.990404118595587e-07,
+      "loss": 0.1074,
+      "step": 771
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.18496526534645386,
+      "learning_rate": 9.990340060073302e-07,
+      "loss": 0.1308,
+      "step": 772
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.19599395432566782,
+      "learning_rate": 9.990275788653074e-07,
+      "loss": 0.1107,
+      "step": 773
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.38743758561766645,
+      "learning_rate": 9.99021130433765e-07,
+      "loss": 0.281,
+      "step": 774
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.29353828597777165,
+      "learning_rate": 9.99014660712978e-07,
+      "loss": 0.1374,
+      "step": 775
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.38702552932139134,
+      "learning_rate": 9.990081697032226e-07,
+      "loss": 0.3723,
+      "step": 776
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5511249787455814,
+      "learning_rate": 9.990016574047755e-07,
+      "loss": 0.3179,
+      "step": 777
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.19069921374797308,
+      "learning_rate": 9.989951238179146e-07,
+      "loss": 0.1863,
+      "step": 778
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5477637035431495,
+      "learning_rate": 9.989885689429187e-07,
+      "loss": 0.2018,
+      "step": 779
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.23073750329632237,
+      "learning_rate": 9.989819927800671e-07,
+      "loss": 0.1418,
+      "step": 780
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.030750283996359234,
+      "learning_rate": 9.989753953296408e-07,
+      "loss": 0.0027,
+      "step": 781
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1145215007398251,
+      "learning_rate": 9.98968776591921e-07,
+      "loss": 0.147,
+      "step": 782
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.25940045290858127,
+      "learning_rate": 9.989621365671902e-07,
+      "loss": 0.0713,
+      "step": 783
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.43694900746613674,
+      "learning_rate": 9.989554752557314e-07,
+      "loss": 0.1064,
+      "step": 784
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3971010506393241,
+      "learning_rate": 9.989487926578291e-07,
+      "loss": 0.1655,
+      "step": 785
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1909020753793466,
+      "learning_rate": 9.989420887737683e-07,
+      "loss": 0.2461,
+      "step": 786
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3099298618303888,
+      "learning_rate": 9.989353636038351e-07,
+      "loss": 0.2658,
+      "step": 787
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9535739179506793,
+      "learning_rate": 9.989286171483161e-07,
+      "loss": 0.271,
+      "step": 788
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.36664103535127324,
+      "learning_rate": 9.989218494074995e-07,
+      "loss": 0.2692,
+      "step": 789
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.36033331083904596,
+      "learning_rate": 9.989150603816738e-07,
+      "loss": 0.1979,
+      "step": 790
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0563505863364127,
+      "learning_rate": 9.989082500711287e-07,
+      "loss": 0.2973,
+      "step": 791
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.35337153350301953,
+      "learning_rate": 9.989014184761546e-07,
+      "loss": 0.0985,
+      "step": 792
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.41319436850610153,
+      "learning_rate": 9.988945655970434e-07,
+      "loss": 0.1578,
+      "step": 793
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.428415309646363,
+      "learning_rate": 9.988876914340868e-07,
+      "loss": 0.3707,
+      "step": 794
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.216637732229521,
+      "learning_rate": 9.988807959875786e-07,
+      "loss": 0.0188,
+      "step": 795
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4181888867376164,
+      "learning_rate": 9.988738792578126e-07,
+      "loss": 0.2444,
+      "step": 796
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2000092320601304,
+      "learning_rate": 9.98866941245084e-07,
+      "loss": 0.1637,
+      "step": 797
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.41985990604917495,
+      "learning_rate": 9.98859981949689e-07,
+      "loss": 0.2563,
+      "step": 798
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3699535318356039,
+      "learning_rate": 9.988530013719243e-07,
+      "loss": 0.1085,
+      "step": 799
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.26470185162460874,
+      "learning_rate": 9.988459995120877e-07,
+      "loss": 0.1144,
+      "step": 800
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.40079513827952984,
+      "learning_rate": 9.988389763704778e-07,
+      "loss": 0.3005,
+      "step": 801
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7079814755560667,
+      "learning_rate": 9.988319319473947e-07,
+      "loss": 0.2295,
+      "step": 802
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.47031904416782294,
+      "learning_rate": 9.988248662431385e-07,
+      "loss": 0.4155,
+      "step": 803
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.35989836180944096,
+      "learning_rate": 9.988177792580106e-07,
+      "loss": 0.0439,
+      "step": 804
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.29549006075515255,
+      "learning_rate": 9.988106709923137e-07,
+      "loss": 0.09,
+      "step": 805
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.17598994399439483,
+      "learning_rate": 9.988035414463507e-07,
+      "loss": 0.0525,
+      "step": 806
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5363124292698066,
+      "learning_rate": 9.98796390620426e-07,
+      "loss": 0.3053,
+      "step": 807
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6041872608092101,
+      "learning_rate": 9.987892185148443e-07,
+      "loss": 0.3073,
+      "step": 808
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4189089818332621,
+      "learning_rate": 9.98782025129912e-07,
+      "loss": 0.3295,
+      "step": 809
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4753315947071798,
+      "learning_rate": 9.987748104659359e-07,
+      "loss": 0.2363,
+      "step": 810
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6384770671401014,
+      "learning_rate": 9.987675745232239e-07,
+      "loss": 0.0567,
+      "step": 811
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.39078600515760487,
+      "learning_rate": 9.987603173020842e-07,
+      "loss": 0.1865,
+      "step": 812
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.41160175106312996,
+      "learning_rate": 9.987530388028267e-07,
+      "loss": 0.2448,
+      "step": 813
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2468671563228391,
+      "learning_rate": 9.987457390257622e-07,
+      "loss": 0.1582,
+      "step": 814
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2780845450730514,
+      "learning_rate": 9.987384179712018e-07,
+      "loss": 0.1083,
+      "step": 815
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5634701131210977,
+      "learning_rate": 9.987310756394578e-07,
+      "loss": 0.2131,
+      "step": 816
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.378584012441739,
+      "learning_rate": 9.987237120308435e-07,
+      "loss": 0.2024,
+      "step": 817
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.24450289137278988,
+      "learning_rate": 9.987163271456733e-07,
+      "loss": 0.2023,
+      "step": 818
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5248509815392669,
+      "learning_rate": 9.987089209842618e-07,
+      "loss": 0.2578,
+      "step": 819
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3488439970835759,
+      "learning_rate": 9.987014935469254e-07,
+      "loss": 0.3219,
+      "step": 820
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4434052624292268,
+      "learning_rate": 9.986940448339807e-07,
+      "loss": 0.1275,
+      "step": 821
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1639172911274137,
+      "learning_rate": 9.986865748457455e-07,
+      "loss": 0.0444,
+      "step": 822
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3198579670432246,
+      "learning_rate": 9.986790835825385e-07,
+      "loss": 0.2233,
+      "step": 823
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6046038247945492,
+      "learning_rate": 9.986715710446795e-07,
+      "loss": 0.2182,
+      "step": 824
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5503062834527775,
+      "learning_rate": 9.986640372324887e-07,
+      "loss": 0.1813,
+      "step": 825
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3311415631719066,
+      "learning_rate": 9.986564821462876e-07,
+      "loss": 0.1052,
+      "step": 826
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3745468369706555,
+      "learning_rate": 9.986489057863986e-07,
+      "loss": 0.2697,
+      "step": 827
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4019031440647441,
+      "learning_rate": 9.986413081531448e-07,
+      "loss": 0.2565,
+      "step": 828
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.48819622858764106,
+      "learning_rate": 9.986336892468506e-07,
+      "loss": 0.158,
+      "step": 829
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.49255007594275985,
+      "learning_rate": 9.986260490678406e-07,
+      "loss": 0.2423,
+      "step": 830
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46966914868829196,
+      "learning_rate": 9.98618387616441e-07,
+      "loss": 0.3574,
+      "step": 831
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5241945042886459,
+      "learning_rate": 9.98610704892979e-07,
+      "loss": 0.1017,
+      "step": 832
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4420068530363283,
+      "learning_rate": 9.986030008977816e-07,
+      "loss": 0.1116,
+      "step": 833
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.42633745870809486,
+      "learning_rate": 9.98595275631178e-07,
+      "loss": 0.1198,
+      "step": 834
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.42117115114059506,
+      "learning_rate": 9.985875290934974e-07,
+      "loss": 0.2213,
+      "step": 835
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5886583825546735,
+      "learning_rate": 9.985797612850709e-07,
+      "loss": 0.2779,
+      "step": 836
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.17939168561944074,
+      "learning_rate": 9.985719722062293e-07,
+      "loss": 0.0299,
+      "step": 837
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.55192345774415,
+      "learning_rate": 9.98564161857305e-07,
+      "loss": 0.1841,
+      "step": 838
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9291312498876918,
+      "learning_rate": 9.985563302386317e-07,
+      "loss": 0.4546,
+      "step": 839
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6257171085637463,
+      "learning_rate": 9.985484773505427e-07,
+      "loss": 0.3799,
+      "step": 840
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6860559930564434,
+      "learning_rate": 9.985406031933737e-07,
+      "loss": 0.2958,
+      "step": 841
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6107884200040208,
+      "learning_rate": 9.985327077674604e-07,
+      "loss": 0.2403,
+      "step": 842
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5212449988275933,
+      "learning_rate": 9.985247910731395e-07,
+      "loss": 0.1581,
+      "step": 843
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5544588418030959,
+      "learning_rate": 9.985168531107488e-07,
+      "loss": 0.2562,
+      "step": 844
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9556498835020415,
+      "learning_rate": 9.985088938806271e-07,
+      "loss": 0.0674,
+      "step": 845
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.06436617935042317,
+      "learning_rate": 9.98500913383114e-07,
+      "loss": 0.0047,
+      "step": 846
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6433903061356261,
+      "learning_rate": 9.984929116185497e-07,
+      "loss": 0.4003,
+      "step": 847
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.18427332635185062,
+      "learning_rate": 9.984848885872756e-07,
+      "loss": 0.0985,
+      "step": 848
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.326773548187247,
+      "learning_rate": 9.984768442896341e-07,
+      "loss": 0.1337,
+      "step": 849
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7734791586844639,
+      "learning_rate": 9.984687787259683e-07,
+      "loss": 0.2146,
+      "step": 850
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4573907101064241,
+      "learning_rate": 9.984606918966226e-07,
+      "loss": 0.2659,
+      "step": 851
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7185429956776199,
+      "learning_rate": 9.984525838019415e-07,
+      "loss": 0.3559,
+      "step": 852
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5786571028021502,
+      "learning_rate": 9.984444544422712e-07,
+      "loss": 0.2182,
+      "step": 853
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.261193248528463,
+      "learning_rate": 9.984363038179587e-07,
+      "loss": 0.0158,
+      "step": 854
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1607690185548112,
+      "learning_rate": 9.984281319293513e-07,
+      "loss": 0.0183,
+      "step": 855
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.43391240870556813,
+      "learning_rate": 9.984199387767978e-07,
+      "loss": 0.3064,
+      "step": 856
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3421269906808885,
+      "learning_rate": 9.984117243606478e-07,
+      "loss": 0.2376,
+      "step": 857
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3914337813170445,
+      "learning_rate": 9.984034886812519e-07,
+      "loss": 0.1398,
+      "step": 858
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.22670431645526753,
+      "learning_rate": 9.983952317389609e-07,
+      "loss": 0.1208,
+      "step": 859
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.551276633488919,
+      "learning_rate": 9.983869535341276e-07,
+      "loss": 0.2405,
+      "step": 860
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5254185456481869,
+      "learning_rate": 9.98378654067105e-07,
+      "loss": 0.1736,
+      "step": 861
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1252888813245433,
+      "learning_rate": 9.98370333338247e-07,
+      "loss": 0.085,
+      "step": 862
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.24475631176376744,
+      "learning_rate": 9.98361991347909e-07,
+      "loss": 0.1545,
+      "step": 863
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8985437187589743,
+      "learning_rate": 9.983536280964463e-07,
+      "loss": 0.1804,
+      "step": 864
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5835744721955892,
+      "learning_rate": 9.983452435842161e-07,
+      "loss": 0.2285,
+      "step": 865
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4454473512156972,
+      "learning_rate": 9.98336837811576e-07,
+      "loss": 0.1476,
+      "step": 866
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5508829195580042,
+      "learning_rate": 9.983284107788849e-07,
+      "loss": 0.2139,
+      "step": 867
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46970387871543356,
+      "learning_rate": 9.983199624865019e-07,
+      "loss": 0.2151,
+      "step": 868
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3177141191160735,
+      "learning_rate": 9.983114929347875e-07,
+      "loss": 0.0975,
+      "step": 869
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7368498456487158,
+      "learning_rate": 9.98303002124103e-07,
+      "loss": 0.2882,
+      "step": 870
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4349333901871027,
+      "learning_rate": 9.982944900548106e-07,
+      "loss": 0.2377,
+      "step": 871
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14115678908192322,
+      "learning_rate": 9.982859567272738e-07,
+      "loss": 0.0091,
+      "step": 872
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.37967986469455783,
+      "learning_rate": 9.982774021418564e-07,
+      "loss": 0.2847,
+      "step": 873
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.10136774015629568,
+      "learning_rate": 9.982688262989235e-07,
+      "loss": 0.0098,
+      "step": 874
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3824427964589409,
+      "learning_rate": 9.982602291988404e-07,
+      "loss": 0.013,
+      "step": 875
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.749030640530025,
+      "learning_rate": 9.982516108419745e-07,
+      "loss": 0.2335,
+      "step": 876
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5183944961616913,
+      "learning_rate": 9.982429712286933e-07,
+      "loss": 0.2373,
+      "step": 877
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.24214020912983317,
+      "learning_rate": 9.982343103593654e-07,
+      "loss": 0.0894,
+      "step": 878
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4661896309772867,
+      "learning_rate": 9.982256282343603e-07,
+      "loss": 0.1716,
+      "step": 879
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.416058283139024,
+      "learning_rate": 9.98216924854048e-07,
+      "loss": 0.2509,
+      "step": 880
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0653442194630949,
+      "learning_rate": 9.982082002188004e-07,
+      "loss": 0.1884,
+      "step": 881
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45427821871010937,
+      "learning_rate": 9.981994543289895e-07,
+      "loss": 0.0431,
+      "step": 882
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4367716379240054,
+      "learning_rate": 9.981906871849884e-07,
+      "loss": 0.1209,
+      "step": 883
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1250590942922611,
+      "learning_rate": 9.981818987871708e-07,
+      "loss": 0.2175,
+      "step": 884
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31067847406138993,
+      "learning_rate": 9.98173089135912e-07,
+      "loss": 0.019,
+      "step": 885
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7127744823338241,
+      "learning_rate": 9.98164258231588e-07,
+      "loss": 0.4501,
+      "step": 886
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5454463047150027,
+      "learning_rate": 9.981554060745754e-07,
+      "loss": 0.2822,
+      "step": 887
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7384467024144157,
+      "learning_rate": 9.981465326652515e-07,
+      "loss": 0.1541,
+      "step": 888
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2491223286847421,
+      "learning_rate": 9.98137638003995e-07,
+      "loss": 0.1961,
+      "step": 889
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3060126226204753,
+      "learning_rate": 9.981287220911857e-07,
+      "loss": 0.1382,
+      "step": 890
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3969689694949935,
+      "learning_rate": 9.981197849272038e-07,
+      "loss": 0.2625,
+      "step": 891
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.27307865038414464,
+      "learning_rate": 9.981108265124303e-07,
+      "loss": 0.2458,
+      "step": 892
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6276549621281022,
+      "learning_rate": 9.981018468472478e-07,
+      "loss": 0.2624,
+      "step": 893
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3623102171189811,
+      "learning_rate": 9.98092845932039e-07,
+      "loss": 0.1402,
+      "step": 894
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6124999407880806,
+      "learning_rate": 9.980838237671883e-07,
+      "loss": 0.1638,
+      "step": 895
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7329529551589599,
+      "learning_rate": 9.980747803530804e-07,
+      "loss": 0.1128,
+      "step": 896
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.22111671966192745,
+      "learning_rate": 9.98065715690101e-07,
+      "loss": 0.1134,
+      "step": 897
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5603531292210565,
+      "learning_rate": 9.98056629778637e-07,
+      "loss": 0.117,
+      "step": 898
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1477932072556478,
+      "learning_rate": 9.98047522619076e-07,
+      "loss": 0.2562,
+      "step": 899
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5238308145115796,
+      "learning_rate": 9.980383942118065e-07,
+      "loss": 0.1441,
+      "step": 900
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5232892018074623,
+      "learning_rate": 9.980292445572179e-07,
+      "loss": 0.2693,
+      "step": 901
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5262895713718663,
+      "learning_rate": 9.980200736557004e-07,
+      "loss": 0.3365,
+      "step": 902
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.25404175535715173,
+      "learning_rate": 9.980108815076455e-07,
+      "loss": 0.0636,
+      "step": 903
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14217573604570202,
+      "learning_rate": 9.980016681134454e-07,
+      "loss": 0.0077,
+      "step": 904
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.27577195936313037,
+      "learning_rate": 9.979924334734929e-07,
+      "loss": 0.1928,
+      "step": 905
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44262819970244743,
+      "learning_rate": 9.979831775881819e-07,
+      "loss": 0.1663,
+      "step": 906
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44267162341035937,
+      "learning_rate": 9.979739004579077e-07,
+      "loss": 0.1427,
+      "step": 907
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6146554850708117,
+      "learning_rate": 9.97964602083066e-07,
+      "loss": 0.1161,
+      "step": 908
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9272958776949122,
+      "learning_rate": 9.979552824640531e-07,
+      "loss": 0.1589,
+      "step": 909
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.21303002979582747,
+      "learning_rate": 9.97945941601267e-07,
+      "loss": 0.0127,
+      "step": 910
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.41194094422912997,
+      "learning_rate": 9.979365794951058e-07,
+      "loss": 0.1445,
+      "step": 911
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46353461993783285,
+      "learning_rate": 9.979271961459694e-07,
+      "loss": 0.1449,
+      "step": 912
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.20054834219323922,
+      "learning_rate": 9.979177915542578e-07,
+      "loss": 0.0069,
+      "step": 913
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2708297641113638,
+      "learning_rate": 9.979083657203723e-07,
+      "loss": 0.1858,
+      "step": 914
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6109363257222682,
+      "learning_rate": 9.978989186447146e-07,
+      "loss": 0.1778,
+      "step": 915
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5028189692038627,
+      "learning_rate": 9.978894503276885e-07,
+      "loss": 0.1534,
+      "step": 916
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.465761413142227,
+      "learning_rate": 9.978799607696975e-07,
+      "loss": 0.0239,
+      "step": 917
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.379612151226625,
+      "learning_rate": 9.978704499711466e-07,
+      "loss": 0.296,
+      "step": 918
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.36307380563278896,
+      "learning_rate": 9.978609179324414e-07,
+      "loss": 0.1662,
+      "step": 919
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5121961709273829,
+      "learning_rate": 9.978513646539886e-07,
+      "loss": 0.1688,
+      "step": 920
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5762208409026472,
+      "learning_rate": 9.978417901361957e-07,
+      "loss": 0.4547,
+      "step": 921
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7750293229894867,
+      "learning_rate": 9.978321943794715e-07,
+      "loss": 0.2777,
+      "step": 922
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1923763827464195,
+      "learning_rate": 9.978225773842248e-07,
+      "loss": 0.065,
+      "step": 923
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.21859226925702638,
+      "learning_rate": 9.978129391508663e-07,
+      "loss": 0.1765,
+      "step": 924
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4779201707557506,
+      "learning_rate": 9.97803279679807e-07,
+      "loss": 0.4008,
+      "step": 925
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4703071639715682,
+      "learning_rate": 9.977935989714594e-07,
+      "loss": 0.1467,
+      "step": 926
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9150128592447756,
+      "learning_rate": 9.97783897026236e-07,
+      "loss": 0.4139,
+      "step": 927
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6190921533181195,
+      "learning_rate": 9.977741738445507e-07,
+      "loss": 0.3267,
+      "step": 928
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2143610086216308,
+      "learning_rate": 9.977644294268187e-07,
+      "loss": 0.1645,
+      "step": 929
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6172031300911652,
+      "learning_rate": 9.977546637734556e-07,
+      "loss": 0.1381,
+      "step": 930
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6070008665072504,
+      "learning_rate": 9.977448768848777e-07,
+      "loss": 0.426,
+      "step": 931
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5518807559900349,
+      "learning_rate": 9.977350687615027e-07,
+      "loss": 0.1151,
+      "step": 932
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5046020486044156,
+      "learning_rate": 9.977252394037492e-07,
+      "loss": 0.1508,
+      "step": 933
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.21914492471035532,
+      "learning_rate": 9.977153888120366e-07,
+      "loss": 0.1708,
+      "step": 934
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6464788062297534,
+      "learning_rate": 9.977055169867848e-07,
+      "loss": 0.3474,
+      "step": 935
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.33618861236827735,
+      "learning_rate": 9.976956239284151e-07,
+      "loss": 0.1884,
+      "step": 936
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6457467960715154,
+      "learning_rate": 9.976857096373499e-07,
+      "loss": 0.1886,
+      "step": 937
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6412063637630472,
+      "learning_rate": 9.976757741140115e-07,
+      "loss": 0.1398,
+      "step": 938
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4207704803199653,
+      "learning_rate": 9.976658173588243e-07,
+      "loss": 0.2348,
+      "step": 939
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46616906548217213,
+      "learning_rate": 9.976558393722127e-07,
+      "loss": 0.338,
+      "step": 940
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4083200801513483,
+      "learning_rate": 9.976458401546028e-07,
+      "loss": 0.1094,
+      "step": 941
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5406999137425461,
+      "learning_rate": 9.97635819706421e-07,
+      "loss": 0.1616,
+      "step": 942
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3111717063994098,
+      "learning_rate": 9.976257780280945e-07,
+      "loss": 0.2098,
+      "step": 943
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.587578293857706,
+      "learning_rate": 9.976157151200522e-07,
+      "loss": 0.1851,
+      "step": 944
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3849442286716842,
+      "learning_rate": 9.97605630982723e-07,
+      "loss": 0.3162,
+      "step": 945
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.40198988524598256,
+      "learning_rate": 9.975955256165374e-07,
+      "loss": 0.3014,
+      "step": 946
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.28406903973403613,
+      "learning_rate": 9.975853990219264e-07,
+      "loss": 0.053,
+      "step": 947
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2876604500665912,
+      "learning_rate": 9.975752511993218e-07,
+      "loss": 0.1911,
+      "step": 948
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46743378594402885,
+      "learning_rate": 9.97565082149157e-07,
+      "loss": 0.2186,
+      "step": 949
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6255651517071519,
+      "learning_rate": 9.975548918718653e-07,
+      "loss": 0.2079,
+      "step": 950
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5948784357183413,
+      "learning_rate": 9.975446803678817e-07,
+      "loss": 0.23,
+      "step": 951
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7860708313336866,
+      "learning_rate": 9.97534447637642e-07,
+      "loss": 0.1739,
+      "step": 952
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6335474927518889,
+      "learning_rate": 9.975241936815826e-07,
+      "loss": 0.3164,
+      "step": 953
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.34379862224453245,
+      "learning_rate": 9.97513918500141e-07,
+      "loss": 0.0258,
+      "step": 954
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7167515453188937,
+      "learning_rate": 9.975036220937553e-07,
+      "loss": 0.2723,
+      "step": 955
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6856200757666084,
+      "learning_rate": 9.974933044628652e-07,
+      "loss": 0.1118,
+      "step": 956
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.21543958132484572,
+      "learning_rate": 9.974829656079104e-07,
+      "loss": 0.1057,
+      "step": 957
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1921174412206408,
+      "learning_rate": 9.974726055293325e-07,
+      "loss": 0.1333,
+      "step": 958
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.33487542565410683,
+      "learning_rate": 9.97462224227573e-07,
+      "loss": 0.0938,
+      "step": 959
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6336804368411803,
+      "learning_rate": 9.97451821703075e-07,
+      "loss": 0.1854,
+      "step": 960
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5751970727441501,
+      "learning_rate": 9.974413979562823e-07,
+      "loss": 0.2631,
+      "step": 961
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6658528447001772,
+      "learning_rate": 9.974309529876396e-07,
+      "loss": 0.1177,
+      "step": 962
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7153463446728332,
+      "learning_rate": 9.974204867975926e-07,
+      "loss": 0.2134,
+      "step": 963
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.556809473274764,
+      "learning_rate": 9.974099993865877e-07,
+      "loss": 0.0977,
+      "step": 964
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4850417655109695,
+      "learning_rate": 9.97399490755072e-07,
+      "loss": 0.3,
+      "step": 965
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.393049910278882,
+      "learning_rate": 9.973889609034944e-07,
+      "loss": 0.1262,
+      "step": 966
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.21258533130898213,
+      "learning_rate": 9.973784098323038e-07,
+      "loss": 0.0703,
+      "step": 967
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7971081695534867,
+      "learning_rate": 9.973678375419504e-07,
+      "loss": 0.1148,
+      "step": 968
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6055105582764296,
+      "learning_rate": 9.97357244032885e-07,
+      "loss": 0.1464,
+      "step": 969
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.98527191471172,
+      "learning_rate": 9.973466293055602e-07,
+      "loss": 0.1548,
+      "step": 970
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8005624121428562,
+      "learning_rate": 9.97335993360428e-07,
+      "loss": 0.2071,
+      "step": 971
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7641223533278209,
+      "learning_rate": 9.973253361979427e-07,
+      "loss": 0.2064,
+      "step": 972
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5262702698149297,
+      "learning_rate": 9.973146578185588e-07,
+      "loss": 0.3942,
+      "step": 973
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44707227252640414,
+      "learning_rate": 9.973039582227319e-07,
+      "loss": 0.5105,
+      "step": 974
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.459651156249005,
+      "learning_rate": 9.972932374109182e-07,
+      "loss": 0.3136,
+      "step": 975
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8895166874354026,
+      "learning_rate": 9.972824953835756e-07,
+      "loss": 0.2322,
+      "step": 976
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.29423753592284146,
+      "learning_rate": 9.97271732141162e-07,
+      "loss": 0.1969,
+      "step": 977
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8961796479193413,
+      "learning_rate": 9.972609476841365e-07,
+      "loss": 0.0635,
+      "step": 978
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.27029945533215904,
+      "learning_rate": 9.972501420129595e-07,
+      "loss": 0.0962,
+      "step": 979
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4174636997470639,
+      "learning_rate": 9.97239315128092e-07,
+      "loss": 0.1965,
+      "step": 980
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5576671873511183,
+      "learning_rate": 9.972284670299955e-07,
+      "loss": 0.1702,
+      "step": 981
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8479045439406689,
+      "learning_rate": 9.972175977191332e-07,
+      "loss": 0.4184,
+      "step": 982
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4090059840814684,
+      "learning_rate": 9.972067071959685e-07,
+      "loss": 0.5159,
+      "step": 983
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7488220660318784,
+      "learning_rate": 9.971957954609662e-07,
+      "loss": 0.2249,
+      "step": 984
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0557110648302839,
+      "learning_rate": 9.971848625145919e-07,
+      "loss": 0.1822,
+      "step": 985
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6860523194772445,
+      "learning_rate": 9.97173908357312e-07,
+      "loss": 0.2638,
+      "step": 986
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5146231791989171,
+      "learning_rate": 9.971629329895934e-07,
+      "loss": 0.2392,
+      "step": 987
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6204843920733871,
+      "learning_rate": 9.971519364119048e-07,
+      "loss": 0.1253,
+      "step": 988
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6831185038719864,
+      "learning_rate": 9.971409186247151e-07,
+      "loss": 0.1564,
+      "step": 989
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5145580830055223,
+      "learning_rate": 9.971298796284947e-07,
+      "loss": 0.1791,
+      "step": 990
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3080111600263855,
+      "learning_rate": 9.97118819423714e-07,
+      "loss": 0.1089,
+      "step": 991
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.026427565716047,
+      "learning_rate": 9.97107738010845e-07,
+      "loss": 0.1174,
+      "step": 992
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.556784534737722,
+      "learning_rate": 9.97096635390361e-07,
+      "loss": 0.0677,
+      "step": 993
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3100620984644595,
+      "learning_rate": 9.97085511562735e-07,
+      "loss": 0.1863,
+      "step": 994
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6374941098365919,
+      "learning_rate": 9.970743665284418e-07,
+      "loss": 0.0677,
+      "step": 995
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4408399328420816,
+      "learning_rate": 9.97063200287957e-07,
+      "loss": 0.0862,
+      "step": 996
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4235597994350044,
+      "learning_rate": 9.970520128417567e-07,
+      "loss": 0.3549,
+      "step": 997
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5956360358683012,
+      "learning_rate": 9.970408041903185e-07,
+      "loss": 0.165,
+      "step": 998
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5629107483257062,
+      "learning_rate": 9.970295743341205e-07,
+      "loss": 0.1385,
+      "step": 999
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2858967054426137,
+      "learning_rate": 9.970183232736414e-07,
+      "loss": 0.0118,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5001823321468504,
+      "learning_rate": 9.970070510093616e-07,
+      "loss": 0.3272,
+      "step": 1001
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8804405518087354,
+      "learning_rate": 9.96995757541762e-07,
+      "loss": 0.3571,
+      "step": 1002
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5344573086959735,
+      "learning_rate": 9.969844428713242e-07,
+      "loss": 0.2215,
+      "step": 1003
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7997486073780382,
+      "learning_rate": 9.96973106998531e-07,
+      "loss": 0.2332,
+      "step": 1004
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6109746072637622,
+      "learning_rate": 9.96961749923866e-07,
+      "loss": 0.4813,
+      "step": 1005
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31869720912455685,
+      "learning_rate": 9.969503716478138e-07,
+      "loss": 0.2018,
+      "step": 1006
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5995328400952173,
+      "learning_rate": 9.9693897217086e-07,
+      "loss": 0.1437,
+      "step": 1007
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1796887188031795,
+      "learning_rate": 9.969275514934903e-07,
+      "loss": 0.1808,
+      "step": 1008
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5658398713474428,
+      "learning_rate": 9.969161096161924e-07,
+      "loss": 0.1765,
+      "step": 1009
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4490361150588233,
+      "learning_rate": 9.969046465394544e-07,
+      "loss": 0.1756,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5629561790593844,
+      "learning_rate": 9.968931622637651e-07,
+      "loss": 0.0333,
+      "step": 1011
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5323933653130326,
+      "learning_rate": 9.968816567896148e-07,
+      "loss": 0.2166,
+      "step": 1012
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3418637697195922,
+      "learning_rate": 9.96870130117494e-07,
+      "loss": 0.2775,
+      "step": 1013
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5571011756013858,
+      "learning_rate": 9.968585822478948e-07,
+      "loss": 0.3092,
+      "step": 1014
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.470738521801038,
+      "learning_rate": 9.968470131813096e-07,
+      "loss": 0.21,
+      "step": 1015
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2805505635889866,
+      "learning_rate": 9.968354229182319e-07,
+      "loss": 0.0548,
+      "step": 1016
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6034060234356918,
+      "learning_rate": 9.968238114591565e-07,
+      "loss": 0.3312,
+      "step": 1017
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5386064080951632,
+      "learning_rate": 9.968121788045784e-07,
+      "loss": 0.2165,
+      "step": 1018
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31641615699874504,
+      "learning_rate": 9.968005249549942e-07,
+      "loss": 0.0959,
+      "step": 1019
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3656653627529067,
+      "learning_rate": 9.967888499109007e-07,
+      "loss": 0.2005,
+      "step": 1020
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.34502538080044065,
+      "learning_rate": 9.967771536727963e-07,
+      "loss": 0.2918,
+      "step": 1021
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6263498899361163,
+      "learning_rate": 9.967654362411798e-07,
+      "loss": 0.3505,
+      "step": 1022
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.22916347440126475,
+      "learning_rate": 9.967536976165515e-07,
+      "loss": 0.0495,
+      "step": 1023
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.36512885889328833,
+      "learning_rate": 9.967419377994116e-07,
+      "loss": 0.2453,
+      "step": 1024
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6159879197644346,
+      "learning_rate": 9.967301567902619e-07,
+      "loss": 0.0777,
+      "step": 1025
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5848397426298377,
+      "learning_rate": 9.967183545896055e-07,
+      "loss": 0.1947,
+      "step": 1026
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3063476765681408,
+      "learning_rate": 9.967065311979452e-07,
+      "loss": 0.0198,
+      "step": 1027
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.12462664133673665,
+      "learning_rate": 9.96694686615786e-07,
+      "loss": 0.0142,
+      "step": 1028
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.33720844756144086,
+      "learning_rate": 9.96682820843633e-07,
+      "loss": 0.1659,
+      "step": 1029
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5796779985194487,
+      "learning_rate": 9.966709338819924e-07,
+      "loss": 0.1904,
+      "step": 1030
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4410799052200912,
+      "learning_rate": 9.966590257313713e-07,
+      "loss": 0.2489,
+      "step": 1031
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5506144502649098,
+      "learning_rate": 9.966470963922778e-07,
+      "loss": 0.1857,
+      "step": 1032
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8245813418322644,
+      "learning_rate": 9.966351458652207e-07,
+      "loss": 0.0588,
+      "step": 1033
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7573390896270649,
+      "learning_rate": 9.9662317415071e-07,
+      "loss": 0.2642,
+      "step": 1034
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6741845741174953,
+      "learning_rate": 9.96611181249256e-07,
+      "loss": 0.1297,
+      "step": 1035
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5176367178481315,
+      "learning_rate": 9.965991671613712e-07,
+      "loss": 0.1497,
+      "step": 1036
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.24016814478297469,
+      "learning_rate": 9.965871318875674e-07,
+      "loss": 0.1805,
+      "step": 1037
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5975343726265553,
+      "learning_rate": 9.965750754283581e-07,
+      "loss": 0.3547,
+      "step": 1038
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.48165162801011546,
+      "learning_rate": 9.965629977842583e-07,
+      "loss": 0.0807,
+      "step": 1039
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3968531251659212,
+      "learning_rate": 9.965508989557825e-07,
+      "loss": 0.3349,
+      "step": 1040
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.27580191225200595,
+      "learning_rate": 9.965387789434473e-07,
+      "loss": 0.011,
+      "step": 1041
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44955503042974304,
+      "learning_rate": 9.965266377477694e-07,
+      "loss": 0.2546,
+      "step": 1042
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3241293984826068,
+      "learning_rate": 9.965144753692672e-07,
+      "loss": 0.0935,
+      "step": 1043
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3984129298375122,
+      "learning_rate": 9.965022918084591e-07,
+      "loss": 0.0321,
+      "step": 1044
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.22176833947800315,
+      "learning_rate": 9.964900870658653e-07,
+      "loss": 0.1971,
+      "step": 1045
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4374868508187839,
+      "learning_rate": 9.964778611420063e-07,
+      "loss": 0.1572,
+      "step": 1046
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6606499413383291,
+      "learning_rate": 9.964656140374038e-07,
+      "loss": 0.1915,
+      "step": 1047
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6208050665040733,
+      "learning_rate": 9.9645334575258e-07,
+      "loss": 0.0846,
+      "step": 1048
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3445588204772651,
+      "learning_rate": 9.964410562880587e-07,
+      "loss": 0.0497,
+      "step": 1049
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4460992279579737,
+      "learning_rate": 9.964287456443639e-07,
+      "loss": 0.3039,
+      "step": 1050
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3220445790766819,
+      "learning_rate": 9.964164138220207e-07,
+      "loss": 0.0138,
+      "step": 1051
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6758744160821144,
+      "learning_rate": 9.964040608215557e-07,
+      "loss": 0.2955,
+      "step": 1052
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3339475440618883,
+      "learning_rate": 9.963916866434952e-07,
+      "loss": 0.0671,
+      "step": 1053
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1073303272937873,
+      "learning_rate": 9.963792912883676e-07,
+      "loss": 0.2129,
+      "step": 1054
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6059768593260848,
+      "learning_rate": 9.963668747567018e-07,
+      "loss": 0.2493,
+      "step": 1055
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4349080240496784,
+      "learning_rate": 9.963544370490268e-07,
+      "loss": 0.408,
+      "step": 1056
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6005730259160773,
+      "learning_rate": 9.963419781658742e-07,
+      "loss": 0.095,
+      "step": 1057
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6750176494329789,
+      "learning_rate": 9.963294981077747e-07,
+      "loss": 0.3842,
+      "step": 1058
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3515007208986595,
+      "learning_rate": 9.963169968752613e-07,
+      "loss": 0.128,
+      "step": 1059
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8845299459356517,
+      "learning_rate": 9.96304474468867e-07,
+      "loss": 0.3846,
+      "step": 1060
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7393735816661642,
+      "learning_rate": 9.962919308891263e-07,
+      "loss": 0.3981,
+      "step": 1061
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44839149358820335,
+      "learning_rate": 9.96279366136574e-07,
+      "loss": 0.2503,
+      "step": 1062
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5404046180966962,
+      "learning_rate": 9.962667802117463e-07,
+      "loss": 0.1419,
+      "step": 1063
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.16238469118518242,
+      "learning_rate": 9.962541731151802e-07,
+      "loss": 0.0093,
+      "step": 1064
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5738195741439209,
+      "learning_rate": 9.962415448474133e-07,
+      "loss": 0.2628,
+      "step": 1065
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.25126688476863107,
+      "learning_rate": 9.962288954089845e-07,
+      "loss": 0.0424,
+      "step": 1066
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.35598606064613925,
+      "learning_rate": 9.962162248004337e-07,
+      "loss": 0.1254,
+      "step": 1067
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.46534900163923154,
+      "learning_rate": 9.96203533022301e-07,
+      "loss": 0.0417,
+      "step": 1068
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.469743894149503,
+      "learning_rate": 9.961908200751283e-07,
+      "loss": 0.1853,
+      "step": 1069
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5671191935457041,
+      "learning_rate": 9.961780859594578e-07,
+      "loss": 0.1609,
+      "step": 1070
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5871653801245706,
+      "learning_rate": 9.961653306758325e-07,
+      "loss": 0.2998,
+      "step": 1071
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5748645345770612,
+      "learning_rate": 9.961525542247968e-07,
+      "loss": 0.1314,
+      "step": 1072
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8097576392881596,
+      "learning_rate": 9.961397566068958e-07,
+      "loss": 0.2395,
+      "step": 1073
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3934771452321142,
+      "learning_rate": 9.961269378226755e-07,
+      "loss": 0.0575,
+      "step": 1074
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7713689273114455,
+      "learning_rate": 9.961140978726827e-07,
+      "loss": 0.3685,
+      "step": 1075
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6497760757747596,
+      "learning_rate": 9.96101236757465e-07,
+      "loss": 0.2248,
+      "step": 1076
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.141576142325507,
+      "learning_rate": 9.960883544775713e-07,
+      "loss": 0.1828,
+      "step": 1077
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.348158927625783,
+      "learning_rate": 9.960754510335513e-07,
+      "loss": 0.2588,
+      "step": 1078
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7260341348660807,
+      "learning_rate": 9.960625264259552e-07,
+      "loss": 0.3234,
+      "step": 1079
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.23468077749983832,
+      "learning_rate": 9.960495806553345e-07,
+      "loss": 0.1675,
+      "step": 1080
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.140786165743733,
+      "learning_rate": 9.960366137222413e-07,
+      "loss": 0.0529,
+      "step": 1081
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5568296291928382,
+      "learning_rate": 9.960236256272293e-07,
+      "loss": 0.1255,
+      "step": 1082
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5604993219180562,
+      "learning_rate": 9.960106163708522e-07,
+      "loss": 0.3344,
+      "step": 1083
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4688437202999613,
+      "learning_rate": 9.959975859536652e-07,
+      "loss": 0.0343,
+      "step": 1084
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42635694146014674,
+      "learning_rate": 9.95984534376224e-07,
+      "loss": 0.1534,
+      "step": 1085
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.17090463730875813,
+      "learning_rate": 9.959714616390855e-07,
+      "loss": 0.0088,
+      "step": 1086
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5101292094283796,
+      "learning_rate": 9.959583677428074e-07,
+      "loss": 0.2843,
+      "step": 1087
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6069779617803412,
+      "learning_rate": 9.959452526879484e-07,
+      "loss": 0.1746,
+      "step": 1088
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1833347923100763,
+      "learning_rate": 9.959321164750678e-07,
+      "loss": 0.2469,
+      "step": 1089
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5081377136538748,
+      "learning_rate": 9.959189591047264e-07,
+      "loss": 0.3232,
+      "step": 1090
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4076384920062124,
+      "learning_rate": 9.959057805774851e-07,
+      "loss": 0.2807,
+      "step": 1091
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3374042679952573,
+      "learning_rate": 9.958925808939063e-07,
+      "loss": 0.1914,
+      "step": 1092
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6500977046862573,
+      "learning_rate": 9.958793600545531e-07,
+      "loss": 0.4011,
+      "step": 1093
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.17172487725261068,
+      "learning_rate": 9.958661180599898e-07,
+      "loss": 0.0724,
+      "step": 1094
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5200887718325509,
+      "learning_rate": 9.95852854910781e-07,
+      "loss": 0.109,
+      "step": 1095
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0243269067900986,
+      "learning_rate": 9.958395706074925e-07,
+      "loss": 0.3678,
+      "step": 1096
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.21424244960150624,
+      "learning_rate": 9.958262651506913e-07,
+      "loss": 0.0498,
+      "step": 1097
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43342568758247785,
+      "learning_rate": 9.958129385409447e-07,
+      "loss": 0.2587,
+      "step": 1098
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44996930849030936,
+      "learning_rate": 9.957995907788217e-07,
+      "loss": 0.0959,
+      "step": 1099
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3958412973964965,
+      "learning_rate": 9.95786221864891e-07,
+      "loss": 0.2296,
+      "step": 1100
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6945523383410414,
+      "learning_rate": 9.95772831799724e-07,
+      "loss": 0.2633,
+      "step": 1101
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5773111265925281,
+      "learning_rate": 9.95759420583891e-07,
+      "loss": 0.0299,
+      "step": 1102
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.49091192136696843,
+      "learning_rate": 9.957459882179647e-07,
+      "loss": 0.2164,
+      "step": 1103
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6774248050977482,
+      "learning_rate": 9.957325347025178e-07,
+      "loss": 0.3325,
+      "step": 1104
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5824953378957358,
+      "learning_rate": 9.957190600381245e-07,
+      "loss": 0.2733,
+      "step": 1105
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.565278403467531,
+      "learning_rate": 9.957055642253596e-07,
+      "loss": 0.376,
+      "step": 1106
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.27982614461170324,
+      "learning_rate": 9.95692047264799e-07,
+      "loss": 0.0116,
+      "step": 1107
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6146706873778525,
+      "learning_rate": 9.956785091570189e-07,
+      "loss": 0.2508,
+      "step": 1108
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6467482493773037,
+      "learning_rate": 9.956649499025973e-07,
+      "loss": 0.2231,
+      "step": 1109
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41524716018566904,
+      "learning_rate": 9.956513695021124e-07,
+      "loss": 0.3621,
+      "step": 1110
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.39463424844312994,
+      "learning_rate": 9.956377679561439e-07,
+      "loss": 0.3311,
+      "step": 1111
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.20254282800084308,
+      "learning_rate": 9.956241452652717e-07,
+      "loss": 0.1893,
+      "step": 1112
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6996924986838188,
+      "learning_rate": 9.95610501430077e-07,
+      "loss": 0.2643,
+      "step": 1113
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.45664575882710157,
+      "learning_rate": 9.955968364511425e-07,
+      "loss": 0.2615,
+      "step": 1114
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4008275526455585,
+      "learning_rate": 9.955831503290502e-07,
+      "loss": 0.2214,
+      "step": 1115
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6087192129796571,
+      "learning_rate": 9.955694430643847e-07,
+      "loss": 0.2212,
+      "step": 1116
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.30902609825608435,
+      "learning_rate": 9.955557146577305e-07,
+      "loss": 0.2269,
+      "step": 1117
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6729231869763611,
+      "learning_rate": 9.955419651096733e-07,
+      "loss": 0.155,
+      "step": 1118
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6423072684982292,
+      "learning_rate": 9.955281944207996e-07,
+      "loss": 0.2317,
+      "step": 1119
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3614719727445751,
+      "learning_rate": 9.955144025916972e-07,
+      "loss": 0.0284,
+      "step": 1120
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.31847134873224486,
+      "learning_rate": 9.955005896229541e-07,
+      "loss": 0.059,
+      "step": 1121
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6640613090116112,
+      "learning_rate": 9.954867555151599e-07,
+      "loss": 0.1176,
+      "step": 1122
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2870218793153667,
+      "learning_rate": 9.954729002689046e-07,
+      "loss": 0.1116,
+      "step": 1123
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4299444902143463,
+      "learning_rate": 9.954590238847792e-07,
+      "loss": 0.177,
+      "step": 1124
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5042856873929278,
+      "learning_rate": 9.954451263633761e-07,
+      "loss": 0.2856,
+      "step": 1125
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5624574567029191,
+      "learning_rate": 9.95431207705288e-07,
+      "loss": 0.2922,
+      "step": 1126
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7033341472119826,
+      "learning_rate": 9.954172679111083e-07,
+      "loss": 0.19,
+      "step": 1127
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5166063605840293,
+      "learning_rate": 9.954033069814323e-07,
+      "loss": 0.1983,
+      "step": 1128
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3111036588703034,
+      "learning_rate": 9.953893249168552e-07,
+      "loss": 0.2572,
+      "step": 1129
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5850201954455693,
+      "learning_rate": 9.953753217179737e-07,
+      "loss": 0.3981,
+      "step": 1130
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.31148056278206443,
+      "learning_rate": 9.95361297385385e-07,
+      "loss": 0.1387,
+      "step": 1131
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5011527650666843,
+      "learning_rate": 9.953472519196876e-07,
+      "loss": 0.1981,
+      "step": 1132
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2952964681892801,
+      "learning_rate": 9.95333185321481e-07,
+      "loss": 0.0314,
+      "step": 1133
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2969760901552938,
+      "learning_rate": 9.953190975913645e-07,
+      "loss": 0.0142,
+      "step": 1134
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.303357352799507,
+      "learning_rate": 9.953049887299399e-07,
+      "loss": 0.0174,
+      "step": 1135
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5223668998963273,
+      "learning_rate": 9.952908587378088e-07,
+      "loss": 0.0619,
+      "step": 1136
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4011876020727392,
+      "learning_rate": 9.95276707615574e-07,
+      "loss": 0.1629,
+      "step": 1137
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6661033093782415,
+      "learning_rate": 9.95262535363839e-07,
+      "loss": 0.3111,
+      "step": 1138
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.20087502918394676,
+      "learning_rate": 9.952483419832087e-07,
+      "loss": 0.0984,
+      "step": 1139
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5523294888749696,
+      "learning_rate": 9.952341274742885e-07,
+      "loss": 0.3158,
+      "step": 1140
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4577651432739096,
+      "learning_rate": 9.952198918376852e-07,
+      "loss": 0.233,
+      "step": 1141
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6495563574623383,
+      "learning_rate": 9.952056350740055e-07,
+      "loss": 0.231,
+      "step": 1142
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1953624200184034,
+      "learning_rate": 9.95191357183858e-07,
+      "loss": 0.1671,
+      "step": 1143
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.656737350758839,
+      "learning_rate": 9.951770581678517e-07,
+      "loss": 0.2997,
+      "step": 1144
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4068216497777626,
+      "learning_rate": 9.951627380265966e-07,
+      "loss": 0.4044,
+      "step": 1145
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7505436657249192,
+      "learning_rate": 9.95148396760704e-07,
+      "loss": 0.1125,
+      "step": 1146
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.649769295135062,
+      "learning_rate": 9.95134034370785e-07,
+      "loss": 0.0573,
+      "step": 1147
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3886561767416595,
+      "learning_rate": 9.95119650857453e-07,
+      "loss": 0.0267,
+      "step": 1148
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3897086233521129,
+      "learning_rate": 9.951052462213214e-07,
+      "loss": 0.133,
+      "step": 1149
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.29633369766252005,
+      "learning_rate": 9.950908204630047e-07,
+      "loss": 0.1935,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4656970089994235,
+      "learning_rate": 9.950763735831182e-07,
+      "loss": 0.3169,
+      "step": 1151
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3022354535153058,
+      "learning_rate": 9.950619055822786e-07,
+      "loss": 0.0651,
+      "step": 1152
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7774838289878337,
+      "learning_rate": 9.950474164611028e-07,
+      "loss": 0.1482,
+      "step": 1153
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5852672827206604,
+      "learning_rate": 9.95032906220209e-07,
+      "loss": 0.1979,
+      "step": 1154
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4731884099599208,
+      "learning_rate": 9.950183748602163e-07,
+      "loss": 0.2213,
+      "step": 1155
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6184415335068383,
+      "learning_rate": 9.950038223817447e-07,
+      "loss": 0.073,
+      "step": 1156
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.47959352443439546,
+      "learning_rate": 9.949892487854149e-07,
+      "loss": 0.252,
+      "step": 1157
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7619017568194408,
+      "learning_rate": 9.949746540718487e-07,
+      "loss": 0.2122,
+      "step": 1158
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4304366317713998,
+      "learning_rate": 9.949600382416685e-07,
+      "loss": 0.0904,
+      "step": 1159
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6021043497073343,
+      "learning_rate": 9.949454012954985e-07,
+      "loss": 0.1185,
+      "step": 1160
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6665744147667394,
+      "learning_rate": 9.949307432339624e-07,
+      "loss": 0.1825,
+      "step": 1161
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.47650417959695823,
+      "learning_rate": 9.94916064057686e-07,
+      "loss": 0.3691,
+      "step": 1162
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.15356554982909978,
+      "learning_rate": 9.949013637672953e-07,
+      "loss": 0.1171,
+      "step": 1163
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5157615623872764,
+      "learning_rate": 9.948866423634176e-07,
+      "loss": 0.1202,
+      "step": 1164
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4315926796169466,
+      "learning_rate": 9.94871899846681e-07,
+      "loss": 0.152,
+      "step": 1165
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.40499839612104527,
+      "learning_rate": 9.94857136217714e-07,
+      "loss": 0.3212,
+      "step": 1166
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2255758734858464,
+      "learning_rate": 9.948423514771472e-07,
+      "loss": 0.1532,
+      "step": 1167
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6682922622170951,
+      "learning_rate": 9.948275456256108e-07,
+      "loss": 0.2367,
+      "step": 1168
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7060494908892788,
+      "learning_rate": 9.948127186637365e-07,
+      "loss": 0.0218,
+      "step": 1169
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3161321311597232,
+      "learning_rate": 9.94797870592157e-07,
+      "loss": 0.2219,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3369016471937861,
+      "learning_rate": 9.947830014115056e-07,
+      "loss": 0.2064,
+      "step": 1171
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1991222681659008,
+      "learning_rate": 9.947681111224166e-07,
+      "loss": 0.4302,
+      "step": 1172
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4410691788944219,
+      "learning_rate": 9.947531997255256e-07,
+      "loss": 0.2865,
+      "step": 1173
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3624040415600144,
+      "learning_rate": 9.947382672214684e-07,
+      "loss": 0.2802,
+      "step": 1174
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5548278217941871,
+      "learning_rate": 9.947233136108822e-07,
+      "loss": 0.1076,
+      "step": 1175
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5653568715153751,
+      "learning_rate": 9.94708338894405e-07,
+      "loss": 0.3068,
+      "step": 1176
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.18076817660824668,
+      "learning_rate": 9.946933430726753e-07,
+      "loss": 0.0119,
+      "step": 1177
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4426239335022008,
+      "learning_rate": 9.946783261463333e-07,
+      "loss": 0.1841,
+      "step": 1178
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.582521438640547,
+      "learning_rate": 9.946632881160196e-07,
+      "loss": 0.2756,
+      "step": 1179
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7782034172939633,
+      "learning_rate": 9.946482289823755e-07,
+      "loss": 0.1649,
+      "step": 1180
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3751614071969384,
+      "learning_rate": 9.946331487460435e-07,
+      "loss": 0.048,
+      "step": 1181
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6767317790794916,
+      "learning_rate": 9.946180474076673e-07,
+      "loss": 0.1142,
+      "step": 1182
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.40650271265656646,
+      "learning_rate": 9.946029249678906e-07,
+      "loss": 0.0159,
+      "step": 1183
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4988330139152289,
+      "learning_rate": 9.94587781427359e-07,
+      "loss": 0.1236,
+      "step": 1184
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.741740164161934,
+      "learning_rate": 9.945726167867184e-07,
+      "loss": 0.3357,
+      "step": 1185
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6184543088109902,
+      "learning_rate": 9.945574310466159e-07,
+      "loss": 0.2253,
+      "step": 1186
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5727246845437999,
+      "learning_rate": 9.945422242076989e-07,
+      "loss": 0.3845,
+      "step": 1187
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5432437340533075,
+      "learning_rate": 9.945269962706167e-07,
+      "loss": 0.3971,
+      "step": 1188
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.29682050896542284,
+      "learning_rate": 9.945117472360184e-07,
+      "loss": 0.1106,
+      "step": 1189
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8166999080191234,
+      "learning_rate": 9.944964771045552e-07,
+      "loss": 0.3071,
+      "step": 1190
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7471517122493586,
+      "learning_rate": 9.944811858768782e-07,
+      "loss": 0.1709,
+      "step": 1191
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2994779575470051,
+      "learning_rate": 9.944658735536395e-07,
+      "loss": 0.2743,
+      "step": 1192
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.23381810826721885,
+      "learning_rate": 9.94450540135493e-07,
+      "loss": 0.0987,
+      "step": 1193
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.39938123805950754,
+      "learning_rate": 9.94435185623092e-07,
+      "loss": 0.2987,
+      "step": 1194
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9688135724386572,
+      "learning_rate": 9.944198100170927e-07,
+      "loss": 0.152,
+      "step": 1195
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6587060853941012,
+      "learning_rate": 9.9440441331815e-07,
+      "loss": 0.1525,
+      "step": 1196
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.16281682683361928,
+      "learning_rate": 9.943889955269212e-07,
+      "loss": 0.0904,
+      "step": 1197
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.33697811875794936,
+      "learning_rate": 9.94373556644064e-07,
+      "loss": 0.059,
+      "step": 1198
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3749015720899504,
+      "learning_rate": 9.94358096670237e-07,
+      "loss": 0.2188,
+      "step": 1199
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.010358437978261,
+      "learning_rate": 9.943426156061e-07,
+      "loss": 0.2657,
+      "step": 1200
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5161456376184853,
+      "learning_rate": 9.94327113452313e-07,
+      "loss": 0.0139,
+      "step": 1201
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2871394945981592,
+      "learning_rate": 9.943115902095378e-07,
+      "loss": 0.008,
+      "step": 1202
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6177966823720042,
+      "learning_rate": 9.942960458784364e-07,
+      "loss": 0.3859,
+      "step": 1203
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3777364649984627,
+      "learning_rate": 9.942804804596722e-07,
+      "loss": 0.1185,
+      "step": 1204
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.969121238221322,
+      "learning_rate": 9.942648939539086e-07,
+      "loss": 0.306,
+      "step": 1205
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.146732018703517,
+      "learning_rate": 9.942492863618114e-07,
+      "loss": 0.156,
+      "step": 1206
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.41643345430756434,
+      "learning_rate": 9.942336576840462e-07,
+      "loss": 0.1954,
+      "step": 1207
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.19056838451877037,
+      "learning_rate": 9.942180079212793e-07,
+      "loss": 0.093,
+      "step": 1208
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.39719466309769197,
+      "learning_rate": 9.94202337074179e-07,
+      "loss": 0.1966,
+      "step": 1209
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7120739258858446,
+      "learning_rate": 9.941866451434131e-07,
+      "loss": 0.2299,
+      "step": 1210
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.394798508834025,
+      "learning_rate": 9.94170932129652e-07,
+      "loss": 0.2869,
+      "step": 1211
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5192994332437146,
+      "learning_rate": 9.941551980335652e-07,
+      "loss": 0.2403,
+      "step": 1212
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4367852588222532,
+      "learning_rate": 9.941394428558244e-07,
+      "loss": 0.1468,
+      "step": 1213
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.74335549781076,
+      "learning_rate": 9.941236665971015e-07,
+      "loss": 0.2553,
+      "step": 1214
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3262131805085108,
+      "learning_rate": 9.941078692580698e-07,
+      "loss": 0.1993,
+      "step": 1215
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3145273998877605,
+      "learning_rate": 9.94092050839403e-07,
+      "loss": 0.1359,
+      "step": 1216
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43561714541066404,
+      "learning_rate": 9.94076211341776e-07,
+      "loss": 0.0249,
+      "step": 1217
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2378825638605044,
+      "learning_rate": 9.940603507658648e-07,
+      "loss": 0.0156,
+      "step": 1218
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6998547645217111,
+      "learning_rate": 9.940444691123458e-07,
+      "loss": 0.3879,
+      "step": 1219
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9523241648092471,
+      "learning_rate": 9.940285663818967e-07,
+      "loss": 0.3711,
+      "step": 1220
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8063499293177215,
+      "learning_rate": 9.940126425751956e-07,
+      "loss": 0.0743,
+      "step": 1221
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6669138093510243,
+      "learning_rate": 9.939966976929222e-07,
+      "loss": 0.4219,
+      "step": 1222
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5257776310542316,
+      "learning_rate": 9.939807317357566e-07,
+      "loss": 0.3033,
+      "step": 1223
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.23137183935848096,
+      "learning_rate": 9.939647447043798e-07,
+      "loss": 0.0905,
+      "step": 1224
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1987866191319867,
+      "learning_rate": 9.939487365994741e-07,
+      "loss": 0.0111,
+      "step": 1225
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4266213404580218,
+      "learning_rate": 9.939327074217225e-07,
+      "loss": 0.1496,
+      "step": 1226
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8760138859606063,
+      "learning_rate": 9.939166571718084e-07,
+      "loss": 0.5151,
+      "step": 1227
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46525354145023406,
+      "learning_rate": 9.93900585850417e-07,
+      "loss": 0.15,
+      "step": 1228
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.48283921693480103,
+      "learning_rate": 9.938844934582337e-07,
+      "loss": 0.2343,
+      "step": 1229
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6795500594554379,
+      "learning_rate": 9.938683799959452e-07,
+      "loss": 0.2176,
+      "step": 1230
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4507303170281007,
+      "learning_rate": 9.938522454642387e-07,
+      "loss": 0.1081,
+      "step": 1231
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4182381054561944,
+      "learning_rate": 9.938360898638026e-07,
+      "loss": 0.1998,
+      "step": 1232
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1807866975834911,
+      "learning_rate": 9.938199131953263e-07,
+      "loss": 0.0178,
+      "step": 1233
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6328066547132252,
+      "learning_rate": 9.938037154594996e-07,
+      "loss": 0.1559,
+      "step": 1234
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4285963389738819,
+      "learning_rate": 9.937874966570139e-07,
+      "loss": 0.1104,
+      "step": 1235
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.689788062574801,
+      "learning_rate": 9.937712567885608e-07,
+      "loss": 0.1247,
+      "step": 1236
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4376887501103226,
+      "learning_rate": 9.937549958548335e-07,
+      "loss": 0.3802,
+      "step": 1237
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42673112844316574,
+      "learning_rate": 9.937387138565255e-07,
+      "loss": 0.2974,
+      "step": 1238
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.48626121488061697,
+      "learning_rate": 9.93722410794331e-07,
+      "loss": 0.3471,
+      "step": 1239
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6003795063350325,
+      "learning_rate": 9.937060866689463e-07,
+      "loss": 0.2365,
+      "step": 1240
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5696774630629393,
+      "learning_rate": 9.936897414810676e-07,
+      "loss": 0.2721,
+      "step": 1241
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.20690602525821974,
+      "learning_rate": 9.936733752313918e-07,
+      "loss": 0.1156,
+      "step": 1242
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7178905498679734,
+      "learning_rate": 9.936569879206175e-07,
+      "loss": 0.2117,
+      "step": 1243
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5812984740921069,
+      "learning_rate": 9.936405795494438e-07,
+      "loss": 0.0232,
+      "step": 1244
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.12484029407011618,
+      "learning_rate": 9.936241501185705e-07,
+      "loss": 0.0247,
+      "step": 1245
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5954917585734902,
+      "learning_rate": 9.936076996286987e-07,
+      "loss": 0.0639,
+      "step": 1246
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4379080912973191,
+      "learning_rate": 9.935912280805302e-07,
+      "loss": 0.1497,
+      "step": 1247
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8882229683532059,
+      "learning_rate": 9.935747354747677e-07,
+      "loss": 0.0951,
+      "step": 1248
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.29518732119725033,
+      "learning_rate": 9.935582218121147e-07,
+      "loss": 0.0143,
+      "step": 1249
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4637263032560623,
+      "learning_rate": 9.935416870932757e-07,
+      "loss": 0.1727,
+      "step": 1250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.28543800842752126,
+      "learning_rate": 9.935251313189563e-07,
+      "loss": 0.0648,
+      "step": 1251
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3134866725587153,
+      "learning_rate": 9.935085544898627e-07,
+      "loss": 0.0368,
+      "step": 1252
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0373680656726496,
+      "learning_rate": 9.93491956606702e-07,
+      "loss": 0.1571,
+      "step": 1253
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.484000544280834,
+      "learning_rate": 9.934753376701825e-07,
+      "loss": 0.2467,
+      "step": 1254
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5171668737774494,
+      "learning_rate": 9.93458697681013e-07,
+      "loss": 0.2099,
+      "step": 1255
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.02310506054813,
+      "learning_rate": 9.934420366399036e-07,
+      "loss": 0.2822,
+      "step": 1256
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42189408668989653,
+      "learning_rate": 9.934253545475648e-07,
+      "loss": 0.1756,
+      "step": 1257
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15533316374032066,
+      "learning_rate": 9.934086514047086e-07,
+      "loss": 0.0806,
+      "step": 1258
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6559099183399568,
+      "learning_rate": 9.933919272120474e-07,
+      "loss": 0.2921,
+      "step": 1259
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.17634704006158938,
+      "learning_rate": 9.933751819702948e-07,
+      "loss": 0.0823,
+      "step": 1260
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5371146377850167,
+      "learning_rate": 9.93358415680165e-07,
+      "loss": 0.194,
+      "step": 1261
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6832922633310711,
+      "learning_rate": 9.933416283423736e-07,
+      "loss": 0.154,
+      "step": 1262
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5445353980456783,
+      "learning_rate": 9.933248199576364e-07,
+      "loss": 0.1312,
+      "step": 1263
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7814843978025754,
+      "learning_rate": 9.933079905266707e-07,
+      "loss": 0.1132,
+      "step": 1264
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.952130812885761,
+      "learning_rate": 9.932911400501947e-07,
+      "loss": 0.1473,
+      "step": 1265
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.56061281154319,
+      "learning_rate": 9.932742685289269e-07,
+      "loss": 0.2857,
+      "step": 1266
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8866076182489778,
+      "learning_rate": 9.93257375963587e-07,
+      "loss": 0.1433,
+      "step": 1267
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.21904820241433778,
+      "learning_rate": 9.93240462354896e-07,
+      "loss": 0.0023,
+      "step": 1268
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7858922196798822,
+      "learning_rate": 9.932235277035754e-07,
+      "loss": 0.291,
+      "step": 1269
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7868040611739603,
+      "learning_rate": 9.932065720103476e-07,
+      "loss": 0.4161,
+      "step": 1270
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.657228617660802,
+      "learning_rate": 9.931895952759357e-07,
+      "loss": 0.1842,
+      "step": 1271
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5977616164189764,
+      "learning_rate": 9.931725975010646e-07,
+      "loss": 0.1011,
+      "step": 1272
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6084631591727283,
+      "learning_rate": 9.931555786864589e-07,
+      "loss": 0.2752,
+      "step": 1273
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5876481628805643,
+      "learning_rate": 9.931385388328448e-07,
+      "loss": 0.5346,
+      "step": 1274
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.35573169640134594,
+      "learning_rate": 9.931214779409494e-07,
+      "loss": 0.2471,
+      "step": 1275
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5684956930967539,
+      "learning_rate": 9.931043960115005e-07,
+      "loss": 0.2068,
+      "step": 1276
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5446491705101372,
+      "learning_rate": 9.930872930452267e-07,
+      "loss": 0.0741,
+      "step": 1277
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4690054081002771,
+      "learning_rate": 9.930701690428577e-07,
+      "loss": 0.2964,
+      "step": 1278
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.21813429221611408,
+      "learning_rate": 9.930530240051243e-07,
+      "loss": 0.0949,
+      "step": 1279
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8295842997198223,
+      "learning_rate": 9.930358579327576e-07,
+      "loss": 0.4008,
+      "step": 1280
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.498212880447362,
+      "learning_rate": 9.9301867082649e-07,
+      "loss": 0.2155,
+      "step": 1281
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9418525856185737,
+      "learning_rate": 9.93001462687055e-07,
+      "loss": 0.1317,
+      "step": 1282
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.16956843695330423,
+      "learning_rate": 9.929842335151863e-07,
+      "loss": 0.0192,
+      "step": 1283
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.45637433018782536,
+      "learning_rate": 9.929669833116194e-07,
+      "loss": 0.2064,
+      "step": 1284
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.25788325094536546,
+      "learning_rate": 9.9294971207709e-07,
+      "loss": 0.0865,
+      "step": 1285
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8106476754548811,
+      "learning_rate": 9.929324198123347e-07,
+      "loss": 0.1804,
+      "step": 1286
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.48021474596666486,
+      "learning_rate": 9.929151065180915e-07,
+      "loss": 0.3005,
+      "step": 1287
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.29667317885857586,
+      "learning_rate": 9.928977721950992e-07,
+      "loss": 0.2073,
+      "step": 1288
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43856265730787586,
+      "learning_rate": 9.928804168440969e-07,
+      "loss": 0.1377,
+      "step": 1289
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0318322581898132,
+      "learning_rate": 9.928630404658254e-07,
+      "loss": 0.1567,
+      "step": 1290
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.49896780895805287,
+      "learning_rate": 9.928456430610257e-07,
+      "loss": 0.1008,
+      "step": 1291
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4054032864642662,
+      "learning_rate": 9.9282822463044e-07,
+      "loss": 0.2169,
+      "step": 1292
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5500438098269558,
+      "learning_rate": 9.928107851748118e-07,
+      "loss": 0.2038,
+      "step": 1293
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5264338135127226,
+      "learning_rate": 9.927933246948846e-07,
+      "loss": 0.1568,
+      "step": 1294
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5709867696264319,
+      "learning_rate": 9.927758431914036e-07,
+      "loss": 0.2531,
+      "step": 1295
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3732442385885163,
+      "learning_rate": 9.927583406651145e-07,
+      "loss": 0.1685,
+      "step": 1296
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.13061876615576834,
+      "learning_rate": 9.927408171167641e-07,
+      "loss": 0.0074,
+      "step": 1297
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5839412601840739,
+      "learning_rate": 9.927232725470998e-07,
+      "loss": 0.4051,
+      "step": 1298
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.41830401854019367,
+      "learning_rate": 9.927057069568702e-07,
+      "loss": 0.134,
+      "step": 1299
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9675478024352526,
+      "learning_rate": 9.926881203468248e-07,
+      "loss": 0.1277,
+      "step": 1300
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6413055047498372,
+      "learning_rate": 9.926705127177137e-07,
+      "loss": 0.1473,
+      "step": 1301
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5766913693417655,
+      "learning_rate": 9.92652884070288e-07,
+      "loss": 0.0879,
+      "step": 1302
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6926279196508893,
+      "learning_rate": 9.926352344053001e-07,
+      "loss": 0.0861,
+      "step": 1303
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8660050060857681,
+      "learning_rate": 9.926175637235026e-07,
+      "loss": 0.1214,
+      "step": 1304
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6410176098271414,
+      "learning_rate": 9.925998720256496e-07,
+      "loss": 0.0194,
+      "step": 1305
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.36678590115652,
+      "learning_rate": 9.925821593124959e-07,
+      "loss": 0.0723,
+      "step": 1306
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.64978428647516,
+      "learning_rate": 9.92564425584797e-07,
+      "loss": 0.2001,
+      "step": 1307
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5857836993663871,
+      "learning_rate": 9.925466708433097e-07,
+      "loss": 0.1756,
+      "step": 1308
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6154155311141734,
+      "learning_rate": 9.925288950887912e-07,
+      "loss": 0.147,
+      "step": 1309
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2220469057151436,
+      "learning_rate": 9.925110983219998e-07,
+      "loss": 0.0138,
+      "step": 1310
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44819777842306247,
+      "learning_rate": 9.924932805436948e-07,
+      "loss": 0.2058,
+      "step": 1311
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9549652290680253,
+      "learning_rate": 9.924754417546367e-07,
+      "loss": 0.123,
+      "step": 1312
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6976385869512235,
+      "learning_rate": 9.924575819555861e-07,
+      "loss": 0.2911,
+      "step": 1313
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5428886474113841,
+      "learning_rate": 9.92439701147305e-07,
+      "loss": 0.4047,
+      "step": 1314
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.522051202353723,
+      "learning_rate": 9.924217993305563e-07,
+      "loss": 0.3151,
+      "step": 1315
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4459646433200106,
+      "learning_rate": 9.92403876506104e-07,
+      "loss": 0.1648,
+      "step": 1316
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5060715812929723,
+      "learning_rate": 9.923859326747124e-07,
+      "loss": 0.2347,
+      "step": 1317
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.31687712905787674,
+      "learning_rate": 9.92367967837147e-07,
+      "loss": 0.0478,
+      "step": 1318
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6603126960062933,
+      "learning_rate": 9.923499819941744e-07,
+      "loss": 0.2607,
+      "step": 1319
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6912679532061217,
+      "learning_rate": 9.923319751465615e-07,
+      "loss": 0.2336,
+      "step": 1320
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4936629031027776,
+      "learning_rate": 9.923139472950772e-07,
+      "loss": 0.0684,
+      "step": 1321
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4074506017586614,
+      "learning_rate": 9.922958984404901e-07,
+      "loss": 0.1365,
+      "step": 1322
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.06931220971631999,
+      "learning_rate": 9.922778285835704e-07,
+      "loss": 0.0039,
+      "step": 1323
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0584991047923242,
+      "learning_rate": 9.92259737725089e-07,
+      "loss": 0.1751,
+      "step": 1324
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42512537728409516,
+      "learning_rate": 9.922416258658173e-07,
+      "loss": 0.0221,
+      "step": 1325
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7174700906380793,
+      "learning_rate": 9.922234930065285e-07,
+      "loss": 0.1311,
+      "step": 1326
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5040507640209686,
+      "learning_rate": 9.922053391479961e-07,
+      "loss": 0.1044,
+      "step": 1327
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.47708684920990274,
+      "learning_rate": 9.921871642909944e-07,
+      "loss": 0.097,
+      "step": 1328
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.505050356460811,
+      "learning_rate": 9.92168968436299e-07,
+      "loss": 0.1123,
+      "step": 1329
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5969726272276742,
+      "learning_rate": 9.921507515846856e-07,
+      "loss": 0.1421,
+      "step": 1330
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.606016186383965,
+      "learning_rate": 9.92132513736932e-07,
+      "loss": 0.2951,
+      "step": 1331
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5882241937129672,
+      "learning_rate": 9.92114254893816e-07,
+      "loss": 0.2291,
+      "step": 1332
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.26234108070725803,
+      "learning_rate": 9.920959750561169e-07,
+      "loss": 0.069,
+      "step": 1333
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.26844022808681833,
+      "learning_rate": 9.920776742246142e-07,
+      "loss": 0.0168,
+      "step": 1334
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8127077765743058,
+      "learning_rate": 9.920593524000885e-07,
+      "loss": 0.4992,
+      "step": 1335
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4332594733357405,
+      "learning_rate": 9.920410095833217e-07,
+      "loss": 0.3344,
+      "step": 1336
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15581514468417898,
+      "learning_rate": 9.920226457750964e-07,
+      "loss": 0.0103,
+      "step": 1337
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5601738146747562,
+      "learning_rate": 9.920042609761961e-07,
+      "loss": 0.3145,
+      "step": 1338
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7410327563077606,
+      "learning_rate": 9.919858551874048e-07,
+      "loss": 0.4673,
+      "step": 1339
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.05010858722366252,
+      "learning_rate": 9.919674284095078e-07,
+      "loss": 0.0021,
+      "step": 1340
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28155993225973236,
+      "learning_rate": 9.919489806432914e-07,
+      "loss": 0.1986,
+      "step": 1341
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9192755585612271,
+      "learning_rate": 9.919305118895424e-07,
+      "loss": 0.4798,
+      "step": 1342
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0614597046619378,
+      "learning_rate": 9.919120221490492e-07,
+      "loss": 0.2899,
+      "step": 1343
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.16533559228097142,
+      "learning_rate": 9.918935114226e-07,
+      "loss": 0.0089,
+      "step": 1344
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8236466378681705,
+      "learning_rate": 9.918749797109848e-07,
+      "loss": 0.2607,
+      "step": 1345
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.26840855921958284,
+      "learning_rate": 9.918564270149942e-07,
+      "loss": 0.1592,
+      "step": 1346
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3758093577499254,
+      "learning_rate": 9.918378533354197e-07,
+      "loss": 0.115,
+      "step": 1347
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2804836172746173,
+      "learning_rate": 9.918192586730538e-07,
+      "loss": 0.2261,
+      "step": 1348
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8855784879112487,
+      "learning_rate": 9.918006430286893e-07,
+      "loss": 0.1226,
+      "step": 1349
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2574103117154745,
+      "learning_rate": 9.917820064031211e-07,
+      "loss": 0.1307,
+      "step": 1350
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6820728818513073,
+      "learning_rate": 9.917633487971438e-07,
+      "loss": 0.255,
+      "step": 1351
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4839934500545702,
+      "learning_rate": 9.917446702115533e-07,
+      "loss": 0.218,
+      "step": 1352
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9974031137381807,
+      "learning_rate": 9.917259706471467e-07,
+      "loss": 0.1613,
+      "step": 1353
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6288966585844071,
+      "learning_rate": 9.917072501047217e-07,
+      "loss": 0.3266,
+      "step": 1354
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5899033809330453,
+      "learning_rate": 9.91688508585077e-07,
+      "loss": 0.0135,
+      "step": 1355
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.48974260277611126,
+      "learning_rate": 9.91669746089012e-07,
+      "loss": 0.2727,
+      "step": 1356
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.892130586122228,
+      "learning_rate": 9.916509626173275e-07,
+      "loss": 0.2681,
+      "step": 1357
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5759666173641834,
+      "learning_rate": 9.916321581708245e-07,
+      "loss": 0.2102,
+      "step": 1358
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1415334650198193,
+      "learning_rate": 9.916133327503052e-07,
+      "loss": 0.0707,
+      "step": 1359
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2950143180306834,
+      "learning_rate": 9.915944863565728e-07,
+      "loss": 0.0184,
+      "step": 1360
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.622270261736345,
+      "learning_rate": 9.915756189904316e-07,
+      "loss": 0.2563,
+      "step": 1361
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6687547593629255,
+      "learning_rate": 9.915567306526862e-07,
+      "loss": 0.2629,
+      "step": 1362
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.813456436767718,
+      "learning_rate": 9.915378213441425e-07,
+      "loss": 0.059,
+      "step": 1363
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8296119799260199,
+      "learning_rate": 9.915188910656073e-07,
+      "loss": 0.1444,
+      "step": 1364
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7155707282550502,
+      "learning_rate": 9.91499939817888e-07,
+      "loss": 0.0603,
+      "step": 1365
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4621222371647438,
+      "learning_rate": 9.914809676017935e-07,
+      "loss": 0.1742,
+      "step": 1366
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1300100528341723,
+      "learning_rate": 9.914619744181326e-07,
+      "loss": 0.5811,
+      "step": 1367
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9614231224301606,
+      "learning_rate": 9.914429602677161e-07,
+      "loss": 0.3166,
+      "step": 1368
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0794458569950924,
+      "learning_rate": 9.914239251513549e-07,
+      "loss": 0.18,
+      "step": 1369
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6467544649663229,
+      "learning_rate": 9.91404869069861e-07,
+      "loss": 0.2362,
+      "step": 1370
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6154057398537042,
+      "learning_rate": 9.91385792024048e-07,
+      "loss": 0.084,
+      "step": 1371
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5645728368622048,
+      "learning_rate": 9.913666940147289e-07,
+      "loss": 0.3736,
+      "step": 1372
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6206794190429861,
+      "learning_rate": 9.91347575042719e-07,
+      "loss": 0.1729,
+      "step": 1373
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.46476261881865205,
+      "learning_rate": 9.913284351088338e-07,
+      "loss": 0.0415,
+      "step": 1374
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9571329913919955,
+      "learning_rate": 9.9130927421389e-07,
+      "loss": 0.4317,
+      "step": 1375
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5332665697299472,
+      "learning_rate": 9.912900923587047e-07,
+      "loss": 0.2812,
+      "step": 1376
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3770817447669728,
+      "learning_rate": 9.912708895440966e-07,
+      "loss": 0.0834,
+      "step": 1377
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6260119806099472,
+      "learning_rate": 9.912516657708847e-07,
+      "loss": 0.0474,
+      "step": 1378
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5076190706164107,
+      "learning_rate": 9.912324210398892e-07,
+      "loss": 0.2409,
+      "step": 1379
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5911689240300607,
+      "learning_rate": 9.91213155351931e-07,
+      "loss": 0.3276,
+      "step": 1380
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.527706445220485,
+      "learning_rate": 9.911938687078323e-07,
+      "loss": 0.1643,
+      "step": 1381
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6533182877886919,
+      "learning_rate": 9.911745611084156e-07,
+      "loss": 0.1749,
+      "step": 1382
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2121348501912437,
+      "learning_rate": 9.91155232554505e-07,
+      "loss": 0.2173,
+      "step": 1383
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5489449013024819,
+      "learning_rate": 9.911358830469247e-07,
+      "loss": 0.1405,
+      "step": 1384
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4199252225993026,
+      "learning_rate": 9.911165125865001e-07,
+      "loss": 0.2352,
+      "step": 1385
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5663838266086673,
+      "learning_rate": 9.91097121174058e-07,
+      "loss": 0.208,
+      "step": 1386
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.41766510964406783,
+      "learning_rate": 9.910777088104256e-07,
+      "loss": 0.0624,
+      "step": 1387
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7570359673996523,
+      "learning_rate": 9.910582754964306e-07,
+      "loss": 0.3536,
+      "step": 1388
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8684024534790326,
+      "learning_rate": 9.910388212329027e-07,
+      "loss": 0.2251,
+      "step": 1389
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0843837871077118,
+      "learning_rate": 9.910193460206716e-07,
+      "loss": 0.163,
+      "step": 1390
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6591320256330364,
+      "learning_rate": 9.90999849860568e-07,
+      "loss": 0.4135,
+      "step": 1391
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.38250931088260687,
+      "learning_rate": 9.909803327534239e-07,
+      "loss": 0.0429,
+      "step": 1392
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5092062891712076,
+      "learning_rate": 9.909607947000717e-07,
+      "loss": 0.2099,
+      "step": 1393
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5321071463340533,
+      "learning_rate": 9.90941235701345e-07,
+      "loss": 0.089,
+      "step": 1394
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5105319453775725,
+      "learning_rate": 9.909216557580784e-07,
+      "loss": 0.1912,
+      "step": 1395
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0240034951819497,
+      "learning_rate": 9.90902054871107e-07,
+      "loss": 0.2655,
+      "step": 1396
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.35749500684136865,
+      "learning_rate": 9.90882433041267e-07,
+      "loss": 0.0129,
+      "step": 1397
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7430532137923848,
+      "learning_rate": 9.908627902693957e-07,
+      "loss": 0.1514,
+      "step": 1398
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4790887423592187,
+      "learning_rate": 9.908431265563313e-07,
+      "loss": 0.1774,
+      "step": 1399
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8233676472290283,
+      "learning_rate": 9.90823441902912e-07,
+      "loss": 0.1219,
+      "step": 1400
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5939786949019252,
+      "learning_rate": 9.908037363099782e-07,
+      "loss": 0.0689,
+      "step": 1401
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5939627617523665,
+      "learning_rate": 9.907840097783704e-07,
+      "loss": 0.1783,
+      "step": 1402
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.49141165452616525,
+      "learning_rate": 9.9076426230893e-07,
+      "loss": 0.1545,
+      "step": 1403
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7278504336264148,
+      "learning_rate": 9.907444939024997e-07,
+      "loss": 0.2825,
+      "step": 1404
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28442755361920813,
+      "learning_rate": 9.907247045599226e-07,
+      "loss": 0.1271,
+      "step": 1405
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.45846238394230227,
+      "learning_rate": 9.907048942820432e-07,
+      "loss": 0.0251,
+      "step": 1406
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5143526782461665,
+      "learning_rate": 9.906850630697066e-07,
+      "loss": 0.2793,
+      "step": 1407
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6207469331297467,
+      "learning_rate": 9.906652109237588e-07,
+      "loss": 0.0464,
+      "step": 1408
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.874109918494243,
+      "learning_rate": 9.90645337845047e-07,
+      "loss": 0.2319,
+      "step": 1409
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4302973166798125,
+      "learning_rate": 9.906254438344185e-07,
+      "loss": 0.2829,
+      "step": 1410
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.37521259962395315,
+      "learning_rate": 9.906055288927221e-07,
+      "loss": 0.1903,
+      "step": 1411
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9284162240136209,
+      "learning_rate": 9.90585593020808e-07,
+      "loss": 0.359,
+      "step": 1412
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7100888419280851,
+      "learning_rate": 9.905656362195261e-07,
+      "loss": 0.2603,
+      "step": 1413
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.34959608011317916,
+      "learning_rate": 9.90545658489728e-07,
+      "loss": 0.2394,
+      "step": 1414
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4230948498712605,
+      "learning_rate": 9.90525659832266e-07,
+      "loss": 0.2377,
+      "step": 1415
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.48756417643515504,
+      "learning_rate": 9.905056402479933e-07,
+      "loss": 0.142,
+      "step": 1416
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.27732749330665324,
+      "learning_rate": 9.904855997377638e-07,
+      "loss": 0.1023,
+      "step": 1417
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.21866130982516352,
+      "learning_rate": 9.904655383024327e-07,
+      "loss": 0.0283,
+      "step": 1418
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4805676435876558,
+      "learning_rate": 9.90445455942856e-07,
+      "loss": 0.3483,
+      "step": 1419
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6717010660990245,
+      "learning_rate": 9.9042535265989e-07,
+      "loss": 0.0295,
+      "step": 1420
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42242652013031196,
+      "learning_rate": 9.904052284543925e-07,
+      "loss": 0.3359,
+      "step": 1421
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9660930421018699,
+      "learning_rate": 9.903850833272222e-07,
+      "loss": 0.4354,
+      "step": 1422
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4133245581699803,
+      "learning_rate": 9.903649172792386e-07,
+      "loss": 0.1681,
+      "step": 1423
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8389060031095329,
+      "learning_rate": 9.903447303113017e-07,
+      "loss": 0.3308,
+      "step": 1424
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2505562843920614,
+      "learning_rate": 9.90324522424273e-07,
+      "loss": 0.1088,
+      "step": 1425
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.46521286222480235,
+      "learning_rate": 9.903042936190145e-07,
+      "loss": 0.1423,
+      "step": 1426
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5018964774218596,
+      "learning_rate": 9.90284043896389e-07,
+      "loss": 0.2328,
+      "step": 1427
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5915417271474269,
+      "learning_rate": 9.90263773257261e-07,
+      "loss": 0.4089,
+      "step": 1428
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.524245641467094,
+      "learning_rate": 9.902434817024945e-07,
+      "loss": 0.3119,
+      "step": 1429
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4189219922182432,
+      "learning_rate": 9.902231692329556e-07,
+      "loss": 0.1085,
+      "step": 1430
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.276746429325565,
+      "learning_rate": 9.90202835849511e-07,
+      "loss": 0.3031,
+      "step": 1431
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8251646253861777,
+      "learning_rate": 9.901824815530277e-07,
+      "loss": 0.1873,
+      "step": 1432
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5937526854920018,
+      "learning_rate": 9.901621063443746e-07,
+      "loss": 0.1396,
+      "step": 1433
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4455253693138965,
+      "learning_rate": 9.901417102244207e-07,
+      "loss": 0.2131,
+      "step": 1434
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.582796698479084,
+      "learning_rate": 9.90121293194036e-07,
+      "loss": 0.5477,
+      "step": 1435
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4848414087429192,
+      "learning_rate": 9.901008552540918e-07,
+      "loss": 0.1672,
+      "step": 1436
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28593497405984064,
+      "learning_rate": 9.900803964054597e-07,
+      "loss": 0.0338,
+      "step": 1437
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6782419152092031,
+      "learning_rate": 9.900599166490129e-07,
+      "loss": 0.2368,
+      "step": 1438
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5081115265738643,
+      "learning_rate": 9.900394159856249e-07,
+      "loss": 0.3658,
+      "step": 1439
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9275593554404851,
+      "learning_rate": 9.900188944161701e-07,
+      "loss": 0.2623,
+      "step": 1440
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.400317797796568,
+      "learning_rate": 9.899983519415244e-07,
+      "loss": 0.1618,
+      "step": 1441
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8491504770105957,
+      "learning_rate": 9.89977788562564e-07,
+      "loss": 0.2061,
+      "step": 1442
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6179973314813763,
+      "learning_rate": 9.899572042801661e-07,
+      "loss": 0.1308,
+      "step": 1443
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6112656200925553,
+      "learning_rate": 9.89936599095209e-07,
+      "loss": 0.0589,
+      "step": 1444
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.48155520352962544,
+      "learning_rate": 9.899159730085714e-07,
+      "loss": 0.197,
+      "step": 1445
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2113338467664378,
+      "learning_rate": 9.898953260211337e-07,
+      "loss": 0.1121,
+      "step": 1446
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5039106787821244,
+      "learning_rate": 9.898746581337766e-07,
+      "loss": 0.2074,
+      "step": 1447
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4526936579946449,
+      "learning_rate": 9.89853969347382e-07,
+      "loss": 0.1398,
+      "step": 1448
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5788236603227206,
+      "learning_rate": 9.898332596628322e-07,
+      "loss": 0.1798,
+      "step": 1449
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.20559894082176153,
+      "learning_rate": 9.898125290810107e-07,
+      "loss": 0.1567,
+      "step": 1450
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6075111709699883,
+      "learning_rate": 9.897917776028022e-07,
+      "loss": 0.2502,
+      "step": 1451
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.31621268799203,
+      "learning_rate": 9.89771005229092e-07,
+      "loss": 0.0991,
+      "step": 1452
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.27102036808554036,
+      "learning_rate": 9.89750211960766e-07,
+      "loss": 0.1116,
+      "step": 1453
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.19562214729796376,
+      "learning_rate": 9.897293977987112e-07,
+      "loss": 0.1442,
+      "step": 1454
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2897769074366858,
+      "learning_rate": 9.897085627438162e-07,
+      "loss": 0.1174,
+      "step": 1455
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9699879807136473,
+      "learning_rate": 9.896877067969694e-07,
+      "loss": 0.3481,
+      "step": 1456
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4334282701949201,
+      "learning_rate": 9.896668299590606e-07,
+      "loss": 0.1089,
+      "step": 1457
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.263748101801689,
+      "learning_rate": 9.896459322309802e-07,
+      "loss": 0.0101,
+      "step": 1458
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5811943469647456,
+      "learning_rate": 9.896250136136203e-07,
+      "loss": 0.0801,
+      "step": 1459
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8318265996009885,
+      "learning_rate": 9.89604074107873e-07,
+      "loss": 0.1777,
+      "step": 1460
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2212493744552875,
+      "learning_rate": 9.895831137146318e-07,
+      "loss": 0.0888,
+      "step": 1461
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.26580603276275505,
+      "learning_rate": 9.895621324347908e-07,
+      "loss": 0.178,
+      "step": 1462
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.277147620795641,
+      "learning_rate": 9.895411302692448e-07,
+      "loss": 0.051,
+      "step": 1463
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5628289413299168,
+      "learning_rate": 9.895201072188903e-07,
+      "loss": 0.3389,
+      "step": 1464
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6444211883774726,
+      "learning_rate": 9.89499063284624e-07,
+      "loss": 0.1766,
+      "step": 1465
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4995024859452174,
+      "learning_rate": 9.894779984673433e-07,
+      "loss": 0.266,
+      "step": 1466
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5980645482596011,
+      "learning_rate": 9.894569127679476e-07,
+      "loss": 0.3054,
+      "step": 1467
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4433154938139935,
+      "learning_rate": 9.894358061873358e-07,
+      "loss": 0.0984,
+      "step": 1468
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2491109776587946,
+      "learning_rate": 9.894146787264088e-07,
+      "loss": 0.0295,
+      "step": 1469
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.36524548467967133,
+      "learning_rate": 9.893935303860677e-07,
+      "loss": 0.0157,
+      "step": 1470
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3223149265797452,
+      "learning_rate": 9.893723611672147e-07,
+      "loss": 0.4159,
+      "step": 1471
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5802469108703582,
+      "learning_rate": 9.89351171070753e-07,
+      "loss": 0.1877,
+      "step": 1472
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3402492212622589,
+      "learning_rate": 9.89329960097587e-07,
+      "loss": 0.1217,
+      "step": 1473
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7272394862500775,
+      "learning_rate": 9.893087282486208e-07,
+      "loss": 0.2007,
+      "step": 1474
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1860676868954452,
+      "learning_rate": 9.892874755247608e-07,
+      "loss": 0.1,
+      "step": 1475
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.533114804369231,
+      "learning_rate": 9.892662019269136e-07,
+      "loss": 0.2238,
+      "step": 1476
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6122280427665033,
+      "learning_rate": 9.892449074559864e-07,
+      "loss": 0.039,
+      "step": 1477
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6477653508701525,
+      "learning_rate": 9.892235921128881e-07,
+      "loss": 0.2751,
+      "step": 1478
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6317612010148661,
+      "learning_rate": 9.89202255898528e-07,
+      "loss": 0.2251,
+      "step": 1479
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.47801365439872195,
+      "learning_rate": 9.89180898813816e-07,
+      "loss": 0.1295,
+      "step": 1480
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3439080532007902,
+      "learning_rate": 9.891595208596638e-07,
+      "loss": 0.161,
+      "step": 1481
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2094726595885454,
+      "learning_rate": 9.891381220369827e-07,
+      "loss": 0.0166,
+      "step": 1482
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4794332543917334,
+      "learning_rate": 9.891167023466864e-07,
+      "loss": 0.1335,
+      "step": 1483
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6044568815496913,
+      "learning_rate": 9.890952617896882e-07,
+      "loss": 0.3535,
+      "step": 1484
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.36745720152941697,
+      "learning_rate": 9.890738003669027e-07,
+      "loss": 0.0173,
+      "step": 1485
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7129460235783216,
+      "learning_rate": 9.89052318079246e-07,
+      "loss": 0.0579,
+      "step": 1486
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0361603130817678,
+      "learning_rate": 9.890308149276342e-07,
+      "loss": 0.1412,
+      "step": 1487
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.18823015993972705,
+      "learning_rate": 9.890092909129848e-07,
+      "loss": 0.0465,
+      "step": 1488
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3016928228509086,
+      "learning_rate": 9.88987746036216e-07,
+      "loss": 0.1725,
+      "step": 1489
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3818409880923631,
+      "learning_rate": 9.889661802982468e-07,
+      "loss": 0.251,
+      "step": 1490
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3649233229388234,
+      "learning_rate": 9.889445936999976e-07,
+      "loss": 0.2045,
+      "step": 1491
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9295062256605674,
+      "learning_rate": 9.889229862423892e-07,
+      "loss": 0.1487,
+      "step": 1492
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6602399601320765,
+      "learning_rate": 9.889013579263433e-07,
+      "loss": 0.2343,
+      "step": 1493
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6796608325296203,
+      "learning_rate": 9.888797087527826e-07,
+      "loss": 0.2295,
+      "step": 1494
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3587957490130416,
+      "learning_rate": 9.888580387226307e-07,
+      "loss": 0.0863,
+      "step": 1495
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5379181859428669,
+      "learning_rate": 9.88836347836812e-07,
+      "loss": 0.2108,
+      "step": 1496
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2348044214295755,
+      "learning_rate": 9.88814636096252e-07,
+      "loss": 0.3544,
+      "step": 1497
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.294621699605234,
+      "learning_rate": 9.887929035018773e-07,
+      "loss": 0.1577,
+      "step": 1498
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1004985743456288,
+      "learning_rate": 9.887711500546147e-07,
+      "loss": 0.1324,
+      "step": 1499
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.44690192051575695,
+      "learning_rate": 9.887493757553923e-07,
+      "loss": 0.3278,
+      "step": 1500
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7855658632665525,
+      "learning_rate": 9.887275806051388e-07,
+      "loss": 0.3881,
+      "step": 1501
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9292787724738459,
+      "learning_rate": 9.887057646047846e-07,
+      "loss": 0.2462,
+      "step": 1502
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.712761817448678,
+      "learning_rate": 9.886839277552597e-07,
+      "loss": 0.0251,
+      "step": 1503
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5377398737419191,
+      "learning_rate": 9.886620700574962e-07,
+      "loss": 0.2777,
+      "step": 1504
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3413670328124251,
+      "learning_rate": 9.886401915124263e-07,
+      "loss": 0.1447,
+      "step": 1505
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5039823574572356,
+      "learning_rate": 9.886182921209839e-07,
+      "loss": 0.2739,
+      "step": 1506
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4076212630655461,
+      "learning_rate": 9.885963718841025e-07,
+      "loss": 0.2969,
+      "step": 1507
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.44300072813030744,
+      "learning_rate": 9.88574430802718e-07,
+      "loss": 0.2196,
+      "step": 1508
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5676336561032818,
+      "learning_rate": 9.88552468877766e-07,
+      "loss": 0.2615,
+      "step": 1509
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.93740836315491,
+      "learning_rate": 9.885304861101835e-07,
+      "loss": 0.2727,
+      "step": 1510
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8297037804825235,
+      "learning_rate": 9.885084825009084e-07,
+      "loss": 0.3854,
+      "step": 1511
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.640175390536145,
+      "learning_rate": 9.884864580508795e-07,
+      "loss": 0.4684,
+      "step": 1512
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.861467133913177,
+      "learning_rate": 9.884644127610365e-07,
+      "loss": 0.2164,
+      "step": 1513
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2962211416169949,
+      "learning_rate": 9.884423466323192e-07,
+      "loss": 0.1421,
+      "step": 1514
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.22972637966336357,
+      "learning_rate": 9.8842025966567e-07,
+      "loss": 0.1644,
+      "step": 1515
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.26510495004736484,
+      "learning_rate": 9.883981518620304e-07,
+      "loss": 0.104,
+      "step": 1516
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4411577210188472,
+      "learning_rate": 9.88376023222344e-07,
+      "loss": 0.0874,
+      "step": 1517
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4871797030131832,
+      "learning_rate": 9.883538737475544e-07,
+      "loss": 0.2601,
+      "step": 1518
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7531400614799489,
+      "learning_rate": 9.883317034386073e-07,
+      "loss": 0.3019,
+      "step": 1519
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7616916491544177,
+      "learning_rate": 9.883095122964477e-07,
+      "loss": 0.4258,
+      "step": 1520
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4315192948478817,
+      "learning_rate": 9.882873003220227e-07,
+      "loss": 0.2612,
+      "step": 1521
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.33962815028522214,
+      "learning_rate": 9.8826506751628e-07,
+      "loss": 0.0532,
+      "step": 1522
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.31640458172068525,
+      "learning_rate": 9.88242813880168e-07,
+      "loss": 0.2545,
+      "step": 1523
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.24711460479640185,
+      "learning_rate": 9.88220539414636e-07,
+      "loss": 0.1543,
+      "step": 1524
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5478387891792209,
+      "learning_rate": 9.881982441206343e-07,
+      "loss": 0.4665,
+      "step": 1525
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.097502839498695,
+      "learning_rate": 9.881759279991143e-07,
+      "loss": 0.0322,
+      "step": 1526
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3476195653312959,
+      "learning_rate": 9.881535910510276e-07,
+      "loss": 0.0366,
+      "step": 1527
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5333026791706978,
+      "learning_rate": 9.881312332773275e-07,
+      "loss": 0.4375,
+      "step": 1528
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3112252104360692,
+      "learning_rate": 9.881088546789677e-07,
+      "loss": 0.1351,
+      "step": 1529
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4399471696451268,
+      "learning_rate": 9.88086455256903e-07,
+      "loss": 0.1491,
+      "step": 1530
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6295850881492374,
+      "learning_rate": 9.88064035012089e-07,
+      "loss": 0.0689,
+      "step": 1531
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.38561365987020313,
+      "learning_rate": 9.88041593945482e-07,
+      "loss": 0.2467,
+      "step": 1532
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5938699720510816,
+      "learning_rate": 9.880191320580396e-07,
+      "loss": 0.3254,
+      "step": 1533
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1697906908430658,
+      "learning_rate": 9.879966493507198e-07,
+      "loss": 0.0077,
+      "step": 1534
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3496812653264161,
+      "learning_rate": 9.879741458244822e-07,
+      "loss": 0.0746,
+      "step": 1535
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.31029222651159144,
+      "learning_rate": 9.879516214802866e-07,
+      "loss": 0.216,
+      "step": 1536
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4859938049276853,
+      "learning_rate": 9.879290763190937e-07,
+      "loss": 0.15,
+      "step": 1537
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.37417699629811346,
+      "learning_rate": 9.879065103418657e-07,
+      "loss": 0.0962,
+      "step": 1538
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5437353087596696,
+      "learning_rate": 9.87883923549565e-07,
+      "loss": 0.145,
+      "step": 1539
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8925268366823884,
+      "learning_rate": 9.878613159431554e-07,
+      "loss": 0.2945,
+      "step": 1540
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5877766008896336,
+      "learning_rate": 9.878386875236016e-07,
+      "loss": 0.17,
+      "step": 1541
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.13553921863758533,
+      "learning_rate": 9.878160382918683e-07,
+      "loss": 0.0674,
+      "step": 1542
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.614686376491471,
+      "learning_rate": 9.877933682489224e-07,
+      "loss": 0.154,
+      "step": 1543
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4224147413842552,
+      "learning_rate": 9.877706773957308e-07,
+      "loss": 0.2191,
+      "step": 1544
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6719867186441478,
+      "learning_rate": 9.877479657332617e-07,
+      "loss": 0.2382,
+      "step": 1545
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.20322204392227114,
+      "learning_rate": 9.877252332624837e-07,
+      "loss": 0.0097,
+      "step": 1546
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8891487395054584,
+      "learning_rate": 9.877024799843667e-07,
+      "loss": 0.0891,
+      "step": 1547
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1298489880351104,
+      "learning_rate": 9.876797058998817e-07,
+      "loss": 0.165,
+      "step": 1548
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.514652923651593,
+      "learning_rate": 9.8765691101e-07,
+      "loss": 0.1222,
+      "step": 1549
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6397997624940633,
+      "learning_rate": 9.876340953156943e-07,
+      "loss": 0.2687,
+      "step": 1550
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7785255189564557,
+      "learning_rate": 9.876112588179376e-07,
+      "loss": 0.1656,
+      "step": 1551
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7121913562734088,
+      "learning_rate": 9.875884015177045e-07,
+      "loss": 0.2444,
+      "step": 1552
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6605143159514306,
+      "learning_rate": 9.8756552341597e-07,
+      "loss": 0.2287,
+      "step": 1553
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.36886705029343614,
+      "learning_rate": 9.875426245137101e-07,
+      "loss": 0.1895,
+      "step": 1554
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7679070950832787,
+      "learning_rate": 9.87519704811902e-07,
+      "loss": 0.2849,
+      "step": 1555
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6701865236270271,
+      "learning_rate": 9.87496764311523e-07,
+      "loss": 0.2261,
+      "step": 1556
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.42004633192356616,
+      "learning_rate": 9.874738030135521e-07,
+      "loss": 0.2492,
+      "step": 1557
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.339808480879583,
+      "learning_rate": 9.874508209189689e-07,
+      "loss": 0.2003,
+      "step": 1558
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7761775197322209,
+      "learning_rate": 9.874278180287536e-07,
+      "loss": 0.2574,
+      "step": 1559
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9695043231557823,
+      "learning_rate": 9.874047943438878e-07,
+      "loss": 0.2385,
+      "step": 1560
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.004971613166862,
+      "learning_rate": 9.873817498653537e-07,
+      "loss": 0.2249,
+      "step": 1561
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5756796278033042,
+      "learning_rate": 9.873586845941344e-07,
+      "loss": 0.1366,
+      "step": 1562
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2083389990401696,
+      "learning_rate": 9.87335598531214e-07,
+      "loss": 0.1626,
+      "step": 1563
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9287886982072556,
+      "learning_rate": 9.87312491677577e-07,
+      "loss": 0.0379,
+      "step": 1564
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4611163777873519,
+      "learning_rate": 9.872893640342098e-07,
+      "loss": 0.1681,
+      "step": 1565
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6474301274319696,
+      "learning_rate": 9.872662156020986e-07,
+      "loss": 0.2599,
+      "step": 1566
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.41372820467170063,
+      "learning_rate": 9.872430463822312e-07,
+      "loss": 0.2448,
+      "step": 1567
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5639025728086583,
+      "learning_rate": 9.872198563755959e-07,
+      "loss": 0.2386,
+      "step": 1568
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3475531518259124,
+      "learning_rate": 9.87196645583182e-07,
+      "loss": 0.1789,
+      "step": 1569
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6908444271896197,
+      "learning_rate": 9.8717341400598e-07,
+      "loss": 0.1217,
+      "step": 1570
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5832040661187229,
+      "learning_rate": 9.871501616449806e-07,
+      "loss": 0.1626,
+      "step": 1571
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3531856618380542,
+      "learning_rate": 9.87126888501176e-07,
+      "loss": 0.1372,
+      "step": 1572
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.48973224252772235,
+      "learning_rate": 9.871035945755593e-07,
+      "loss": 0.098,
+      "step": 1573
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40065693580039274,
+      "learning_rate": 9.87080279869124e-07,
+      "loss": 0.0152,
+      "step": 1574
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6889147943802033,
+      "learning_rate": 9.87056944382865e-07,
+      "loss": 0.3114,
+      "step": 1575
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3608933592686679,
+      "learning_rate": 9.870335881177773e-07,
+      "loss": 0.2242,
+      "step": 1576
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7705750279452553,
+      "learning_rate": 9.870102110748577e-07,
+      "loss": 0.4654,
+      "step": 1577
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5027787671169268,
+      "learning_rate": 9.869868132551036e-07,
+      "loss": 0.0991,
+      "step": 1578
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.45869951418894883,
+      "learning_rate": 9.86963394659513e-07,
+      "loss": 0.1745,
+      "step": 1579
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5005370133604695,
+      "learning_rate": 9.869399552890852e-07,
+      "loss": 0.1244,
+      "step": 1580
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6014441781373475,
+      "learning_rate": 9.8691649514482e-07,
+      "loss": 0.522,
+      "step": 1581
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4004666041309846,
+      "learning_rate": 9.868930142277183e-07,
+      "loss": 0.0131,
+      "step": 1582
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.503409687553413,
+      "learning_rate": 9.868695125387817e-07,
+      "loss": 0.2341,
+      "step": 1583
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.226301001193882,
+      "learning_rate": 9.868459900790131e-07,
+      "loss": 0.2083,
+      "step": 1584
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.36025836315339727,
+      "learning_rate": 9.868224468494158e-07,
+      "loss": 0.2862,
+      "step": 1585
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6341257291602369,
+      "learning_rate": 9.867988828509943e-07,
+      "loss": 0.0529,
+      "step": 1586
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6686417639942993,
+      "learning_rate": 9.86775298084754e-07,
+      "loss": 0.2086,
+      "step": 1587
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15439815770014853,
+      "learning_rate": 9.867516925517008e-07,
+      "loss": 0.0469,
+      "step": 1588
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6521696811939802,
+      "learning_rate": 9.867280662528418e-07,
+      "loss": 0.2327,
+      "step": 1589
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4687022140046831,
+      "learning_rate": 9.867044191891853e-07,
+      "loss": 0.016,
+      "step": 1590
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.42445870053570794,
+      "learning_rate": 9.866807513617396e-07,
+      "loss": 0.2598,
+      "step": 1591
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4178296998710727,
+      "learning_rate": 9.86657062771515e-07,
+      "loss": 0.3339,
+      "step": 1592
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1190587607436973,
+      "learning_rate": 9.866333534195214e-07,
+      "loss": 0.0508,
+      "step": 1593
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7492173303820396,
+      "learning_rate": 9.86609623306771e-07,
+      "loss": 0.0173,
+      "step": 1594
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.17268774248057586,
+      "learning_rate": 9.86585872434276e-07,
+      "loss": 0.0098,
+      "step": 1595
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6882161608394176,
+      "learning_rate": 9.865621008030492e-07,
+      "loss": 0.4026,
+      "step": 1596
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.18914030733008877,
+      "learning_rate": 9.865383084141051e-07,
+      "loss": 0.0209,
+      "step": 1597
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8754944883215988,
+      "learning_rate": 9.865144952684588e-07,
+      "loss": 0.064,
+      "step": 1598
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2758435168021218,
+      "learning_rate": 9.864906613671262e-07,
+      "loss": 0.1013,
+      "step": 1599
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6569159055960563,
+      "learning_rate": 9.864668067111238e-07,
+      "loss": 0.2068,
+      "step": 1600
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4573223639888256,
+      "learning_rate": 9.864429313014698e-07,
+      "loss": 0.245,
+      "step": 1601
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6549498989924214,
+      "learning_rate": 9.86419035139182e-07,
+      "loss": 0.0784,
+      "step": 1602
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4307212536134745,
+      "learning_rate": 9.863951182252808e-07,
+      "loss": 0.0911,
+      "step": 1603
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9246695738009061,
+      "learning_rate": 9.863711805607858e-07,
+      "loss": 0.1848,
+      "step": 1604
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8584093011942867,
+      "learning_rate": 9.863472221467188e-07,
+      "loss": 0.2089,
+      "step": 1605
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1959506485643055,
+      "learning_rate": 9.863232429841013e-07,
+      "loss": 0.0927,
+      "step": 1606
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1312076550436183,
+      "learning_rate": 9.862992430739569e-07,
+      "loss": 0.2254,
+      "step": 1607
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7194644325458003,
+      "learning_rate": 9.862752224173089e-07,
+      "loss": 0.1336,
+      "step": 1608
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.35806296749475824,
+      "learning_rate": 9.862511810151827e-07,
+      "loss": 0.091,
+      "step": 1609
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3224249152615377,
+      "learning_rate": 9.862271188686036e-07,
+      "loss": 0.1694,
+      "step": 1610
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.35529476439744356,
+      "learning_rate": 9.86203035978598e-07,
+      "loss": 0.084,
+      "step": 1611
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4637912150396062,
+      "learning_rate": 9.861789323461936e-07,
+      "loss": 0.1505,
+      "step": 1612
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.695547042266105,
+      "learning_rate": 9.861548079724184e-07,
+      "loss": 0.1855,
+      "step": 1613
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9573012386923306,
+      "learning_rate": 9.86130662858302e-07,
+      "loss": 0.4129,
+      "step": 1614
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8909234017172247,
+      "learning_rate": 9.861064970048742e-07,
+      "loss": 0.3612,
+      "step": 1615
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5461979307132618,
+      "learning_rate": 9.860823104131661e-07,
+      "loss": 0.1456,
+      "step": 1616
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.450693533453506,
+      "learning_rate": 9.860581030842094e-07,
+      "loss": 0.3491,
+      "step": 1617
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5700827078183561,
+      "learning_rate": 9.86033875019037e-07,
+      "loss": 0.236,
+      "step": 1618
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6427023992944825,
+      "learning_rate": 9.860096262186822e-07,
+      "loss": 0.1428,
+      "step": 1619
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5089999767887992,
+      "learning_rate": 9.8598535668418e-07,
+      "loss": 0.1872,
+      "step": 1620
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.22674102403414914,
+      "learning_rate": 9.859610664165657e-07,
+      "loss": 0.0153,
+      "step": 1621
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6010049723526243,
+      "learning_rate": 9.85936755416875e-07,
+      "loss": 0.2255,
+      "step": 1622
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.46927568870917785,
+      "learning_rate": 9.859124236861458e-07,
+      "loss": 0.2207,
+      "step": 1623
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6100554901334297,
+      "learning_rate": 9.858880712254156e-07,
+      "loss": 0.3904,
+      "step": 1624
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5253195391980957,
+      "learning_rate": 9.858636980357236e-07,
+      "loss": 0.1064,
+      "step": 1625
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3186766994435842,
+      "learning_rate": 9.858393041181094e-07,
+      "loss": 0.0091,
+      "step": 1626
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.49194077776033,
+      "learning_rate": 9.85814889473614e-07,
+      "loss": 0.1151,
+      "step": 1627
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3014511666074654,
+      "learning_rate": 9.857904541032788e-07,
+      "loss": 0.0477,
+      "step": 1628
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6800381451103563,
+      "learning_rate": 9.857659980081462e-07,
+      "loss": 0.2274,
+      "step": 1629
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5048897009278337,
+      "learning_rate": 9.857415211892597e-07,
+      "loss": 0.2173,
+      "step": 1630
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.38878687723503236,
+      "learning_rate": 9.857170236476634e-07,
+      "loss": 0.1196,
+      "step": 1631
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4921853470158305,
+      "learning_rate": 9.856925053844024e-07,
+      "loss": 0.2371,
+      "step": 1632
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6741135439542971,
+      "learning_rate": 9.856679664005227e-07,
+      "loss": 0.2392,
+      "step": 1633
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7293732326588783,
+      "learning_rate": 9.856434066970713e-07,
+      "loss": 0.4094,
+      "step": 1634
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2996126678591897,
+      "learning_rate": 9.85618826275096e-07,
+      "loss": 0.3037,
+      "step": 1635
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.10446453277853311,
+      "learning_rate": 9.855942251356452e-07,
+      "loss": 0.0599,
+      "step": 1636
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7036396263243663,
+      "learning_rate": 9.855696032797687e-07,
+      "loss": 0.1621,
+      "step": 1637
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6439254075434799,
+      "learning_rate": 9.855449607085168e-07,
+      "loss": 0.146,
+      "step": 1638
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0861699865315655,
+      "learning_rate": 9.855202974229407e-07,
+      "loss": 0.1874,
+      "step": 1639
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7652036360756149,
+      "learning_rate": 9.854956134240929e-07,
+      "loss": 0.2367,
+      "step": 1640
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.491924900336481,
+      "learning_rate": 9.85470908713026e-07,
+      "loss": 0.2121,
+      "step": 1641
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.09698943161845,
+      "learning_rate": 9.854461832907943e-07,
+      "loss": 0.2639,
+      "step": 1642
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3646403434772786,
+      "learning_rate": 9.854214371584526e-07,
+      "loss": 0.1764,
+      "step": 1643
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5071519005871684,
+      "learning_rate": 9.853966703170566e-07,
+      "loss": 0.1177,
+      "step": 1644
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6921879817661868,
+      "learning_rate": 9.853718827676628e-07,
+      "loss": 0.3415,
+      "step": 1645
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1163835110493454,
+      "learning_rate": 9.853470745113288e-07,
+      "loss": 0.1166,
+      "step": 1646
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.25022652331553863,
+      "learning_rate": 9.85322245549113e-07,
+      "loss": 0.0072,
+      "step": 1647
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6749871432923831,
+      "learning_rate": 9.852973958820746e-07,
+      "loss": 0.176,
+      "step": 1648
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4975505186456713,
+      "learning_rate": 9.852725255112734e-07,
+      "loss": 0.1886,
+      "step": 1649
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6232905693402038,
+      "learning_rate": 9.85247634437771e-07,
+      "loss": 0.3126,
+      "step": 1650
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8002825993038455,
+      "learning_rate": 9.852227226626292e-07,
+      "loss": 0.1474,
+      "step": 1651
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10381943547798866,
+      "learning_rate": 9.851977901869105e-07,
+      "loss": 0.0839,
+      "step": 1652
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7543358346895345,
+      "learning_rate": 9.851728370116786e-07,
+      "loss": 0.3721,
+      "step": 1653
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8141386475697114,
+      "learning_rate": 9.851478631379982e-07,
+      "loss": 0.1532,
+      "step": 1654
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6566212458996052,
+      "learning_rate": 9.851228685669347e-07,
+      "loss": 0.2333,
+      "step": 1655
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3501251624972066,
+      "learning_rate": 9.850978532995545e-07,
+      "loss": 0.1383,
+      "step": 1656
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7267284790649643,
+      "learning_rate": 9.850728173369246e-07,
+      "loss": 0.228,
+      "step": 1657
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.581595390627329,
+      "learning_rate": 9.850477606801132e-07,
+      "loss": 0.2769,
+      "step": 1658
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4579902043564122,
+      "learning_rate": 9.850226833301892e-07,
+      "loss": 0.2163,
+      "step": 1659
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9122505674135036,
+      "learning_rate": 9.849975852882226e-07,
+      "loss": 0.057,
+      "step": 1660
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3111394917274574,
+      "learning_rate": 9.84972466555284e-07,
+      "loss": 0.0714,
+      "step": 1661
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8783409127064085,
+      "learning_rate": 9.849473271324452e-07,
+      "loss": 0.3154,
+      "step": 1662
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8169232392745213,
+      "learning_rate": 9.849221670207783e-07,
+      "loss": 0.1238,
+      "step": 1663
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.34584369689229244,
+      "learning_rate": 9.848969862213572e-07,
+      "loss": 0.0439,
+      "step": 1664
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.20439358764503512,
+      "learning_rate": 9.848717847352557e-07,
+      "loss": 0.0118,
+      "step": 1665
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8297774778868178,
+      "learning_rate": 9.848465625635494e-07,
+      "loss": 0.0591,
+      "step": 1666
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.21684237184432614,
+      "learning_rate": 9.848213197073138e-07,
+      "loss": 0.0772,
+      "step": 1667
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5148735373998339,
+      "learning_rate": 9.847960561676263e-07,
+      "loss": 0.3261,
+      "step": 1668
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6875334741208333,
+      "learning_rate": 9.847707719455643e-07,
+      "loss": 0.2342,
+      "step": 1669
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.21305224593370214,
+      "learning_rate": 9.847454670422067e-07,
+      "loss": 0.0965,
+      "step": 1670
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.30857334239297074,
+      "learning_rate": 9.84720141458633e-07,
+      "loss": 0.1509,
+      "step": 1671
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5982979108595999,
+      "learning_rate": 9.846947951959237e-07,
+      "loss": 0.1248,
+      "step": 1672
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3881707818725426,
+      "learning_rate": 9.846694282551601e-07,
+      "loss": 0.0306,
+      "step": 1673
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7056799214476155,
+      "learning_rate": 9.846440406374244e-07,
+      "loss": 0.1658,
+      "step": 1674
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8837410663483006,
+      "learning_rate": 9.846186323437998e-07,
+      "loss": 0.2022,
+      "step": 1675
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8431493174441403,
+      "learning_rate": 9.8459320337537e-07,
+      "loss": 0.3112,
+      "step": 1676
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.24393374157755976,
+      "learning_rate": 9.8456775373322e-07,
+      "loss": 0.0037,
+      "step": 1677
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5249565529368295,
+      "learning_rate": 9.845422834184354e-07,
+      "loss": 0.2402,
+      "step": 1678
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5373256341212908,
+      "learning_rate": 9.84516792432103e-07,
+      "loss": 0.2174,
+      "step": 1679
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5148799176302914,
+      "learning_rate": 9.844912807753102e-07,
+      "loss": 0.0178,
+      "step": 1680
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2654821914220962,
+      "learning_rate": 9.844657484491455e-07,
+      "loss": 0.5374,
+      "step": 1681
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.34012022508386563,
+      "learning_rate": 9.844401954546982e-07,
+      "loss": 0.2204,
+      "step": 1682
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6504776717129892,
+      "learning_rate": 9.84414621793058e-07,
+      "loss": 0.4022,
+      "step": 1683
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5094296043064444,
+      "learning_rate": 9.843890274653164e-07,
+      "loss": 0.2673,
+      "step": 1684
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9684770030115625,
+      "learning_rate": 9.843634124725653e-07,
+      "loss": 0.1952,
+      "step": 1685
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6669948245222741,
+      "learning_rate": 9.843377768158971e-07,
+      "loss": 0.2342,
+      "step": 1686
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4466624802946475,
+      "learning_rate": 9.843121204964057e-07,
+      "loss": 0.1335,
+      "step": 1687
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.60892533892139,
+      "learning_rate": 9.842864435151859e-07,
+      "loss": 0.3891,
+      "step": 1688
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4968500382457546,
+      "learning_rate": 9.842607458733325e-07,
+      "loss": 0.2572,
+      "step": 1689
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.670389115941225,
+      "learning_rate": 9.842350275719426e-07,
+      "loss": 0.0334,
+      "step": 1690
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6025367438324835,
+      "learning_rate": 9.842092886121127e-07,
+      "loss": 0.1054,
+      "step": 1691
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4752594685453358,
+      "learning_rate": 9.841835289949412e-07,
+      "loss": 0.2078,
+      "step": 1692
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7298000651694033,
+      "learning_rate": 9.84157748721527e-07,
+      "loss": 0.0499,
+      "step": 1693
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.44105453765342617,
+      "learning_rate": 9.841319477929702e-07,
+      "loss": 0.2132,
+      "step": 1694
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5848802588162659,
+      "learning_rate": 9.841061262103712e-07,
+      "loss": 0.3941,
+      "step": 1695
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5989033355167053,
+      "learning_rate": 9.840802839748313e-07,
+      "loss": 0.2946,
+      "step": 1696
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4800551168081081,
+      "learning_rate": 9.84054421087454e-07,
+      "loss": 0.1525,
+      "step": 1697
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.17172942266235122,
+      "learning_rate": 9.840285375493416e-07,
+      "loss": 0.0979,
+      "step": 1698
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6392176185366084,
+      "learning_rate": 9.840026333615987e-07,
+      "loss": 0.0169,
+      "step": 1699
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.846588000790528,
+      "learning_rate": 9.839767085253307e-07,
+      "loss": 0.2467,
+      "step": 1700
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9009554247070751,
+      "learning_rate": 9.839507630416436e-07,
+      "loss": 0.1856,
+      "step": 1701
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.30164687180032423,
+      "learning_rate": 9.839247969116437e-07,
+      "loss": 0.2101,
+      "step": 1702
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0101689422412143,
+      "learning_rate": 9.838988101364394e-07,
+      "loss": 0.0621,
+      "step": 1703
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6279014376251323,
+      "learning_rate": 9.838728027171388e-07,
+      "loss": 0.3739,
+      "step": 1704
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4724623998971567,
+      "learning_rate": 9.83846774654852e-07,
+      "loss": 0.3858,
+      "step": 1705
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3424337059410447,
+      "learning_rate": 9.83820725950689e-07,
+      "loss": 0.343,
+      "step": 1706
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2760635915213745,
+      "learning_rate": 9.837946566057614e-07,
+      "loss": 0.1629,
+      "step": 1707
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7490635728472637,
+      "learning_rate": 9.83768566621181e-07,
+      "loss": 0.1155,
+      "step": 1708
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8384511238522884,
+      "learning_rate": 9.837424559980612e-07,
+      "loss": 0.1987,
+      "step": 1709
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8215913083693003,
+      "learning_rate": 9.837163247375157e-07,
+      "loss": 0.2006,
+      "step": 1710
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5060243417944514,
+      "learning_rate": 9.836901728406594e-07,
+      "loss": 0.3762,
+      "step": 1711
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5476501914212019,
+      "learning_rate": 9.83664000308608e-07,
+      "loss": 0.2034,
+      "step": 1712
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.44823886681977815,
+      "learning_rate": 9.83637807142478e-07,
+      "loss": 0.2989,
+      "step": 1713
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5106181915582283,
+      "learning_rate": 9.83611593343387e-07,
+      "loss": 0.2934,
+      "step": 1714
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.282226581431579,
+      "learning_rate": 9.835853589124531e-07,
+      "loss": 0.2696,
+      "step": 1715
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6971792264132003,
+      "learning_rate": 9.83559103850796e-07,
+      "loss": 0.2182,
+      "step": 1716
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.18208110199973887,
+      "learning_rate": 9.835328281595351e-07,
+      "loss": 0.1607,
+      "step": 1717
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7076367669794615,
+      "learning_rate": 9.83506531839792e-07,
+      "loss": 0.3098,
+      "step": 1718
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7060300322733915,
+      "learning_rate": 9.834802148926882e-07,
+      "loss": 0.1326,
+      "step": 1719
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5726378716837955,
+      "learning_rate": 9.834538773193463e-07,
+      "loss": 0.2372,
+      "step": 1720
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.09794177964706838,
+      "learning_rate": 9.834275191208902e-07,
+      "loss": 0.004,
+      "step": 1721
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.33448780003817236,
+      "learning_rate": 9.834011402984445e-07,
+      "loss": 0.119,
+      "step": 1722
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.570240963744505,
+      "learning_rate": 9.833747408531344e-07,
+      "loss": 0.4368,
+      "step": 1723
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.553949900799787,
+      "learning_rate": 9.833483207860859e-07,
+      "loss": 0.0991,
+      "step": 1724
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.29333690283732444,
+      "learning_rate": 9.833218800984266e-07,
+      "loss": 0.0875,
+      "step": 1725
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3069001815477144,
+      "learning_rate": 9.832954187912843e-07,
+      "loss": 0.0407,
+      "step": 1726
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5057713466762443,
+      "learning_rate": 9.832689368657879e-07,
+      "loss": 0.2316,
+      "step": 1727
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4303615533563246,
+      "learning_rate": 9.83242434323067e-07,
+      "loss": 0.0915,
+      "step": 1728
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6236747153961416,
+      "learning_rate": 9.832159111642526e-07,
+      "loss": 0.0566,
+      "step": 1729
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3078982598280512,
+      "learning_rate": 9.831893673904759e-07,
+      "loss": 0.1209,
+      "step": 1730
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2862679414751008,
+      "learning_rate": 9.831628030028696e-07,
+      "loss": 0.2588,
+      "step": 1731
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3677271815970271,
+      "learning_rate": 9.831362180025666e-07,
+      "loss": 0.3474,
+      "step": 1732
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6741049769726314,
+      "learning_rate": 9.831096123907015e-07,
+      "loss": 0.2151,
+      "step": 1733
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6127332879081245,
+      "learning_rate": 9.83082986168409e-07,
+      "loss": 0.2339,
+      "step": 1734
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.31925479786507616,
+      "learning_rate": 9.830563393368255e-07,
+      "loss": 0.0889,
+      "step": 1735
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.48704935384966197,
+      "learning_rate": 9.830296718970872e-07,
+      "loss": 0.1856,
+      "step": 1736
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7564513724340943,
+      "learning_rate": 9.830029838503322e-07,
+      "loss": 0.1297,
+      "step": 1737
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8743928928257639,
+      "learning_rate": 9.829762751976991e-07,
+      "loss": 0.3113,
+      "step": 1738
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4996887780505269,
+      "learning_rate": 9.82949545940327e-07,
+      "loss": 0.2354,
+      "step": 1739
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7579165889952131,
+      "learning_rate": 9.829227960793564e-07,
+      "loss": 0.196,
+      "step": 1740
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2615387161682434,
+      "learning_rate": 9.828960256159287e-07,
+      "loss": 0.0289,
+      "step": 1741
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3813425408339454,
+      "learning_rate": 9.828692345511857e-07,
+      "loss": 0.1092,
+      "step": 1742
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.27618223710505885,
+      "learning_rate": 9.828424228862703e-07,
+      "loss": 0.0345,
+      "step": 1743
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5707560816531704,
+      "learning_rate": 9.828155906223267e-07,
+      "loss": 0.0679,
+      "step": 1744
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8166632332306311,
+      "learning_rate": 9.827887377604995e-07,
+      "loss": 0.2064,
+      "step": 1745
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.13638838624686567,
+      "learning_rate": 9.827618643019339e-07,
+      "loss": 0.081,
+      "step": 1746
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.10007512364566283,
+      "learning_rate": 9.82734970247777e-07,
+      "loss": 0.0394,
+      "step": 1747
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6577255730064722,
+      "learning_rate": 9.827080555991759e-07,
+      "loss": 0.4635,
+      "step": 1748
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.30082004733304346,
+      "learning_rate": 9.826811203572785e-07,
+      "loss": 0.051,
+      "step": 1749
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.970628457968785,
+      "learning_rate": 9.826541645232344e-07,
+      "loss": 0.1083,
+      "step": 1750
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.45907943999897577,
+      "learning_rate": 9.826271880981934e-07,
+      "loss": 0.2844,
+      "step": 1751
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.34774585567832594,
+      "learning_rate": 9.826001910833062e-07,
+      "loss": 0.4047,
+      "step": 1752
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5476944428879885,
+      "learning_rate": 9.825731734797246e-07,
+      "loss": 0.2125,
+      "step": 1753
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5065585280997207,
+      "learning_rate": 9.825461352886016e-07,
+      "loss": 0.3552,
+      "step": 1754
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5620406168997668,
+      "learning_rate": 9.825190765110904e-07,
+      "loss": 0.1863,
+      "step": 1755
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8806592318224842,
+      "learning_rate": 9.824919971483451e-07,
+      "loss": 0.3026,
+      "step": 1756
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.647794205861562,
+      "learning_rate": 9.824648972015218e-07,
+      "loss": 0.1532,
+      "step": 1757
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3841639771848569,
+      "learning_rate": 9.824377766717758e-07,
+      "loss": 0.1178,
+      "step": 1758
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.47163025371174877,
+      "learning_rate": 9.824106355602643e-07,
+      "loss": 0.0146,
+      "step": 1759
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.917001845279771,
+      "learning_rate": 9.823834738681454e-07,
+      "loss": 0.1631,
+      "step": 1760
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3793247641315772,
+      "learning_rate": 9.823562915965779e-07,
+      "loss": 0.265,
+      "step": 1761
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9563626167521188,
+      "learning_rate": 9.823290887467213e-07,
+      "loss": 0.3483,
+      "step": 1762
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.42631993786465006,
+      "learning_rate": 9.82301865319736e-07,
+      "loss": 0.1072,
+      "step": 1763
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4259209625898786,
+      "learning_rate": 9.82274621316784e-07,
+      "loss": 0.1468,
+      "step": 1764
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4871455260003286,
+      "learning_rate": 9.822473567390269e-07,
+      "loss": 0.4063,
+      "step": 1765
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5626342707483665,
+      "learning_rate": 9.82220071587628e-07,
+      "loss": 0.2594,
+      "step": 1766
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3353387516102091,
+      "learning_rate": 9.821927658637517e-07,
+      "loss": 0.0104,
+      "step": 1767
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7804416552276265,
+      "learning_rate": 9.821654395685626e-07,
+      "loss": 0.1929,
+      "step": 1768
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.23512946681285746,
+      "learning_rate": 9.821380927032264e-07,
+      "loss": 0.1229,
+      "step": 1769
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5555530016103668,
+      "learning_rate": 9.821107252689102e-07,
+      "loss": 0.2054,
+      "step": 1770
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5443962688924795,
+      "learning_rate": 9.820833372667812e-07,
+      "loss": 0.1442,
+      "step": 1771
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39619915615185125,
+      "learning_rate": 9.82055928698008e-07,
+      "loss": 0.2412,
+      "step": 1772
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3305805628162596,
+      "learning_rate": 9.820284995637595e-07,
+      "loss": 0.2055,
+      "step": 1773
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6046842976387699,
+      "learning_rate": 9.820010498652064e-07,
+      "loss": 0.4141,
+      "step": 1774
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4031768210777634,
+      "learning_rate": 9.819735796035197e-07,
+      "loss": 0.1269,
+      "step": 1775
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6229167909879683,
+      "learning_rate": 9.819460887798713e-07,
+      "loss": 0.1951,
+      "step": 1776
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.40434643232921913,
+      "learning_rate": 9.819185773954335e-07,
+      "loss": 0.1349,
+      "step": 1777
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.433914946149375,
+      "learning_rate": 9.818910454513808e-07,
+      "loss": 0.2466,
+      "step": 1778
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5308097070441203,
+      "learning_rate": 9.818634929488872e-07,
+      "loss": 0.1349,
+      "step": 1779
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3651917806208901,
+      "learning_rate": 9.818359198891284e-07,
+      "loss": 0.083,
+      "step": 1780
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4163734535240564,
+      "learning_rate": 9.818083262732806e-07,
+      "loss": 0.2195,
+      "step": 1781
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.16223345234529754,
+      "learning_rate": 9.81780712102521e-07,
+      "loss": 0.2932,
+      "step": 1782
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6132952793487771,
+      "learning_rate": 9.817530773780276e-07,
+      "loss": 0.1361,
+      "step": 1783
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.45207397666049487,
+      "learning_rate": 9.817254221009798e-07,
+      "loss": 0.448,
+      "step": 1784
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.34689781529461494,
+      "learning_rate": 9.81697746272557e-07,
+      "loss": 0.1557,
+      "step": 1785
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5331420713596847,
+      "learning_rate": 9.816700498939399e-07,
+      "loss": 0.0829,
+      "step": 1786
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.49071365729277694,
+      "learning_rate": 9.8164233296631e-07,
+      "loss": 0.124,
+      "step": 1787
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.596526687363129,
+      "learning_rate": 9.816145954908503e-07,
+      "loss": 0.154,
+      "step": 1788
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39534540517531486,
+      "learning_rate": 9.815868374687436e-07,
+      "loss": 0.0784,
+      "step": 1789
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.42546356363112214,
+      "learning_rate": 9.815590589011746e-07,
+      "loss": 0.1693,
+      "step": 1790
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.31141271989092173,
+      "learning_rate": 9.815312597893278e-07,
+      "loss": 0.0798,
+      "step": 1791
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4670777628172246,
+      "learning_rate": 9.815034401343896e-07,
+      "loss": 0.1561,
+      "step": 1792
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6577163324054129,
+      "learning_rate": 9.814755999375466e-07,
+      "loss": 0.2182,
+      "step": 1793
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.624311653075215,
+      "learning_rate": 9.814477391999867e-07,
+      "loss": 0.3685,
+      "step": 1794
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4001669059106104,
+      "learning_rate": 9.814198579228985e-07,
+      "loss": 0.2468,
+      "step": 1795
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6506927806465743,
+      "learning_rate": 9.81391956107471e-07,
+      "loss": 0.2173,
+      "step": 1796
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6632104392930209,
+      "learning_rate": 9.813640337548954e-07,
+      "loss": 0.213,
+      "step": 1797
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5259042393102924,
+      "learning_rate": 9.813360908663621e-07,
+      "loss": 0.1574,
+      "step": 1798
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6201969188977602,
+      "learning_rate": 9.813081274430636e-07,
+      "loss": 0.1324,
+      "step": 1799
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1144665506804912,
+      "learning_rate": 9.81280143486193e-07,
+      "loss": 0.1556,
+      "step": 1800
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4889335289184677,
+      "learning_rate": 9.81252138996944e-07,
+      "loss": 0.1724,
+      "step": 1801
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.18328905865515413,
+      "learning_rate": 9.812241139765112e-07,
+      "loss": 0.0105,
+      "step": 1802
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1410813212247752,
+      "learning_rate": 9.811960684260906e-07,
+      "loss": 0.1477,
+      "step": 1803
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.24938386629140047,
+      "learning_rate": 9.81168002346878e-07,
+      "loss": 0.0134,
+      "step": 1804
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5242348551646572,
+      "learning_rate": 9.811399157400712e-07,
+      "loss": 0.315,
+      "step": 1805
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.19186140756534406,
+      "learning_rate": 9.811118086068687e-07,
+      "loss": 0.0754,
+      "step": 1806
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3366882759389052,
+      "learning_rate": 9.810836809484689e-07,
+      "loss": 0.1253,
+      "step": 1807
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.31379195803906235,
+      "learning_rate": 9.810555327660723e-07,
+      "loss": 0.3143,
+      "step": 1808
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.798839770675499,
+      "learning_rate": 9.810273640608798e-07,
+      "loss": 0.151,
+      "step": 1809
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5005194608080867,
+      "learning_rate": 9.809991748340926e-07,
+      "loss": 0.1507,
+      "step": 1810
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.32667542686374074,
+      "learning_rate": 9.80970965086914e-07,
+      "loss": 0.1517,
+      "step": 1811
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2176304917892244,
+      "learning_rate": 9.80942734820547e-07,
+      "loss": 0.0326,
+      "step": 1812
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7991156169284748,
+      "learning_rate": 9.809144840361963e-07,
+      "loss": 0.3402,
+      "step": 1813
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7622023192574338,
+      "learning_rate": 9.808862127350668e-07,
+      "loss": 0.4426,
+      "step": 1814
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4018101359851353,
+      "learning_rate": 9.808579209183648e-07,
+      "loss": 0.088,
+      "step": 1815
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.42093073883317284,
+      "learning_rate": 9.808296085872971e-07,
+      "loss": 0.158,
+      "step": 1816
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.31664305537217885,
+      "learning_rate": 9.80801275743072e-07,
+      "loss": 0.1224,
+      "step": 1817
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.34672074617306375,
+      "learning_rate": 9.807729223868978e-07,
+      "loss": 0.1798,
+      "step": 1818
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3266763970395985,
+      "learning_rate": 9.807445485199842e-07,
+      "loss": 0.0133,
+      "step": 1819
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7941496523634355,
+      "learning_rate": 9.807161541435417e-07,
+      "loss": 0.4453,
+      "step": 1820
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5296019648078376,
+      "learning_rate": 9.80687739258782e-07,
+      "loss": 0.205,
+      "step": 1821
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6639361373691804,
+      "learning_rate": 9.806593038669167e-07,
+      "loss": 0.3553,
+      "step": 1822
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.20200816961047624,
+      "learning_rate": 9.806308479691594e-07,
+      "loss": 0.1893,
+      "step": 1823
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4058296945525068,
+      "learning_rate": 9.80602371566724e-07,
+      "loss": 0.3192,
+      "step": 1824
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.31605906656818067,
+      "learning_rate": 9.805738746608251e-07,
+      "loss": 0.1119,
+      "step": 1825
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.22778975778087518,
+      "learning_rate": 9.805453572526787e-07,
+      "loss": 0.0946,
+      "step": 1826
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4639766727675487,
+      "learning_rate": 9.805168193435014e-07,
+      "loss": 0.1883,
+      "step": 1827
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6688244519223763,
+      "learning_rate": 9.804882609345106e-07,
+      "loss": 0.0828,
+      "step": 1828
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4956690050397604,
+      "learning_rate": 9.804596820269249e-07,
+      "loss": 0.1604,
+      "step": 1829
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.441950423118919,
+      "learning_rate": 9.80431082621963e-07,
+      "loss": 0.2277,
+      "step": 1830
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4657042253417374,
+      "learning_rate": 9.804024627208456e-07,
+      "loss": 0.1976,
+      "step": 1831
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4808679457578324,
+      "learning_rate": 9.80373822324793e-07,
+      "loss": 0.0222,
+      "step": 1832
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6024860333388573,
+      "learning_rate": 9.803451614350278e-07,
+      "loss": 0.1906,
+      "step": 1833
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.37904837274505715,
+      "learning_rate": 9.803164800527723e-07,
+      "loss": 0.2039,
+      "step": 1834
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3253697459300525,
+      "learning_rate": 9.802877781792503e-07,
+      "loss": 0.1406,
+      "step": 1835
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3089051198714158,
+      "learning_rate": 9.802590558156861e-07,
+      "loss": 0.1815,
+      "step": 1836
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7715997059014499,
+      "learning_rate": 9.802303129633052e-07,
+      "loss": 0.3338,
+      "step": 1837
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9373057696770362,
+      "learning_rate": 9.802015496233336e-07,
+      "loss": 0.3327,
+      "step": 1838
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.39383632011750946,
+      "learning_rate": 9.801727657969987e-07,
+      "loss": 0.1205,
+      "step": 1839
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4442621155258366,
+      "learning_rate": 9.801439614855285e-07,
+      "loss": 0.086,
+      "step": 1840
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.36569334529732345,
+      "learning_rate": 9.801151366901514e-07,
+      "loss": 0.1447,
+      "step": 1841
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6557356390290756,
+      "learning_rate": 9.800862914120975e-07,
+      "loss": 0.3614,
+      "step": 1842
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2791723331236794,
+      "learning_rate": 9.800574256525974e-07,
+      "loss": 0.0677,
+      "step": 1843
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4177283348243584,
+      "learning_rate": 9.800285394128824e-07,
+      "loss": 0.1132,
+      "step": 1844
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5267707960046631,
+      "learning_rate": 9.799996326941849e-07,
+      "loss": 0.1599,
+      "step": 1845
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.349914681365973,
+      "learning_rate": 9.799707054977382e-07,
+      "loss": 0.3591,
+      "step": 1846
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.22452490998338034,
+      "learning_rate": 9.799417578247764e-07,
+      "loss": 0.1453,
+      "step": 1847
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6393751850329222,
+      "learning_rate": 9.799127896765344e-07,
+      "loss": 0.2642,
+      "step": 1848
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.27700682909654584,
+      "learning_rate": 9.798838010542482e-07,
+      "loss": 0.1942,
+      "step": 1849
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8556120715818918,
+      "learning_rate": 9.79854791959154e-07,
+      "loss": 0.1704,
+      "step": 1850
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.285872714361692,
+      "learning_rate": 9.798257623924899e-07,
+      "loss": 0.0514,
+      "step": 1851
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4804723649355166,
+      "learning_rate": 9.797967123554943e-07,
+      "loss": 0.2293,
+      "step": 1852
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8830097789972284,
+      "learning_rate": 9.797676418494063e-07,
+      "loss": 0.1981,
+      "step": 1853
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.39143972961572615,
+      "learning_rate": 9.797385508754664e-07,
+      "loss": 0.1841,
+      "step": 1854
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5713067759027882,
+      "learning_rate": 9.797094394349152e-07,
+      "loss": 0.4405,
+      "step": 1855
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1961522534586305,
+      "learning_rate": 9.796803075289953e-07,
+      "loss": 0.1621,
+      "step": 1856
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.39600435921384797,
+      "learning_rate": 9.79651155158949e-07,
+      "loss": 0.1969,
+      "step": 1857
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.28276205103716984,
+      "learning_rate": 9.796219823260203e-07,
+      "loss": 0.1092,
+      "step": 1858
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.18106232221884677,
+      "learning_rate": 9.795927890314536e-07,
+      "loss": 0.1051,
+      "step": 1859
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.49736151366559794,
+      "learning_rate": 9.795635752764947e-07,
+      "loss": 0.2484,
+      "step": 1860
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6512118286875864,
+      "learning_rate": 9.795343410623893e-07,
+      "loss": 0.3593,
+      "step": 1861
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6993336560272034,
+      "learning_rate": 9.795050863903851e-07,
+      "loss": 0.1588,
+      "step": 1862
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7299234256592234,
+      "learning_rate": 9.7947581126173e-07,
+      "loss": 0.43,
+      "step": 1863
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8753700912961243,
+      "learning_rate": 9.794465156776727e-07,
+      "loss": 0.2569,
+      "step": 1864
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7229558812877958,
+      "learning_rate": 9.794171996394636e-07,
+      "loss": 0.4534,
+      "step": 1865
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3767583159896383,
+      "learning_rate": 9.793878631483528e-07,
+      "loss": 0.2364,
+      "step": 1866
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0080189319702062,
+      "learning_rate": 9.79358506205592e-07,
+      "loss": 0.2474,
+      "step": 1867
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.32366932783785746,
+      "learning_rate": 9.793291288124339e-07,
+      "loss": 0.2354,
+      "step": 1868
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.41794399068829174,
+      "learning_rate": 9.792997309701314e-07,
+      "loss": 0.2055,
+      "step": 1869
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8669801849769293,
+      "learning_rate": 9.79270312679939e-07,
+      "loss": 0.3206,
+      "step": 1870
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1628062491234834,
+      "learning_rate": 9.792408739431115e-07,
+      "loss": 0.1916,
+      "step": 1871
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.22763899811533927,
+      "learning_rate": 9.79211414760905e-07,
+      "loss": 0.0115,
+      "step": 1872
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8178078855246522,
+      "learning_rate": 9.79181935134576e-07,
+      "loss": 0.2622,
+      "step": 1873
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5902380529764373,
+      "learning_rate": 9.791524350653825e-07,
+      "loss": 0.3348,
+      "step": 1874
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3034829566721241,
+      "learning_rate": 9.791229145545832e-07,
+      "loss": 0.163,
+      "step": 1875
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6596373558294532,
+      "learning_rate": 9.790933736034367e-07,
+      "loss": 0.2436,
+      "step": 1876
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2110218039827778,
+      "learning_rate": 9.790638122132042e-07,
+      "loss": 0.3134,
+      "step": 1877
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6319633247353574,
+      "learning_rate": 9.790342303851462e-07,
+      "loss": 0.6044,
+      "step": 1878
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.678909030370214,
+      "learning_rate": 9.79004628120525e-07,
+      "loss": 0.3747,
+      "step": 1879
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.31815956324995,
+      "learning_rate": 9.789750054206035e-07,
+      "loss": 0.1646,
+      "step": 1880
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8783055343424494,
+      "learning_rate": 9.789453622866453e-07,
+      "loss": 0.1787,
+      "step": 1881
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.20925826497237446,
+      "learning_rate": 9.789156987199154e-07,
+      "loss": 0.0833,
+      "step": 1882
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5741813855041612,
+      "learning_rate": 9.788860147216788e-07,
+      "loss": 0.3773,
+      "step": 1883
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9213884749639869,
+      "learning_rate": 9.788563102932021e-07,
+      "loss": 0.2624,
+      "step": 1884
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4642293176264571,
+      "learning_rate": 9.788265854357527e-07,
+      "loss": 0.2079,
+      "step": 1885
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7433205464784334,
+      "learning_rate": 9.787968401505987e-07,
+      "loss": 0.2458,
+      "step": 1886
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3769902608715658,
+      "learning_rate": 9.787670744390088e-07,
+      "loss": 0.4972,
+      "step": 1887
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2847382639762413,
+      "learning_rate": 9.787372883022531e-07,
+      "loss": 0.2299,
+      "step": 1888
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4660825408621928,
+      "learning_rate": 9.787074817416023e-07,
+      "loss": 0.233,
+      "step": 1889
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8257495471447167,
+      "learning_rate": 9.786776547583281e-07,
+      "loss": 0.2907,
+      "step": 1890
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4428450110946486,
+      "learning_rate": 9.786478073537029e-07,
+      "loss": 0.1748,
+      "step": 1891
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9232552991844433,
+      "learning_rate": 9.78617939529e-07,
+      "loss": 0.0626,
+      "step": 1892
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3262947492739359,
+      "learning_rate": 9.785880512854935e-07,
+      "loss": 0.1915,
+      "step": 1893
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.42104398880951505,
+      "learning_rate": 9.785581426244588e-07,
+      "loss": 0.0546,
+      "step": 1894
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4887469321417179,
+      "learning_rate": 9.785282135471718e-07,
+      "loss": 0.0486,
+      "step": 1895
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7994460322423196,
+      "learning_rate": 9.78498264054909e-07,
+      "loss": 0.1747,
+      "step": 1896
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.43210748121813813,
+      "learning_rate": 9.784682941489484e-07,
+      "loss": 0.4854,
+      "step": 1897
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4616134580844671,
+      "learning_rate": 9.784383038305687e-07,
+      "loss": 0.2818,
+      "step": 1898
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4322884457982743,
+      "learning_rate": 9.78408293101049e-07,
+      "loss": 0.2974,
+      "step": 1899
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.569075289211957,
+      "learning_rate": 9.7837826196167e-07,
+      "loss": 0.0971,
+      "step": 1900
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5319529424582372,
+      "learning_rate": 9.783482104137127e-07,
+      "loss": 0.1652,
+      "step": 1901
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.41210358606979236,
+      "learning_rate": 9.783181384584589e-07,
+      "loss": 0.0598,
+      "step": 1902
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.48852613679855594,
+      "learning_rate": 9.782880460971918e-07,
+      "loss": 0.2678,
+      "step": 1903
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6702462521981983,
+      "learning_rate": 9.782579333311954e-07,
+      "loss": 0.1188,
+      "step": 1904
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5641643685730272,
+      "learning_rate": 9.78227800161754e-07,
+      "loss": 0.039,
+      "step": 1905
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6205175895376341,
+      "learning_rate": 9.781976465901532e-07,
+      "loss": 0.288,
+      "step": 1906
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7442542022079933,
+      "learning_rate": 9.781674726176797e-07,
+      "loss": 0.1192,
+      "step": 1907
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4395414020640999,
+      "learning_rate": 9.781372782456204e-07,
+      "loss": 0.2025,
+      "step": 1908
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2712850025784868,
+      "learning_rate": 9.781070634752637e-07,
+      "loss": 0.2414,
+      "step": 1909
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.459579176644862,
+      "learning_rate": 9.780768283078986e-07,
+      "loss": 0.1618,
+      "step": 1910
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.37994028366968474,
+      "learning_rate": 9.780465727448149e-07,
+      "loss": 0.1511,
+      "step": 1911
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.332826870108756,
+      "learning_rate": 9.780162967873034e-07,
+      "loss": 0.1548,
+      "step": 1912
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7704599714078209,
+      "learning_rate": 9.779860004366559e-07,
+      "loss": 0.4119,
+      "step": 1913
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6648900381526484,
+      "learning_rate": 9.779556836941645e-07,
+      "loss": 0.2656,
+      "step": 1914
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4220809764577071,
+      "learning_rate": 9.77925346561123e-07,
+      "loss": 0.2177,
+      "step": 1915
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6930813390204651,
+      "learning_rate": 9.778949890388254e-07,
+      "loss": 0.1372,
+      "step": 1916
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5281738830083382,
+      "learning_rate": 9.778646111285667e-07,
+      "loss": 0.1023,
+      "step": 1917
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.14482613321771215,
+      "learning_rate": 9.778342128316432e-07,
+      "loss": 0.0347,
+      "step": 1918
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3616251398336747,
+      "learning_rate": 9.778037941493518e-07,
+      "loss": 0.1497,
+      "step": 1919
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3712443218442867,
+      "learning_rate": 9.7777335508299e-07,
+      "loss": 0.4352,
+      "step": 1920
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6481102215820237,
+      "learning_rate": 9.777428956338562e-07,
+      "loss": 0.2932,
+      "step": 1921
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6974521070601287,
+      "learning_rate": 9.777124158032502e-07,
+      "loss": 0.244,
+      "step": 1922
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8303239942182551,
+      "learning_rate": 9.776819155924724e-07,
+      "loss": 0.1638,
+      "step": 1923
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3098626885906418,
+      "learning_rate": 9.776513950028235e-07,
+      "loss": 0.2514,
+      "step": 1924
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6372159472920058,
+      "learning_rate": 9.77620854035606e-07,
+      "loss": 0.3313,
+      "step": 1925
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.41749912640256737,
+      "learning_rate": 9.775902926921228e-07,
+      "loss": 0.1376,
+      "step": 1926
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7259535137665333,
+      "learning_rate": 9.775597109736774e-07,
+      "loss": 0.1202,
+      "step": 1927
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7724715954337328,
+      "learning_rate": 9.77529108881575e-07,
+      "loss": 0.3796,
+      "step": 1928
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.36036089208488853,
+      "learning_rate": 9.77498486417121e-07,
+      "loss": 0.3772,
+      "step": 1929
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3278680317156285,
+      "learning_rate": 9.774678435816212e-07,
+      "loss": 0.3091,
+      "step": 1930
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8595227782650138,
+      "learning_rate": 9.774371803763837e-07,
+      "loss": 0.2001,
+      "step": 1931
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4584120445869711,
+      "learning_rate": 9.774064968027162e-07,
+      "loss": 0.0665,
+      "step": 1932
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3755435083092351,
+      "learning_rate": 9.77375792861928e-07,
+      "loss": 0.0687,
+      "step": 1933
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2137723759695154,
+      "learning_rate": 9.773450685553285e-07,
+      "loss": 0.1917,
+      "step": 1934
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9278592061036838,
+      "learning_rate": 9.773143238842293e-07,
+      "loss": 0.1079,
+      "step": 1935
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8340899317370496,
+      "learning_rate": 9.77283558849941e-07,
+      "loss": 0.1827,
+      "step": 1936
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3445475578282629,
+      "learning_rate": 9.77252773453777e-07,
+      "loss": 0.1261,
+      "step": 1937
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4587220861021094,
+      "learning_rate": 9.7722196769705e-07,
+      "loss": 0.2183,
+      "step": 1938
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5146792374457008,
+      "learning_rate": 9.771911415810746e-07,
+      "loss": 0.4111,
+      "step": 1939
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5514524419459433,
+      "learning_rate": 9.77160295107166e-07,
+      "loss": 0.1075,
+      "step": 1940
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2040956696373666,
+      "learning_rate": 9.771294282766399e-07,
+      "loss": 0.1792,
+      "step": 1941
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3809124391047485,
+      "learning_rate": 9.77098541090813e-07,
+      "loss": 0.155,
+      "step": 1942
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6677320796447382,
+      "learning_rate": 9.770676335510036e-07,
+      "loss": 0.0639,
+      "step": 1943
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6887584945356219,
+      "learning_rate": 9.770367056585298e-07,
+      "loss": 0.0538,
+      "step": 1944
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1499177396189217,
+      "learning_rate": 9.77005757414711e-07,
+      "loss": 0.042,
+      "step": 1945
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.508774266944597,
+      "learning_rate": 9.769747888208678e-07,
+      "loss": 0.1313,
+      "step": 1946
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3653623836731963,
+      "learning_rate": 9.769437998783214e-07,
+      "loss": 0.2212,
+      "step": 1947
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2563068904160823,
+      "learning_rate": 9.769127905883937e-07,
+      "loss": 0.0084,
+      "step": 1948
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8803497788997626,
+      "learning_rate": 9.768817609524075e-07,
+      "loss": 0.2744,
+      "step": 1949
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4595795259897732,
+      "learning_rate": 9.768507109716868e-07,
+      "loss": 0.1341,
+      "step": 1950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6557764103477656,
+      "learning_rate": 9.768196406475563e-07,
+      "loss": 0.3394,
+      "step": 1951
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6006556222801455,
+      "learning_rate": 9.76788549981341e-07,
+      "loss": 0.0615,
+      "step": 1952
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2228324990992887,
+      "learning_rate": 9.767574389743681e-07,
+      "loss": 0.1342,
+      "step": 1953
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7179146374553077,
+      "learning_rate": 9.767263076279643e-07,
+      "loss": 0.1997,
+      "step": 1954
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9693337983360613,
+      "learning_rate": 9.76695155943458e-07,
+      "loss": 0.112,
+      "step": 1955
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.667522800453616,
+      "learning_rate": 9.76663983922178e-07,
+      "loss": 0.1727,
+      "step": 1956
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5507093670357065,
+      "learning_rate": 9.766327915654541e-07,
+      "loss": 0.1278,
+      "step": 1957
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6860862337370913,
+      "learning_rate": 9.766015788746173e-07,
+      "loss": 0.3761,
+      "step": 1958
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40121908942416445,
+      "learning_rate": 9.76570345850999e-07,
+      "loss": 0.1011,
+      "step": 1959
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.411413968367731,
+      "learning_rate": 9.765390924959317e-07,
+      "loss": 0.368,
+      "step": 1960
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2786316838571697,
+      "learning_rate": 9.765078188107487e-07,
+      "loss": 0.1497,
+      "step": 1961
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.059609911889508,
+      "learning_rate": 9.764765247967843e-07,
+      "loss": 0.4543,
+      "step": 1962
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1661323197525828,
+      "learning_rate": 9.764452104553736e-07,
+      "loss": 0.2113,
+      "step": 1963
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.534199010122491,
+      "learning_rate": 9.764138757878524e-07,
+      "loss": 0.1222,
+      "step": 1964
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.24197334755360245,
+      "learning_rate": 9.763825207955577e-07,
+      "loss": 0.027,
+      "step": 1965
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.46587338170337794,
+      "learning_rate": 9.763511454798266e-07,
+      "loss": 0.478,
+      "step": 1966
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.32889061524493196,
+      "learning_rate": 9.763197498419984e-07,
+      "loss": 0.1475,
+      "step": 1967
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.42973472107561617,
+      "learning_rate": 9.76288333883412e-07,
+      "loss": 0.1379,
+      "step": 1968
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.34327729390832834,
+      "learning_rate": 9.76256897605408e-07,
+      "loss": 0.1892,
+      "step": 1969
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8733412381690852,
+      "learning_rate": 9.76225441009327e-07,
+      "loss": 0.1151,
+      "step": 1970
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6990838412039561,
+      "learning_rate": 9.761939640965118e-07,
+      "loss": 0.2961,
+      "step": 1971
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.103030258430781,
+      "learning_rate": 9.761624668683044e-07,
+      "loss": 0.184,
+      "step": 1972
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.47537482875737563,
+      "learning_rate": 9.761309493260492e-07,
+      "loss": 0.2184,
+      "step": 1973
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8109631954245425,
+      "learning_rate": 9.760994114710904e-07,
+      "loss": 0.1935,
+      "step": 1974
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7905566985070605,
+      "learning_rate": 9.76067853304774e-07,
+      "loss": 0.1877,
+      "step": 1975
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.31555085957082785,
+      "learning_rate": 9.760362748284456e-07,
+      "loss": 0.1956,
+      "step": 1976
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.145092587492478,
+      "learning_rate": 9.760046760434529e-07,
+      "loss": 0.485,
+      "step": 1977
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7701666012878458,
+      "learning_rate": 9.759730569511438e-07,
+      "loss": 0.3231,
+      "step": 1978
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7721610642616265,
+      "learning_rate": 9.759414175528671e-07,
+      "loss": 0.189,
+      "step": 1979
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0110930986742899,
+      "learning_rate": 9.75909757849973e-07,
+      "loss": 0.1744,
+      "step": 1980
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6197373928767853,
+      "learning_rate": 9.758780778438117e-07,
+      "loss": 0.0371,
+      "step": 1981
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9999683879013466,
+      "learning_rate": 9.758463775357352e-07,
+      "loss": 0.133,
+      "step": 1982
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5769867949119858,
+      "learning_rate": 9.758146569270956e-07,
+      "loss": 0.0779,
+      "step": 1983
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5880562635464869,
+      "learning_rate": 9.757829160192462e-07,
+      "loss": 0.2503,
+      "step": 1984
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3987440069337375,
+      "learning_rate": 9.757511548135411e-07,
+      "loss": 0.1259,
+      "step": 1985
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4624906612642204,
+      "learning_rate": 9.757193733113355e-07,
+      "loss": 0.141,
+      "step": 1986
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.45979987576149456,
+      "learning_rate": 9.75687571513985e-07,
+      "loss": 0.0776,
+      "step": 1987
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5147345228133103,
+      "learning_rate": 9.756557494228464e-07,
+      "loss": 0.1651,
+      "step": 1988
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6293209370364992,
+      "learning_rate": 9.756239070392776e-07,
+      "loss": 0.1009,
+      "step": 1989
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5651852931621907,
+      "learning_rate": 9.755920443646364e-07,
+      "loss": 0.3339,
+      "step": 1990
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9320924431987775,
+      "learning_rate": 9.755601614002828e-07,
+      "loss": 0.0483,
+      "step": 1991
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8842636608675487,
+      "learning_rate": 9.755282581475767e-07,
+      "loss": 0.2291,
+      "step": 1992
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.43139096736901833,
+      "learning_rate": 9.754963346078792e-07,
+      "loss": 0.1092,
+      "step": 1993
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6004594186436094,
+      "learning_rate": 9.754643907825522e-07,
+      "loss": 0.2012,
+      "step": 1994
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7253309011594199,
+      "learning_rate": 9.754324266729583e-07,
+      "loss": 0.5189,
+      "step": 1995
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.26812043890197435,
+      "learning_rate": 9.754004422804616e-07,
+      "loss": 0.1146,
+      "step": 1996
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6805844637181703,
+      "learning_rate": 9.753684376064262e-07,
+      "loss": 0.317,
+      "step": 1997
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9486966659565942,
+      "learning_rate": 9.753364126522177e-07,
+      "loss": 0.4689,
+      "step": 1998
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5718265247519756,
+      "learning_rate": 9.753043674192022e-07,
+      "loss": 0.2353,
+      "step": 1999
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8737459338518712,
+      "learning_rate": 9.75272301908747e-07,
+      "loss": 0.2111,
+      "step": 2000
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39156741362398484,
+      "learning_rate": 9.7524021612222e-07,
+      "loss": 0.201,
+      "step": 2001
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7781864891728763,
+      "learning_rate": 9.7520811006099e-07,
+      "loss": 0.0336,
+      "step": 2002
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5638928565719611,
+      "learning_rate": 9.751759837264267e-07,
+      "loss": 0.2304,
+      "step": 2003
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7209768030488014,
+      "learning_rate": 9.751438371199006e-07,
+      "loss": 0.3343,
+      "step": 2004
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.519758722621288,
+      "learning_rate": 9.751116702427833e-07,
+      "loss": 0.332,
+      "step": 2005
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7667403647507963,
+      "learning_rate": 9.750794830964472e-07,
+      "loss": 0.205,
+      "step": 2006
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4957310847385013,
+      "learning_rate": 9.750472756822652e-07,
+      "loss": 0.2274,
+      "step": 2007
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.29985426137213894,
+      "learning_rate": 9.750150480016114e-07,
+      "loss": 0.1888,
+      "step": 2008
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4795539366511488,
+      "learning_rate": 9.74982800055861e-07,
+      "loss": 0.2566,
+      "step": 2009
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6106050578516774,
+      "learning_rate": 9.749505318463894e-07,
+      "loss": 0.2971,
+      "step": 2010
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3604109131880696,
+      "learning_rate": 9.749182433745732e-07,
+      "loss": 0.1832,
+      "step": 2011
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7558736410748353,
+      "learning_rate": 9.7488593464179e-07,
+      "loss": 0.3393,
+      "step": 2012
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.515464229619306,
+      "learning_rate": 9.748536056494186e-07,
+      "loss": 0.238,
+      "step": 2013
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7572105402428722,
+      "learning_rate": 9.748212563988375e-07,
+      "loss": 0.1486,
+      "step": 2014
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6228214472009372,
+      "learning_rate": 9.74788886891427e-07,
+      "loss": 0.1377,
+      "step": 2015
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5848550671694532,
+      "learning_rate": 9.747564971285684e-07,
+      "loss": 0.2804,
+      "step": 2016
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6466063998668526,
+      "learning_rate": 9.747240871116432e-07,
+      "loss": 0.2071,
+      "step": 2017
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6349211533366779,
+      "learning_rate": 9.74691656842034e-07,
+      "loss": 0.1981,
+      "step": 2018
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3721453325045769,
+      "learning_rate": 9.746592063211246e-07,
+      "loss": 0.2082,
+      "step": 2019
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15608617273693193,
+      "learning_rate": 9.746267355502991e-07,
+      "loss": 0.0133,
+      "step": 2020
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6938914285732772,
+      "learning_rate": 9.74594244530943e-07,
+      "loss": 0.1737,
+      "step": 2021
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3990814093962295,
+      "learning_rate": 9.745617332644424e-07,
+      "loss": 0.2083,
+      "step": 2022
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3651754782509643,
+      "learning_rate": 9.745292017521842e-07,
+      "loss": 0.2143,
+      "step": 2023
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4697368479888514,
+      "learning_rate": 9.744966499955565e-07,
+      "loss": 0.2256,
+      "step": 2024
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.43175384708877673,
+      "learning_rate": 9.744640779959477e-07,
+      "loss": 0.3387,
+      "step": 2025
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.47758884710700433,
+      "learning_rate": 9.744314857547476e-07,
+      "loss": 0.1533,
+      "step": 2026
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9890250844300368,
+      "learning_rate": 9.743988732733466e-07,
+      "loss": 0.2262,
+      "step": 2027
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2470302772005145,
+      "learning_rate": 9.743662405531359e-07,
+      "loss": 0.0653,
+      "step": 2028
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0875355284708483,
+      "learning_rate": 9.74333587595508e-07,
+      "loss": 0.2291,
+      "step": 2029
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.753026424832712,
+      "learning_rate": 9.743009144018556e-07,
+      "loss": 0.3202,
+      "step": 2030
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7075134086327713,
+      "learning_rate": 9.742682209735727e-07,
+      "loss": 0.1801,
+      "step": 2031
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4536577664712018,
+      "learning_rate": 9.742355073120542e-07,
+      "loss": 0.2078,
+      "step": 2032
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8891863029578857,
+      "learning_rate": 9.742027734186955e-07,
+      "loss": 0.3736,
+      "step": 2033
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.38843643792649585,
+      "learning_rate": 9.741700192948934e-07,
+      "loss": 0.1576,
+      "step": 2034
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4305876125032848,
+      "learning_rate": 9.741372449420448e-07,
+      "loss": 0.3007,
+      "step": 2035
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8447349856615494,
+      "learning_rate": 9.741044503615484e-07,
+      "loss": 0.2363,
+      "step": 2036
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.066234023029225,
+      "learning_rate": 9.740716355548028e-07,
+      "loss": 0.3245,
+      "step": 2037
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.49150715679904,
+      "learning_rate": 9.740388005232085e-07,
+      "loss": 0.0474,
+      "step": 2038
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.41928467630095934,
+      "learning_rate": 9.74005945268166e-07,
+      "loss": 0.1455,
+      "step": 2039
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.42009305044203954,
+      "learning_rate": 9.73973069791077e-07,
+      "loss": 0.0243,
+      "step": 2040
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.39566816871536853,
+      "learning_rate": 9.73940174093344e-07,
+      "loss": 0.1708,
+      "step": 2041
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5963582248153219,
+      "learning_rate": 9.739072581763704e-07,
+      "loss": 0.04,
+      "step": 2042
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0059675591387007,
+      "learning_rate": 9.738743220415607e-07,
+      "loss": 0.2017,
+      "step": 2043
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.46461982600549445,
+      "learning_rate": 9.738413656903197e-07,
+      "loss": 0.2319,
+      "step": 2044
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8383189109340576,
+      "learning_rate": 9.738083891240534e-07,
+      "loss": 0.2545,
+      "step": 2045
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5723536110817462,
+      "learning_rate": 9.737753923441687e-07,
+      "loss": 0.2316,
+      "step": 2046
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5474228215810266,
+      "learning_rate": 9.737423753520734e-07,
+      "loss": 0.1426,
+      "step": 2047
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.48171784417932656,
+      "learning_rate": 9.737093381491761e-07,
+      "loss": 0.2324,
+      "step": 2048
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3155652055117869,
+      "learning_rate": 9.73676280736886e-07,
+      "loss": 0.2159,
+      "step": 2049
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8692302547729482,
+      "learning_rate": 9.736432031166138e-07,
+      "loss": 0.2041,
+      "step": 2050
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4674316385906084,
+      "learning_rate": 9.736101052897704e-07,
+      "loss": 0.3144,
+      "step": 2051
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7351636609704858,
+      "learning_rate": 9.735769872577677e-07,
+      "loss": 0.0644,
+      "step": 2052
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4570046401607926,
+      "learning_rate": 9.735438490220186e-07,
+      "loss": 0.022,
+      "step": 2053
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8104489250680039,
+      "learning_rate": 9.735106905839372e-07,
+      "loss": 0.2011,
+      "step": 2054
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4759229744029291,
+      "learning_rate": 9.734775119449378e-07,
+      "loss": 0.2255,
+      "step": 2055
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.594143220537935,
+      "learning_rate": 9.73444313106436e-07,
+      "loss": 0.1352,
+      "step": 2056
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6073301301142678,
+      "learning_rate": 9.73411094069848e-07,
+      "loss": 0.1248,
+      "step": 2057
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5823100815634501,
+      "learning_rate": 9.73377854836591e-07,
+      "loss": 0.2456,
+      "step": 2058
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7276842922878295,
+      "learning_rate": 9.73344595408083e-07,
+      "loss": 0.2906,
+      "step": 2059
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4718660706642817,
+      "learning_rate": 9.733113157857433e-07,
+      "loss": 0.1925,
+      "step": 2060
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3890053909921082,
+      "learning_rate": 9.732780159709912e-07,
+      "loss": 0.0667,
+      "step": 2061
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5305031062772791,
+      "learning_rate": 9.732446959652475e-07,
+      "loss": 0.1155,
+      "step": 2062
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7979789520933933,
+      "learning_rate": 9.732113557699337e-07,
+      "loss": 0.0327,
+      "step": 2063
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.34268174794172396,
+      "learning_rate": 9.731779953864723e-07,
+      "loss": 0.0952,
+      "step": 2064
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6587714619738644,
+      "learning_rate": 9.731446148162866e-07,
+      "loss": 0.1926,
+      "step": 2065
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5532658298962123,
+      "learning_rate": 9.731112140608003e-07,
+      "loss": 0.19,
+      "step": 2066
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2003756135881918,
+      "learning_rate": 9.730777931214383e-07,
+      "loss": 0.4008,
+      "step": 2067
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6556766157080656,
+      "learning_rate": 9.730443519996269e-07,
+      "loss": 0.2006,
+      "step": 2068
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.899811363077273,
+      "learning_rate": 9.730108906967923e-07,
+      "loss": 0.2034,
+      "step": 2069
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1664057054531014,
+      "learning_rate": 9.729774092143626e-07,
+      "loss": 0.1783,
+      "step": 2070
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.18792340268741953,
+      "learning_rate": 9.729439075537655e-07,
+      "loss": 0.099,
+      "step": 2071
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3933597350424269,
+      "learning_rate": 9.729103857164308e-07,
+      "loss": 0.1759,
+      "step": 2072
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8924769614688599,
+      "learning_rate": 9.728768437037882e-07,
+      "loss": 0.1931,
+      "step": 2073
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7779664685121285,
+      "learning_rate": 9.728432815172688e-07,
+      "loss": 0.1541,
+      "step": 2074
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5050616246557152,
+      "learning_rate": 9.728096991583047e-07,
+      "loss": 0.2209,
+      "step": 2075
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9552024925810386,
+      "learning_rate": 9.727760966283283e-07,
+      "loss": 0.1533,
+      "step": 2076
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2126531806060135,
+      "learning_rate": 9.727424739287731e-07,
+      "loss": 0.2265,
+      "step": 2077
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2747233731372835,
+      "learning_rate": 9.727088310610738e-07,
+      "loss": 0.1232,
+      "step": 2078
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6592036867883576,
+      "learning_rate": 9.726751680266652e-07,
+      "loss": 0.3436,
+      "step": 2079
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6027860660602337,
+      "learning_rate": 9.72641484826984e-07,
+      "loss": 0.2032,
+      "step": 2080
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7120638297149764,
+      "learning_rate": 9.72607781463467e-07,
+      "loss": 0.39,
+      "step": 2081
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7869806889949933,
+      "learning_rate": 9.725740579375516e-07,
+      "loss": 0.1947,
+      "step": 2082
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7442396837728845,
+      "learning_rate": 9.725403142506772e-07,
+      "loss": 0.0312,
+      "step": 2083
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2172920833957057,
+      "learning_rate": 9.725065504042832e-07,
+      "loss": 0.2183,
+      "step": 2084
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2403537132885519,
+      "learning_rate": 9.724727663998096e-07,
+      "loss": 0.0849,
+      "step": 2085
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4990650843509188,
+      "learning_rate": 9.724389622386982e-07,
+      "loss": 0.19,
+      "step": 2086
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4999932947105477,
+      "learning_rate": 9.724051379223908e-07,
+      "loss": 0.4414,
+      "step": 2087
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9431875913326456,
+      "learning_rate": 9.723712934523306e-07,
+      "loss": 0.2161,
+      "step": 2088
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9023665974170813,
+      "learning_rate": 9.723374288299614e-07,
+      "loss": 0.3656,
+      "step": 2089
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.21688367093736713,
+      "learning_rate": 9.72303544056728e-07,
+      "loss": 0.1726,
+      "step": 2090
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5682566962497361,
+      "learning_rate": 9.72269639134076e-07,
+      "loss": 0.2856,
+      "step": 2091
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.24899213121243702,
+      "learning_rate": 9.722357140634518e-07,
+      "loss": 0.0443,
+      "step": 2092
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8775025550605597,
+      "learning_rate": 9.722017688463026e-07,
+      "loss": 0.2971,
+      "step": 2093
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4040510763057725,
+      "learning_rate": 9.72167803484077e-07,
+      "loss": 0.1474,
+      "step": 2094
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5417357806702155,
+      "learning_rate": 9.721338179782235e-07,
+      "loss": 0.1652,
+      "step": 2095
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.38910553878469045,
+      "learning_rate": 9.720998123301922e-07,
+      "loss": 0.1546,
+      "step": 2096
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7782408360767257,
+      "learning_rate": 9.720657865414338e-07,
+      "loss": 0.1887,
+      "step": 2097
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8654013443872984,
+      "learning_rate": 9.720317406134002e-07,
+      "loss": 0.1594,
+      "step": 2098
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6049097490226051,
+      "learning_rate": 9.719976745475435e-07,
+      "loss": 0.1257,
+      "step": 2099
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2215042007525334,
+      "learning_rate": 9.719635883453174e-07,
+      "loss": 0.2492,
+      "step": 2100
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6352060577546463,
+      "learning_rate": 9.719294820081755e-07,
+      "loss": 0.1563,
+      "step": 2101
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.627771964044561,
+      "learning_rate": 9.718953555375734e-07,
+      "loss": 0.1785,
+      "step": 2102
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9944669068364482,
+      "learning_rate": 9.718612089349669e-07,
+      "loss": 0.2305,
+      "step": 2103
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5532890803846539,
+      "learning_rate": 9.718270422018124e-07,
+      "loss": 0.1413,
+      "step": 2104
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6236204924539288,
+      "learning_rate": 9.71792855339568e-07,
+      "loss": 0.0515,
+      "step": 2105
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7052950318639188,
+      "learning_rate": 9.71758648349692e-07,
+      "loss": 0.203,
+      "step": 2106
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.42809166385761,
+      "learning_rate": 9.717244212336436e-07,
+      "loss": 0.0337,
+      "step": 2107
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.41174206684281756,
+      "learning_rate": 9.716901739928831e-07,
+      "loss": 0.1834,
+      "step": 2108
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.665161076998366,
+      "learning_rate": 9.716559066288714e-07,
+      "loss": 0.2682,
+      "step": 2109
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4224887983642203,
+      "learning_rate": 9.716216191430708e-07,
+      "loss": 0.1781,
+      "step": 2110
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2939570486585365,
+      "learning_rate": 9.715873115369439e-07,
+      "loss": 0.1703,
+      "step": 2111
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6171708726170932,
+      "learning_rate": 9.71552983811954e-07,
+      "loss": 0.1676,
+      "step": 2112
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8375808528722852,
+      "learning_rate": 9.71518635969566e-07,
+      "loss": 0.1593,
+      "step": 2113
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5639128985805464,
+      "learning_rate": 9.714842680112455e-07,
+      "loss": 0.3167,
+      "step": 2114
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.550662219443129,
+      "learning_rate": 9.714498799384578e-07,
+      "loss": 0.2902,
+      "step": 2115
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5102697864545089,
+      "learning_rate": 9.714154717526708e-07,
+      "loss": 0.2088,
+      "step": 2116
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5303789548006846,
+      "learning_rate": 9.713810434553519e-07,
+      "loss": 0.1205,
+      "step": 2117
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.42365031034017475,
+      "learning_rate": 9.713465950479703e-07,
+      "loss": 0.1049,
+      "step": 2118
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.31375022282377896,
+      "learning_rate": 9.713121265319952e-07,
+      "loss": 0.2105,
+      "step": 2119
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.44697664741525117,
+      "learning_rate": 9.712776379088975e-07,
+      "loss": 0.2007,
+      "step": 2120
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.63302274232392,
+      "learning_rate": 9.712431291801482e-07,
+      "loss": 0.2584,
+      "step": 2121
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9660387719003406,
+      "learning_rate": 9.7120860034722e-07,
+      "loss": 0.1291,
+      "step": 2122
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.706858797654844,
+      "learning_rate": 9.711740514115853e-07,
+      "loss": 0.1969,
+      "step": 2123
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4814523309403148,
+      "learning_rate": 9.711394823747185e-07,
+      "loss": 0.1819,
+      "step": 2124
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.36670555841369007,
+      "learning_rate": 9.711048932380944e-07,
+      "loss": 0.1673,
+      "step": 2125
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6323553144994762,
+      "learning_rate": 9.710702840031885e-07,
+      "loss": 0.4136,
+      "step": 2126
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5761880647757363,
+      "learning_rate": 9.710356546714772e-07,
+      "loss": 0.1918,
+      "step": 2127
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.49994957241813537,
+      "learning_rate": 9.71001005244438e-07,
+      "loss": 0.1708,
+      "step": 2128
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6593229239558963,
+      "learning_rate": 9.709663357235492e-07,
+      "loss": 0.1929,
+      "step": 2129
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.608489529045219,
+      "learning_rate": 9.709316461102897e-07,
+      "loss": 0.125,
+      "step": 2130
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.468802224090361,
+      "learning_rate": 9.708969364061394e-07,
+      "loss": 0.2931,
+      "step": 2131
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.49492412495848126,
+      "learning_rate": 9.708622066125793e-07,
+      "loss": 0.2214,
+      "step": 2132
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.18891886191960253,
+      "learning_rate": 9.708274567310908e-07,
+      "loss": 0.0844,
+      "step": 2133
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7454011611673332,
+      "learning_rate": 9.707926867631566e-07,
+      "loss": 0.211,
+      "step": 2134
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6777257527568694,
+      "learning_rate": 9.707578967102598e-07,
+      "loss": 0.2128,
+      "step": 2135
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7101565817484278,
+      "learning_rate": 9.70723086573885e-07,
+      "loss": 0.0073,
+      "step": 2136
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5642559087375111,
+      "learning_rate": 9.70688256355517e-07,
+      "loss": 0.0641,
+      "step": 2137
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.44727627988655927,
+      "learning_rate": 9.706534060566418e-07,
+      "loss": 0.1266,
+      "step": 2138
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.27122243561276915,
+      "learning_rate": 9.70618535678746e-07,
+      "loss": 0.0889,
+      "step": 2139
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1558674339226116,
+      "learning_rate": 9.705836452233174e-07,
+      "loss": 0.4229,
+      "step": 2140
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.36806759305293124,
+      "learning_rate": 9.705487346918447e-07,
+      "loss": 0.3808,
+      "step": 2141
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6306939368757434,
+      "learning_rate": 9.705138040858169e-07,
+      "loss": 0.0952,
+      "step": 2142
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2239928866755898,
+      "learning_rate": 9.704788534067246e-07,
+      "loss": 0.2595,
+      "step": 2143
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6745385669592355,
+      "learning_rate": 9.704438826560584e-07,
+      "loss": 0.1144,
+      "step": 2144
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8328907271771963,
+      "learning_rate": 9.704088918353107e-07,
+      "loss": 0.0156,
+      "step": 2145
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9253891868737265,
+      "learning_rate": 9.703738809459738e-07,
+      "loss": 0.1115,
+      "step": 2146
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8604095819465852,
+      "learning_rate": 9.703388499895414e-07,
+      "loss": 0.3714,
+      "step": 2147
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5903399009294115,
+      "learning_rate": 9.703037989675086e-07,
+      "loss": 0.2533,
+      "step": 2148
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6681768886008498,
+      "learning_rate": 9.7026872788137e-07,
+      "loss": 0.1211,
+      "step": 2149
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9656416768461943,
+      "learning_rate": 9.702336367326222e-07,
+      "loss": 0.3824,
+      "step": 2150
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8419538800878464,
+      "learning_rate": 9.701985255227624e-07,
+      "loss": 0.1727,
+      "step": 2151
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6389175782612608,
+      "learning_rate": 9.701633942532879e-07,
+      "loss": 0.3651,
+      "step": 2152
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5482342468742762,
+      "learning_rate": 9.701282429256982e-07,
+      "loss": 0.175,
+      "step": 2153
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9234820789636007,
+      "learning_rate": 9.700930715414923e-07,
+      "loss": 0.1268,
+      "step": 2154
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4561882955178448,
+      "learning_rate": 9.70057880102171e-07,
+      "loss": 0.2099,
+      "step": 2155
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6178272401758675,
+      "learning_rate": 9.700226686092357e-07,
+      "loss": 0.2364,
+      "step": 2156
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3762372836218014,
+      "learning_rate": 9.699874370641885e-07,
+      "loss": 0.1661,
+      "step": 2157
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7340596125330184,
+      "learning_rate": 9.699521854685324e-07,
+      "loss": 0.4084,
+      "step": 2158
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4741231741355246,
+      "learning_rate": 9.699169138237714e-07,
+      "loss": 0.2054,
+      "step": 2159
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7000852701286417,
+      "learning_rate": 9.6988162213141e-07,
+      "loss": 0.1913,
+      "step": 2160
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8700922457417857,
+      "learning_rate": 9.698463103929541e-07,
+      "loss": 0.148,
+      "step": 2161
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.21168941718294554,
+      "learning_rate": 9.698109786099103e-07,
+      "loss": 0.0822,
+      "step": 2162
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8618181796727525,
+      "learning_rate": 9.697756267837855e-07,
+      "loss": 0.2028,
+      "step": 2163
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.33734627838567355,
+      "learning_rate": 9.69740254916088e-07,
+      "loss": 0.0987,
+      "step": 2164
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7313910951546378,
+      "learning_rate": 9.697048630083271e-07,
+      "loss": 0.1143,
+      "step": 2165
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5104355299439599,
+      "learning_rate": 9.696694510620126e-07,
+      "loss": 0.1054,
+      "step": 2166
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.26417956245786783,
+      "learning_rate": 9.696340190786551e-07,
+      "loss": 0.013,
+      "step": 2167
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1226080990347633,
+      "learning_rate": 9.695985670597662e-07,
+      "loss": 0.2777,
+      "step": 2168
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5572058221991034,
+      "learning_rate": 9.695630950068585e-07,
+      "loss": 0.0854,
+      "step": 2169
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5143647653811192,
+      "learning_rate": 9.695276029214452e-07,
+      "loss": 0.074,
+      "step": 2170
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4317050042711363,
+      "learning_rate": 9.694920908050405e-07,
+      "loss": 0.1415,
+      "step": 2171
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3311311852121749,
+      "learning_rate": 9.694565586591593e-07,
+      "loss": 0.118,
+      "step": 2172
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6783385015164943,
+      "learning_rate": 9.694210064853176e-07,
+      "loss": 0.2808,
+      "step": 2173
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3257814175320826,
+      "learning_rate": 9.693854342850322e-07,
+      "loss": 0.1957,
+      "step": 2174
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7204555437328546,
+      "learning_rate": 9.693498420598206e-07,
+      "loss": 0.0513,
+      "step": 2175
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.49848661994604865,
+      "learning_rate": 9.693142298112012e-07,
+      "loss": 0.0496,
+      "step": 2176
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4716172545068549,
+      "learning_rate": 9.692785975406933e-07,
+      "loss": 0.2037,
+      "step": 2177
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0155619549011925,
+      "learning_rate": 9.692429452498171e-07,
+      "loss": 0.1208,
+      "step": 2178
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5177398619634386,
+      "learning_rate": 9.692072729400936e-07,
+      "loss": 0.3334,
+      "step": 2179
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6641217851323812,
+      "learning_rate": 9.691715806130445e-07,
+      "loss": 0.3065,
+      "step": 2180
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.24003933327188864,
+      "learning_rate": 9.691358682701926e-07,
+      "loss": 0.1132,
+      "step": 2181
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9836596156748919,
+      "learning_rate": 9.691001359130614e-07,
+      "loss": 0.2834,
+      "step": 2182
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3591091840601756,
+      "learning_rate": 9.690643835431756e-07,
+      "loss": 0.0909,
+      "step": 2183
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7763253196946548,
+      "learning_rate": 9.690286111620601e-07,
+      "loss": 0.3036,
+      "step": 2184
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6384863997698524,
+      "learning_rate": 9.689928187712414e-07,
+      "loss": 0.3303,
+      "step": 2185
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0370332070148198,
+      "learning_rate": 9.68957006372246e-07,
+      "loss": 0.1551,
+      "step": 2186
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6683096996660153,
+      "learning_rate": 9.689211739666022e-07,
+      "loss": 0.1618,
+      "step": 2187
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6021385146080815,
+      "learning_rate": 9.688853215558384e-07,
+      "loss": 0.0951,
+      "step": 2188
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4977924246182438,
+      "learning_rate": 9.688494491414842e-07,
+      "loss": 0.1774,
+      "step": 2189
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6280800769891541,
+      "learning_rate": 9.6881355672507e-07,
+      "loss": 0.3283,
+      "step": 2190
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.49895923109815055,
+      "learning_rate": 9.687776443081268e-07,
+      "loss": 0.2626,
+      "step": 2191
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5822084227475768,
+      "learning_rate": 9.687417118921872e-07,
+      "loss": 0.2192,
+      "step": 2192
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40429637307764493,
+      "learning_rate": 9.68705759478784e-07,
+      "loss": 0.0835,
+      "step": 2193
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3349807403298228,
+      "learning_rate": 9.686697870694507e-07,
+      "loss": 0.0688,
+      "step": 2194
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7251416008599905,
+      "learning_rate": 9.68633794665722e-07,
+      "loss": 0.1115,
+      "step": 2195
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6856654549230466,
+      "learning_rate": 9.685977822691338e-07,
+      "loss": 0.0439,
+      "step": 2196
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4043057449881499,
+      "learning_rate": 9.685617498812221e-07,
+      "loss": 0.1153,
+      "step": 2197
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.24902936667567233,
+      "learning_rate": 9.68525697503524e-07,
+      "loss": 0.0159,
+      "step": 2198
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3618034565837879,
+      "learning_rate": 9.684896251375783e-07,
+      "loss": 0.0098,
+      "step": 2199
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.838832455790905,
+      "learning_rate": 9.684535327849231e-07,
+      "loss": 0.1777,
+      "step": 2200
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0008950477903942,
+      "learning_rate": 9.684174204470985e-07,
+      "loss": 0.44,
+      "step": 2201
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.17549592469279013,
+      "learning_rate": 9.68381288125645e-07,
+      "loss": 0.0694,
+      "step": 2202
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5686764235835852,
+      "learning_rate": 9.683451358221044e-07,
+      "loss": 0.1453,
+      "step": 2203
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5265235396622149,
+      "learning_rate": 9.683089635380185e-07,
+      "loss": 0.167,
+      "step": 2204
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2578997011400992,
+      "learning_rate": 9.682727712749311e-07,
+      "loss": 0.0425,
+      "step": 2205
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.786053876989963,
+      "learning_rate": 9.68236559034386e-07,
+      "loss": 0.1647,
+      "step": 2206
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4285561195686105,
+      "learning_rate": 9.682003268179276e-07,
+      "loss": 0.2602,
+      "step": 2207
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.05317384824930782,
+      "learning_rate": 9.681640746271026e-07,
+      "loss": 0.0014,
+      "step": 2208
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3058819496896352,
+      "learning_rate": 9.681278024634568e-07,
+      "loss": 0.0511,
+      "step": 2209
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.48222105925547604,
+      "learning_rate": 9.680915103285376e-07,
+      "loss": 0.2603,
+      "step": 2210
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.35396914352780745,
+      "learning_rate": 9.68055198223894e-07,
+      "loss": 0.1367,
+      "step": 2211
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8654080235768975,
+      "learning_rate": 9.680188661510746e-07,
+      "loss": 0.3336,
+      "step": 2212
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5383644093618157,
+      "learning_rate": 9.679825141116294e-07,
+      "loss": 0.2385,
+      "step": 2213
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5615487289719443,
+      "learning_rate": 9.679461421071096e-07,
+      "loss": 0.0606,
+      "step": 2214
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8546193412942816,
+      "learning_rate": 9.679097501390666e-07,
+      "loss": 0.084,
+      "step": 2215
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4234781567013046,
+      "learning_rate": 9.67873338209053e-07,
+      "loss": 0.1631,
+      "step": 2216
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5009648123432349,
+      "learning_rate": 9.678369063186222e-07,
+      "loss": 0.3513,
+      "step": 2217
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.46509501902431477,
+      "learning_rate": 9.678004544693285e-07,
+      "loss": 0.1144,
+      "step": 2218
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41297550452765464,
+      "learning_rate": 9.677639826627271e-07,
+      "loss": 0.1154,
+      "step": 2219
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3134990194762429,
+      "learning_rate": 9.67727490900374e-07,
+      "loss": 0.2109,
+      "step": 2220
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6121985257256408,
+      "learning_rate": 9.676909791838257e-07,
+      "loss": 0.2242,
+      "step": 2221
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.245129433818458,
+      "learning_rate": 9.676544475146402e-07,
+      "loss": 0.439,
+      "step": 2222
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6649483068722379,
+      "learning_rate": 9.676178958943756e-07,
+      "loss": 0.1686,
+      "step": 2223
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.35886327928050354,
+      "learning_rate": 9.675813243245918e-07,
+      "loss": 0.131,
+      "step": 2224
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4399401243529808,
+      "learning_rate": 9.675447328068488e-07,
+      "loss": 0.128,
+      "step": 2225
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0505089769550253,
+      "learning_rate": 9.675081213427074e-07,
+      "loss": 0.2093,
+      "step": 2226
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0391963369372383,
+      "learning_rate": 9.674714899337298e-07,
+      "loss": 0.1625,
+      "step": 2227
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8679782758021442,
+      "learning_rate": 9.674348385814786e-07,
+      "loss": 0.1818,
+      "step": 2228
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.34390251896014173,
+      "learning_rate": 9.673981672875178e-07,
+      "loss": 0.1262,
+      "step": 2229
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.550592707040225,
+      "learning_rate": 9.673614760534114e-07,
+      "loss": 0.0191,
+      "step": 2230
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4974107876985201,
+      "learning_rate": 9.673247648807248e-07,
+      "loss": 0.2038,
+      "step": 2231
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5077694190800893,
+      "learning_rate": 9.672880337710246e-07,
+      "loss": 0.2585,
+      "step": 2232
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.37577197112213945,
+      "learning_rate": 9.672512827258773e-07,
+      "loss": 0.1661,
+      "step": 2233
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5223572975879749,
+      "learning_rate": 9.67214511746851e-07,
+      "loss": 0.1267,
+      "step": 2234
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.31077655176892444,
+      "learning_rate": 9.671777208355145e-07,
+      "loss": 0.0997,
+      "step": 2235
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4371748183684701,
+      "learning_rate": 9.671409099934372e-07,
+      "loss": 0.1151,
+      "step": 2236
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.28810427124386667,
+      "learning_rate": 9.671040792221896e-07,
+      "loss": 0.1184,
+      "step": 2237
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6674952069968723,
+      "learning_rate": 9.67067228523343e-07,
+      "loss": 0.2114,
+      "step": 2238
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6422148799994809,
+      "learning_rate": 9.670303578984696e-07,
+      "loss": 0.3188,
+      "step": 2239
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6988796102022719,
+      "learning_rate": 9.669934673491423e-07,
+      "loss": 0.3983,
+      "step": 2240
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3211078676716037,
+      "learning_rate": 9.669565568769347e-07,
+      "loss": 0.1478,
+      "step": 2241
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.882084442419305,
+      "learning_rate": 9.669196264834219e-07,
+      "loss": 0.3688,
+      "step": 2242
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3245513035662919,
+      "learning_rate": 9.668826761701791e-07,
+      "loss": 0.1576,
+      "step": 2243
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8463906820598696,
+      "learning_rate": 9.668457059387826e-07,
+      "loss": 0.1455,
+      "step": 2244
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7189558460863216,
+      "learning_rate": 9.6680871579081e-07,
+      "loss": 0.1305,
+      "step": 2245
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5163374702417902,
+      "learning_rate": 9.667717057278393e-07,
+      "loss": 0.2395,
+      "step": 2246
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7378784092646128,
+      "learning_rate": 9.667346757514492e-07,
+      "loss": 0.1645,
+      "step": 2247
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5902167984125476,
+      "learning_rate": 9.666976258632195e-07,
+      "loss": 0.2945,
+      "step": 2248
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6015192434654577,
+      "learning_rate": 9.66660556064731e-07,
+      "loss": 0.2368,
+      "step": 2249
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5495202520755843,
+      "learning_rate": 9.66623466357565e-07,
+      "loss": 0.1012,
+      "step": 2250
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5985745751093206,
+      "learning_rate": 9.66586356743304e-07,
+      "loss": 0.0957,
+      "step": 2251
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5664242279788742,
+      "learning_rate": 9.665492272235309e-07,
+      "loss": 0.4868,
+      "step": 2252
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3543322807729264,
+      "learning_rate": 9.665120777998302e-07,
+      "loss": 0.2164,
+      "step": 2253
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.012639129934071,
+      "learning_rate": 9.664749084737862e-07,
+      "loss": 0.1544,
+      "step": 2254
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.81779050790222,
+      "learning_rate": 9.664377192469848e-07,
+      "loss": 0.3724,
+      "step": 2255
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8172736485518265,
+      "learning_rate": 9.664005101210128e-07,
+      "loss": 0.0195,
+      "step": 2256
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2967269382121912,
+      "learning_rate": 9.663632810974574e-07,
+      "loss": 0.148,
+      "step": 2257
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.45317727513112177,
+      "learning_rate": 9.66326032177907e-07,
+      "loss": 0.1707,
+      "step": 2258
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7766974709362029,
+      "learning_rate": 9.662887633639505e-07,
+      "loss": 0.266,
+      "step": 2259
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4984038311157407,
+      "learning_rate": 9.66251474657178e-07,
+      "loss": 0.0535,
+      "step": 2260
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1891626931876308,
+      "learning_rate": 9.662141660591804e-07,
+      "loss": 0.0147,
+      "step": 2261
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3164793955238129,
+      "learning_rate": 9.661768375715491e-07,
+      "loss": 0.2204,
+      "step": 2262
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.49853749721826085,
+      "learning_rate": 9.66139489195877e-07,
+      "loss": 0.1248,
+      "step": 2263
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.976423822686046,
+      "learning_rate": 9.661021209337573e-07,
+      "loss": 0.1819,
+      "step": 2264
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.32536603027870525,
+      "learning_rate": 9.660647327867838e-07,
+      "loss": 0.213,
+      "step": 2265
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.41416220367858053,
+      "learning_rate": 9.660273247565522e-07,
+      "loss": 0.1143,
+      "step": 2266
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7376481330398571,
+      "learning_rate": 9.659898968446578e-07,
+      "loss": 0.1566,
+      "step": 2267
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3830419810598921,
+      "learning_rate": 9.659524490526979e-07,
+      "loss": 0.0325,
+      "step": 2268
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9413732448889611,
+      "learning_rate": 9.659149813822697e-07,
+      "loss": 0.3643,
+      "step": 2269
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5660978306764866,
+      "learning_rate": 9.658774938349718e-07,
+      "loss": 0.4425,
+      "step": 2270
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6829693757465235,
+      "learning_rate": 9.658399864124035e-07,
+      "loss": 0.2359,
+      "step": 2271
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4514127261456764,
+      "learning_rate": 9.658024591161651e-07,
+      "loss": 0.1984,
+      "step": 2272
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4498924244285713,
+      "learning_rate": 9.657649119478572e-07,
+      "loss": 0.0579,
+      "step": 2273
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7614178656215915,
+      "learning_rate": 9.65727344909082e-07,
+      "loss": 0.2248,
+      "step": 2274
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6076176180082014,
+      "learning_rate": 9.65689758001442e-07,
+      "loss": 0.121,
+      "step": 2275
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.28470817960489736,
+      "learning_rate": 9.656521512265405e-07,
+      "loss": 0.266,
+      "step": 2276
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3198352442511579,
+      "learning_rate": 9.656145245859823e-07,
+      "loss": 0.1984,
+      "step": 2277
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38328997388644065,
+      "learning_rate": 9.655768780813727e-07,
+      "loss": 0.2533,
+      "step": 2278
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0348423197625813,
+      "learning_rate": 9.655392117143172e-07,
+      "loss": 0.2776,
+      "step": 2279
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8350534328849376,
+      "learning_rate": 9.655015254864233e-07,
+      "loss": 0.0954,
+      "step": 2280
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6564781601015938,
+      "learning_rate": 9.654638193992986e-07,
+      "loss": 0.3759,
+      "step": 2281
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4569331204775347,
+      "learning_rate": 9.654260934545518e-07,
+      "loss": 0.0432,
+      "step": 2282
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8782757935857414,
+      "learning_rate": 9.653883476537918e-07,
+      "loss": 0.0517,
+      "step": 2283
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.27739462172948237,
+      "learning_rate": 9.653505819986297e-07,
+      "loss": 0.1759,
+      "step": 2284
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9304604462430499,
+      "learning_rate": 9.653127964906762e-07,
+      "loss": 0.0239,
+      "step": 2285
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6302011693903139,
+      "learning_rate": 9.652749911315433e-07,
+      "loss": 0.1394,
+      "step": 2286
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7351225115861103,
+      "learning_rate": 9.65237165922844e-07,
+      "loss": 0.3691,
+      "step": 2287
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7122696323469048,
+      "learning_rate": 9.65199320866192e-07,
+      "loss": 0.2106,
+      "step": 2288
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2624750293746841,
+      "learning_rate": 9.651614559632017e-07,
+      "loss": 0.0104,
+      "step": 2289
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6971316836718546,
+      "learning_rate": 9.651235712154887e-07,
+      "loss": 0.331,
+      "step": 2290
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.552145582263909,
+      "learning_rate": 9.650856666246692e-07,
+      "loss": 0.2856,
+      "step": 2291
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8679718247585241,
+      "learning_rate": 9.650477421923602e-07,
+      "loss": 0.1551,
+      "step": 2292
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.44643504149248464,
+      "learning_rate": 9.650097979201795e-07,
+      "loss": 0.0522,
+      "step": 2293
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8235716914964851,
+      "learning_rate": 9.649718338097462e-07,
+      "loss": 0.1463,
+      "step": 2294
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0343432708886806,
+      "learning_rate": 9.649338498626795e-07,
+      "loss": 0.3718,
+      "step": 2295
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7726073379068243,
+      "learning_rate": 9.648958460806003e-07,
+      "loss": 0.2415,
+      "step": 2296
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9999413965119969,
+      "learning_rate": 9.648578224651299e-07,
+      "loss": 0.2055,
+      "step": 2297
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5979325658254782,
+      "learning_rate": 9.6481977901789e-07,
+      "loss": 0.306,
+      "step": 2298
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3014136895704547,
+      "learning_rate": 9.647817157405043e-07,
+      "loss": 0.1481,
+      "step": 2299
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4454559117522062,
+      "learning_rate": 9.64743632634596e-07,
+      "loss": 0.2429,
+      "step": 2300
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6919515419081649,
+      "learning_rate": 9.647055297017902e-07,
+      "loss": 0.0867,
+      "step": 2301
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5852256753102256,
+      "learning_rate": 9.646674069437122e-07,
+      "loss": 0.2837,
+      "step": 2302
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8964603662872278,
+      "learning_rate": 9.646292643619887e-07,
+      "loss": 0.3714,
+      "step": 2303
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8299760482054624,
+      "learning_rate": 9.645911019582465e-07,
+      "loss": 0.3254,
+      "step": 2304
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7201842646276856,
+      "learning_rate": 9.645529197341142e-07,
+      "loss": 0.2404,
+      "step": 2305
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4962911567973602,
+      "learning_rate": 9.645147176912203e-07,
+      "loss": 0.0751,
+      "step": 2306
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9255300170601806,
+      "learning_rate": 9.64476495831195e-07,
+      "loss": 0.395,
+      "step": 2307
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5418588930887981,
+      "learning_rate": 9.644382541556684e-07,
+      "loss": 0.119,
+      "step": 2308
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8536579121334795,
+      "learning_rate": 9.643999926662723e-07,
+      "loss": 0.1972,
+      "step": 2309
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9897793684100308,
+      "learning_rate": 9.643617113646392e-07,
+      "loss": 0.2981,
+      "step": 2310
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6466825868896282,
+      "learning_rate": 9.643234102524017e-07,
+      "loss": 0.1807,
+      "step": 2311
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33705126163694304,
+      "learning_rate": 9.642850893311942e-07,
+      "loss": 0.2265,
+      "step": 2312
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5435083251237673,
+      "learning_rate": 9.642467486026516e-07,
+      "loss": 0.287,
+      "step": 2313
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3867512644692106,
+      "learning_rate": 9.642083880684093e-07,
+      "loss": 0.1887,
+      "step": 2314
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0036525121321793,
+      "learning_rate": 9.64170007730104e-07,
+      "loss": 0.2691,
+      "step": 2315
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5478909722228598,
+      "learning_rate": 9.641316075893731e-07,
+      "loss": 0.1671,
+      "step": 2316
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9728080261290837,
+      "learning_rate": 9.640931876478546e-07,
+      "loss": 0.0473,
+      "step": 2317
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.47594745891612705,
+      "learning_rate": 9.64054747907188e-07,
+      "loss": 0.3223,
+      "step": 2318
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3104477542687964,
+      "learning_rate": 9.64016288369013e-07,
+      "loss": 0.1943,
+      "step": 2319
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1109103011879944,
+      "learning_rate": 9.639778090349705e-07,
+      "loss": 0.3341,
+      "step": 2320
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7667004818078823,
+      "learning_rate": 9.639393099067015e-07,
+      "loss": 0.2947,
+      "step": 2321
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33847381927963693,
+      "learning_rate": 9.639007909858493e-07,
+      "loss": 0.3664,
+      "step": 2322
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4719779746135633,
+      "learning_rate": 9.638622522740565e-07,
+      "loss": 0.2952,
+      "step": 2323
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.32241616455743904,
+      "learning_rate": 9.638236937729678e-07,
+      "loss": 0.1506,
+      "step": 2324
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39757686248786983,
+      "learning_rate": 9.637851154842279e-07,
+      "loss": 0.1362,
+      "step": 2325
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.04706264417641275,
+      "learning_rate": 9.637465174094825e-07,
+      "loss": 0.002,
+      "step": 2326
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.48218257664301306,
+      "learning_rate": 9.637078995503784e-07,
+      "loss": 0.3685,
+      "step": 2327
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6935869139609587,
+      "learning_rate": 9.636692619085633e-07,
+      "loss": 0.2605,
+      "step": 2328
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0172783780874317,
+      "learning_rate": 9.636306044856853e-07,
+      "loss": 0.2496,
+      "step": 2329
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7990364493575447,
+      "learning_rate": 9.635919272833937e-07,
+      "loss": 0.2556,
+      "step": 2330
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6195611723475155,
+      "learning_rate": 9.635532303033385e-07,
+      "loss": 0.2784,
+      "step": 2331
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.969221530484212,
+      "learning_rate": 9.635145135471708e-07,
+      "loss": 0.1225,
+      "step": 2332
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4353391986370538,
+      "learning_rate": 9.63475777016542e-07,
+      "loss": 0.2233,
+      "step": 2333
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4104028272530483,
+      "learning_rate": 9.634370207131047e-07,
+      "loss": 0.1183,
+      "step": 2334
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8507423609935023,
+      "learning_rate": 9.633982446385129e-07,
+      "loss": 0.1365,
+      "step": 2335
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4205435335862078,
+      "learning_rate": 9.633594487944202e-07,
+      "loss": 0.0591,
+      "step": 2336
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5029415640201254,
+      "learning_rate": 9.63320633182482e-07,
+      "loss": 0.1185,
+      "step": 2337
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6350187922324667,
+      "learning_rate": 9.632817978043542e-07,
+      "loss": 0.3711,
+      "step": 2338
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6837763626589852,
+      "learning_rate": 9.632429426616935e-07,
+      "loss": 0.2258,
+      "step": 2339
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1247648723040353,
+      "learning_rate": 9.632040677561577e-07,
+      "loss": 0.1972,
+      "step": 2340
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5807963694314374,
+      "learning_rate": 9.63165173089405e-07,
+      "loss": 0.2031,
+      "step": 2341
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6817570329315167,
+      "learning_rate": 9.631262586630952e-07,
+      "loss": 0.1247,
+      "step": 2342
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6135669974590318,
+      "learning_rate": 9.630873244788882e-07,
+      "loss": 0.341,
+      "step": 2343
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6181313290501207,
+      "learning_rate": 9.63048370538445e-07,
+      "loss": 0.2672,
+      "step": 2344
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4007863289709619,
+      "learning_rate": 9.630093968434275e-07,
+      "loss": 0.1462,
+      "step": 2345
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0453733218873578,
+      "learning_rate": 9.629704033954981e-07,
+      "loss": 0.0337,
+      "step": 2346
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4082676620728667,
+      "learning_rate": 9.62931390196321e-07,
+      "loss": 0.115,
+      "step": 2347
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6161421143518172,
+      "learning_rate": 9.628923572475598e-07,
+      "loss": 0.3096,
+      "step": 2348
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2977332789299097,
+      "learning_rate": 9.628533045508803e-07,
+      "loss": 0.0868,
+      "step": 2349
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8224576365366434,
+      "learning_rate": 9.628142321079485e-07,
+      "loss": 0.0306,
+      "step": 2350
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6319045401424148,
+      "learning_rate": 9.62775139920431e-07,
+      "loss": 0.2043,
+      "step": 2351
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6003548495530138,
+      "learning_rate": 9.627360279899956e-07,
+      "loss": 0.06,
+      "step": 2352
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.985280074315826,
+      "learning_rate": 9.626968963183113e-07,
+      "loss": 0.1847,
+      "step": 2353
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2260737403352769,
+      "learning_rate": 9.626577449070473e-07,
+      "loss": 0.0967,
+      "step": 2354
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33288818264685915,
+      "learning_rate": 9.626185737578737e-07,
+      "loss": 0.1423,
+      "step": 2355
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.076048043320009,
+      "learning_rate": 9.625793828724618e-07,
+      "loss": 0.2161,
+      "step": 2356
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.28497596382596746,
+      "learning_rate": 9.625401722524835e-07,
+      "loss": 0.0466,
+      "step": 2357
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8612805387264523,
+      "learning_rate": 9.625009418996115e-07,
+      "loss": 0.4057,
+      "step": 2358
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.896418765709942,
+      "learning_rate": 9.624616918155197e-07,
+      "loss": 0.1871,
+      "step": 2359
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1315802655313565,
+      "learning_rate": 9.624224220018826e-07,
+      "loss": 0.0523,
+      "step": 2360
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3753553941282118,
+      "learning_rate": 9.623831324603752e-07,
+      "loss": 0.1157,
+      "step": 2361
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.36621201404913223,
+      "learning_rate": 9.623438231926739e-07,
+      "loss": 0.1089,
+      "step": 2362
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5573117921494466,
+      "learning_rate": 9.623044942004557e-07,
+      "loss": 0.1621,
+      "step": 2363
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.78492953971227,
+      "learning_rate": 9.622651454853985e-07,
+      "loss": 0.3114,
+      "step": 2364
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3565610982750958,
+      "learning_rate": 9.62225777049181e-07,
+      "loss": 0.2818,
+      "step": 2365
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3618723782921361,
+      "learning_rate": 9.621863888934824e-07,
+      "loss": 0.1987,
+      "step": 2366
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5948519558003095,
+      "learning_rate": 9.621469810199833e-07,
+      "loss": 0.3976,
+      "step": 2367
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.657854215589905,
+      "learning_rate": 9.621075534303654e-07,
+      "loss": 0.5349,
+      "step": 2368
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1712196716947927,
+      "learning_rate": 9.620681061263098e-07,
+      "loss": 0.3451,
+      "step": 2369
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.16272161629774692,
+      "learning_rate": 9.620286391095003e-07,
+      "loss": 0.1046,
+      "step": 2370
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5854611664085622,
+      "learning_rate": 9.619891523816202e-07,
+      "loss": 0.4168,
+      "step": 2371
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3570465267301881,
+      "learning_rate": 9.61949645944354e-07,
+      "loss": 0.0927,
+      "step": 2372
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5179421249236883,
+      "learning_rate": 9.619101197993874e-07,
+      "loss": 0.2437,
+      "step": 2373
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8658125477907648,
+      "learning_rate": 9.618705739484064e-07,
+      "loss": 0.4816,
+      "step": 2374
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6956739182159973,
+      "learning_rate": 9.618310083930983e-07,
+      "loss": 0.1711,
+      "step": 2375
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7735803198214356,
+      "learning_rate": 9.617914231351508e-07,
+      "loss": 0.2274,
+      "step": 2376
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.775590982406566,
+      "learning_rate": 9.617518181762531e-07,
+      "loss": 0.4208,
+      "step": 2377
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39703609197852757,
+      "learning_rate": 9.617121935180946e-07,
+      "loss": 0.1647,
+      "step": 2378
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3233245062122669,
+      "learning_rate": 9.616725491623657e-07,
+      "loss": 0.1064,
+      "step": 2379
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5076525687307184,
+      "learning_rate": 9.616328851107578e-07,
+      "loss": 0.2319,
+      "step": 2380
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.42553805338670764,
+      "learning_rate": 9.61593201364963e-07,
+      "loss": 0.1853,
+      "step": 2381
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9205030596838517,
+      "learning_rate": 9.615534979266744e-07,
+      "loss": 0.4135,
+      "step": 2382
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5379667422988014,
+      "learning_rate": 9.615137747975857e-07,
+      "loss": 0.3133,
+      "step": 2383
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5742670032412414,
+      "learning_rate": 9.614740319793915e-07,
+      "loss": 0.2185,
+      "step": 2384
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2402152679106932,
+      "learning_rate": 9.614342694737876e-07,
+      "loss": 0.2124,
+      "step": 2385
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7856792205453577,
+      "learning_rate": 9.6139448728247e-07,
+      "loss": 0.3608,
+      "step": 2386
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4222629909090372,
+      "learning_rate": 9.613546854071362e-07,
+      "loss": 0.1599,
+      "step": 2387
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8568168659751535,
+      "learning_rate": 9.613148638494839e-07,
+      "loss": 0.0264,
+      "step": 2388
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.248520687448224,
+      "learning_rate": 9.612750226112122e-07,
+      "loss": 0.25,
+      "step": 2389
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1878111336995245,
+      "learning_rate": 9.612351616940209e-07,
+      "loss": 0.2917,
+      "step": 2390
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.40762221093022905,
+      "learning_rate": 9.611952810996102e-07,
+      "loss": 0.287,
+      "step": 2391
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3397480071058737,
+      "learning_rate": 9.611553808296818e-07,
+      "loss": 0.1617,
+      "step": 2392
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.689809580736861,
+      "learning_rate": 9.611154608859378e-07,
+      "loss": 0.1587,
+      "step": 2393
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5182216233375155,
+      "learning_rate": 9.610755212700814e-07,
+      "loss": 0.0876,
+      "step": 2394
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.223606523800506,
+      "learning_rate": 9.610355619838162e-07,
+      "loss": 0.0617,
+      "step": 2395
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.10688526188776,
+      "learning_rate": 9.609955830288471e-07,
+      "loss": 0.0832,
+      "step": 2396
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5905240026624127,
+      "learning_rate": 9.6095558440688e-07,
+      "loss": 0.2074,
+      "step": 2397
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8937538166310866,
+      "learning_rate": 9.609155661196209e-07,
+      "loss": 0.2132,
+      "step": 2398
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6143162927375626,
+      "learning_rate": 9.60875528168777e-07,
+      "loss": 0.3599,
+      "step": 2399
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.433598283328746,
+      "learning_rate": 9.608354705560567e-07,
+      "loss": 0.1139,
+      "step": 2400
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.811439499470575,
+      "learning_rate": 9.60795393283169e-07,
+      "loss": 0.1481,
+      "step": 2401
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7920695747098897,
+      "learning_rate": 9.607552963518232e-07,
+      "loss": 0.2026,
+      "step": 2402
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9343304919547966,
+      "learning_rate": 9.607151797637305e-07,
+      "loss": 0.2416,
+      "step": 2403
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.44922222916730686,
+      "learning_rate": 9.606750435206019e-07,
+      "loss": 0.0723,
+      "step": 2404
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9007662843357239,
+      "learning_rate": 9.6063488762415e-07,
+      "loss": 0.2407,
+      "step": 2405
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2592833143028035,
+      "learning_rate": 9.605947120760877e-07,
+      "loss": 0.2817,
+      "step": 2406
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8598279190527072,
+      "learning_rate": 9.605545168781291e-07,
+      "loss": 0.0815,
+      "step": 2407
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4779322437716764,
+      "learning_rate": 9.60514302031989e-07,
+      "loss": 0.1776,
+      "step": 2408
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5904274798959935,
+      "learning_rate": 9.60474067539383e-07,
+      "loss": 0.345,
+      "step": 2409
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7881322561250321,
+      "learning_rate": 9.604338134020278e-07,
+      "loss": 0.2772,
+      "step": 2410
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6838207405531467,
+      "learning_rate": 9.603935396216404e-07,
+      "loss": 0.3399,
+      "step": 2411
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5401657859931392,
+      "learning_rate": 9.60353246199939e-07,
+      "loss": 0.2189,
+      "step": 2412
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3508163251196253,
+      "learning_rate": 9.603129331386425e-07,
+      "loss": 0.1616,
+      "step": 2413
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6713099313881182,
+      "learning_rate": 9.602726004394713e-07,
+      "loss": 0.2643,
+      "step": 2414
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1978295176181208,
+      "learning_rate": 9.602322481041457e-07,
+      "loss": 0.1534,
+      "step": 2415
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8870010417763626,
+      "learning_rate": 9.60191876134387e-07,
+      "loss": 0.088,
+      "step": 2416
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37156038596610425,
+      "learning_rate": 9.601514845319178e-07,
+      "loss": 0.0091,
+      "step": 2417
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5673089570323873,
+      "learning_rate": 9.601110732984611e-07,
+      "loss": 0.0818,
+      "step": 2418
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8794165913808688,
+      "learning_rate": 9.600706424357414e-07,
+      "loss": 0.1543,
+      "step": 2419
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8860515292708557,
+      "learning_rate": 9.600301919454832e-07,
+      "loss": 0.1811,
+      "step": 2420
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.41049406078086303,
+      "learning_rate": 9.599897218294121e-07,
+      "loss": 0.2829,
+      "step": 2421
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4387965860439652,
+      "learning_rate": 9.599492320892548e-07,
+      "loss": 0.0282,
+      "step": 2422
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.08653651633786842,
+      "learning_rate": 9.599087227267386e-07,
+      "loss": 0.0023,
+      "step": 2423
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5448737842767393,
+      "learning_rate": 9.598681937435918e-07,
+      "loss": 0.3058,
+      "step": 2424
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7449450919035487,
+      "learning_rate": 9.598276451415436e-07,
+      "loss": 0.1391,
+      "step": 2425
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5389733243028549,
+      "learning_rate": 9.597870769223234e-07,
+      "loss": 0.202,
+      "step": 2426
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9581193222381307,
+      "learning_rate": 9.597464890876623e-07,
+      "loss": 0.0766,
+      "step": 2427
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7255173443642816,
+      "learning_rate": 9.597058816392917e-07,
+      "loss": 0.2205,
+      "step": 2428
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5465860010520189,
+      "learning_rate": 9.596652545789441e-07,
+      "loss": 0.1827,
+      "step": 2429
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.44458216987835736,
+      "learning_rate": 9.59624607908353e-07,
+      "loss": 0.0578,
+      "step": 2430
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9803309693892769,
+      "learning_rate": 9.595839416292516e-07,
+      "loss": 0.1905,
+      "step": 2431
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4503185098868732,
+      "learning_rate": 9.595432557433758e-07,
+      "loss": 0.1836,
+      "step": 2432
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4490701976502874,
+      "learning_rate": 9.595025502524607e-07,
+      "loss": 0.3053,
+      "step": 2433
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5301201782137251,
+      "learning_rate": 9.594618251582433e-07,
+      "loss": 0.1949,
+      "step": 2434
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4630095575443994,
+      "learning_rate": 9.594210804624608e-07,
+      "loss": 0.1619,
+      "step": 2435
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5665595307354613,
+      "learning_rate": 9.593803161668511e-07,
+      "loss": 0.1135,
+      "step": 2436
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4015014708988468,
+      "learning_rate": 9.59339532273154e-07,
+      "loss": 0.1545,
+      "step": 2437
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7615373792978193,
+      "learning_rate": 9.59298728783109e-07,
+      "loss": 0.2178,
+      "step": 2438
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3709279526175329,
+      "learning_rate": 9.59257905698457e-07,
+      "loss": 0.1777,
+      "step": 2439
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16915273068108697,
+      "learning_rate": 9.592170630209393e-07,
+      "loss": 0.0898,
+      "step": 2440
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4554536599748477,
+      "learning_rate": 9.591762007522986e-07,
+      "loss": 0.2354,
+      "step": 2441
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6277257649198804,
+      "learning_rate": 9.591353188942782e-07,
+      "loss": 0.2001,
+      "step": 2442
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7339984100844941,
+      "learning_rate": 9.59094417448622e-07,
+      "loss": 0.1953,
+      "step": 2443
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3210775223225176,
+      "learning_rate": 9.590534964170751e-07,
+      "loss": 0.0939,
+      "step": 2444
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4799228540673482,
+      "learning_rate": 9.590125558013833e-07,
+      "loss": 0.2605,
+      "step": 2445
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5003199535506793,
+      "learning_rate": 9.589715956032931e-07,
+      "loss": 0.1503,
+      "step": 2446
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2093919587016295,
+      "learning_rate": 9.589306158245519e-07,
+      "loss": 0.0823,
+      "step": 2447
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.211062135323147,
+      "learning_rate": 9.588896164669083e-07,
+      "loss": 0.0798,
+      "step": 2448
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8255251624391429,
+      "learning_rate": 9.58848597532111e-07,
+      "loss": 0.3699,
+      "step": 2449
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7480180932204308,
+      "learning_rate": 9.5880755902191e-07,
+      "loss": 0.1169,
+      "step": 2450
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5885945131637391,
+      "learning_rate": 9.587665009380564e-07,
+      "loss": 0.2496,
+      "step": 2451
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6911377401986551,
+      "learning_rate": 9.587254232823017e-07,
+      "loss": 0.1732,
+      "step": 2452
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2802853173063493,
+      "learning_rate": 9.586843260563981e-07,
+      "loss": 0.2303,
+      "step": 2453
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5706821042495621,
+      "learning_rate": 9.586432092620993e-07,
+      "loss": 0.1628,
+      "step": 2454
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5149533528375522,
+      "learning_rate": 9.586020729011591e-07,
+      "loss": 0.0887,
+      "step": 2455
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8646381630516424,
+      "learning_rate": 9.585609169753323e-07,
+      "loss": 0.1842,
+      "step": 2456
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5109092992198471,
+      "learning_rate": 9.585197414863754e-07,
+      "loss": 0.1276,
+      "step": 2457
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.34629214369350364,
+      "learning_rate": 9.584785464360442e-07,
+      "loss": 0.0915,
+      "step": 2458
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7161245704669781,
+      "learning_rate": 9.584373318260968e-07,
+      "loss": 0.1963,
+      "step": 2459
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4039344109309137,
+      "learning_rate": 9.583960976582913e-07,
+      "loss": 0.1153,
+      "step": 2460
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7991041837940455,
+      "learning_rate": 9.583548439343864e-07,
+      "loss": 0.2508,
+      "step": 2461
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39917016746617423,
+      "learning_rate": 9.583135706561427e-07,
+      "loss": 0.1676,
+      "step": 2462
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4471927740248249,
+      "learning_rate": 9.582722778253209e-07,
+      "loss": 0.0567,
+      "step": 2463
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5560947222512375,
+      "learning_rate": 9.582309654436824e-07,
+      "loss": 0.0954,
+      "step": 2464
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7713230345468085,
+      "learning_rate": 9.581896335129896e-07,
+      "loss": 0.1478,
+      "step": 2465
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5992686563278361,
+      "learning_rate": 9.581482820350062e-07,
+      "loss": 0.199,
+      "step": 2466
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5839880897771591,
+      "learning_rate": 9.581069110114959e-07,
+      "loss": 0.0778,
+      "step": 2467
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7665261128257583,
+      "learning_rate": 9.58065520444224e-07,
+      "loss": 0.2288,
+      "step": 2468
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3608221879220619,
+      "learning_rate": 9.58024110334956e-07,
+      "loss": 0.1416,
+      "step": 2469
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6580402781787531,
+      "learning_rate": 9.579826806854587e-07,
+      "loss": 0.2062,
+      "step": 2470
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.616037261545377,
+      "learning_rate": 9.579412314974996e-07,
+      "loss": 0.1413,
+      "step": 2471
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3845961631809863,
+      "learning_rate": 9.578997627728472e-07,
+      "loss": 0.1416,
+      "step": 2472
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6196184136165247,
+      "learning_rate": 9.578582745132702e-07,
+      "loss": 0.4322,
+      "step": 2473
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6062145541363272,
+      "learning_rate": 9.57816766720539e-07,
+      "loss": 0.0266,
+      "step": 2474
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4826225528256578,
+      "learning_rate": 9.57775239396424e-07,
+      "loss": 0.0091,
+      "step": 2475
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39785134353497276,
+      "learning_rate": 9.577336925426972e-07,
+      "loss": 0.1898,
+      "step": 2476
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6171680118428898,
+      "learning_rate": 9.576921261611308e-07,
+      "loss": 0.1495,
+      "step": 2477
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7923348973742145,
+      "learning_rate": 9.576505402534984e-07,
+      "loss": 0.1283,
+      "step": 2478
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.45345901198394867,
+      "learning_rate": 9.576089348215738e-07,
+      "loss": 0.2268,
+      "step": 2479
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.48493822051526925,
+      "learning_rate": 9.575673098671322e-07,
+      "loss": 0.2149,
+      "step": 2480
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7903018784896051,
+      "learning_rate": 9.575256653919492e-07,
+      "loss": 0.1499,
+      "step": 2481
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3313231203351894,
+      "learning_rate": 9.574840013978016e-07,
+      "loss": 0.0865,
+      "step": 2482
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37845476055638494,
+      "learning_rate": 9.57442317886467e-07,
+      "loss": 0.3144,
+      "step": 2483
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8310784224755157,
+      "learning_rate": 9.574006148597237e-07,
+      "loss": 0.2645,
+      "step": 2484
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5913650808489613,
+      "learning_rate": 9.573588923193504e-07,
+      "loss": 0.1877,
+      "step": 2485
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39724934213196095,
+      "learning_rate": 9.573171502671272e-07,
+      "loss": 0.1766,
+      "step": 2486
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.49356068507374784,
+      "learning_rate": 9.572753887048352e-07,
+      "loss": 0.2308,
+      "step": 2487
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6912161352337316,
+      "learning_rate": 9.57233607634256e-07,
+      "loss": 0.2348,
+      "step": 2488
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6673408530739752,
+      "learning_rate": 9.57191807057172e-07,
+      "loss": 0.182,
+      "step": 2489
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.11580191835952931,
+      "learning_rate": 9.57149986975366e-07,
+      "loss": 0.0047,
+      "step": 2490
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3003908593523906,
+      "learning_rate": 9.571081473906231e-07,
+      "loss": 0.0654,
+      "step": 2491
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.43900193929379083,
+      "learning_rate": 9.570662883047273e-07,
+      "loss": 0.131,
+      "step": 2492
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5021973118101629,
+      "learning_rate": 9.570244097194652e-07,
+      "loss": 0.1366,
+      "step": 2493
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.513413478436164,
+      "learning_rate": 9.569825116366229e-07,
+      "loss": 0.1909,
+      "step": 2494
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.975627560256554,
+      "learning_rate": 9.569405940579879e-07,
+      "loss": 0.1584,
+      "step": 2495
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9213510930240534,
+      "learning_rate": 9.568986569853487e-07,
+      "loss": 0.356,
+      "step": 2496
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5005513591545933,
+      "learning_rate": 9.568567004204942e-07,
+      "loss": 0.424,
+      "step": 2497
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3215748174192554,
+      "learning_rate": 9.568147243652145e-07,
+      "loss": 0.2754,
+      "step": 2498
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.592545580116721,
+      "learning_rate": 9.567727288213004e-07,
+      "loss": 0.2503,
+      "step": 2499
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.32801051559945615,
+      "learning_rate": 9.567307137905433e-07,
+      "loss": 0.239,
+      "step": 2500
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7970070432843507,
+      "learning_rate": 9.56688679274736e-07,
+      "loss": 0.2516,
+      "step": 2501
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6526024385015151,
+      "learning_rate": 9.566466252756717e-07,
+      "loss": 0.2077,
+      "step": 2502
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.27808512487183307,
+      "learning_rate": 9.566045517951443e-07,
+      "loss": 0.2123,
+      "step": 2503
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6026659699645546,
+      "learning_rate": 9.565624588349486e-07,
+      "loss": 0.3233,
+      "step": 2504
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.421675142640018,
+      "learning_rate": 9.565203463968807e-07,
+      "loss": 0.2718,
+      "step": 2505
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9978363566766066,
+      "learning_rate": 9.564782144827372e-07,
+      "loss": 0.2785,
+      "step": 2506
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5901030590545348,
+      "learning_rate": 9.564360630943153e-07,
+      "loss": 0.143,
+      "step": 2507
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8699076113639252,
+      "learning_rate": 9.563938922334136e-07,
+      "loss": 0.3662,
+      "step": 2508
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5544078938630749,
+      "learning_rate": 9.563517019018308e-07,
+      "loss": 0.2223,
+      "step": 2509
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9647226312710868,
+      "learning_rate": 9.563094921013672e-07,
+      "loss": 0.0912,
+      "step": 2510
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4990970038813086,
+      "learning_rate": 9.562672628338231e-07,
+      "loss": 0.1941,
+      "step": 2511
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15629989285131096,
+      "learning_rate": 9.562250141010006e-07,
+      "loss": 0.1158,
+      "step": 2512
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9254353943652026,
+      "learning_rate": 9.561827459047017e-07,
+      "loss": 0.2113,
+      "step": 2513
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3158188291288584,
+      "learning_rate": 9.5614045824673e-07,
+      "loss": 0.1268,
+      "step": 2514
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4468005883944894,
+      "learning_rate": 9.560981511288892e-07,
+      "loss": 0.3139,
+      "step": 2515
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3096031642338491,
+      "learning_rate": 9.560558245529847e-07,
+      "loss": 0.0759,
+      "step": 2516
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5855834841151715,
+      "learning_rate": 9.560134785208215e-07,
+      "loss": 0.2879,
+      "step": 2517
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6364203530360595,
+      "learning_rate": 9.55971113034207e-07,
+      "loss": 0.3691,
+      "step": 2518
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5258385288479613,
+      "learning_rate": 9.55928728094948e-07,
+      "loss": 0.1821,
+      "step": 2519
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.31328462882732316,
+      "learning_rate": 9.558863237048528e-07,
+      "loss": 0.2089,
+      "step": 2520
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4188983438063981,
+      "learning_rate": 9.55843899865731e-07,
+      "loss": 0.2094,
+      "step": 2521
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7633257798796065,
+      "learning_rate": 9.558014565793917e-07,
+      "loss": 0.2042,
+      "step": 2522
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3434334507804224,
+      "learning_rate": 9.557589938476462e-07,
+      "loss": 0.0868,
+      "step": 2523
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4018431164075182,
+      "learning_rate": 9.557165116723056e-07,
+      "loss": 0.3267,
+      "step": 2524
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.44637835500273204,
+      "learning_rate": 9.556740100551827e-07,
+      "loss": 0.1597,
+      "step": 2525
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2952894519625469,
+      "learning_rate": 9.556314889980906e-07,
+      "loss": 0.1849,
+      "step": 2526
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.606974143033971,
+      "learning_rate": 9.55588948502843e-07,
+      "loss": 0.2545,
+      "step": 2527
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.808011575425698,
+      "learning_rate": 9.55546388571255e-07,
+      "loss": 0.4429,
+      "step": 2528
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38018435236573644,
+      "learning_rate": 9.555038092051424e-07,
+      "loss": 0.1824,
+      "step": 2529
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5634119634022953,
+      "learning_rate": 9.554612104063218e-07,
+      "loss": 0.2054,
+      "step": 2530
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4168603419719356,
+      "learning_rate": 9.5541859217661e-07,
+      "loss": 0.055,
+      "step": 2531
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7034347148926832,
+      "learning_rate": 9.55375954517826e-07,
+      "loss": 0.4097,
+      "step": 2532
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6641957985703538,
+      "learning_rate": 9.553332974317881e-07,
+      "loss": 0.0477,
+      "step": 2533
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16677956900373406,
+      "learning_rate": 9.552906209203164e-07,
+      "loss": 0.093,
+      "step": 2534
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.672648093655843,
+      "learning_rate": 9.552479249852314e-07,
+      "loss": 0.2203,
+      "step": 2535
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9316536883359925,
+      "learning_rate": 9.55205209628355e-07,
+      "loss": 0.0684,
+      "step": 2536
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7425895926971369,
+      "learning_rate": 9.551624748515093e-07,
+      "loss": 0.3351,
+      "step": 2537
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5577282767065026,
+      "learning_rate": 9.551197206565172e-07,
+      "loss": 0.25,
+      "step": 2538
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7144229171110558,
+      "learning_rate": 9.55076947045203e-07,
+      "loss": 0.0449,
+      "step": 2539
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6754350692010835,
+      "learning_rate": 9.550341540193915e-07,
+      "loss": 0.239,
+      "step": 2540
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.48017233439651535,
+      "learning_rate": 9.549913415809083e-07,
+      "loss": 0.1089,
+      "step": 2541
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2157533579531004,
+      "learning_rate": 9.549485097315797e-07,
+      "loss": 0.0096,
+      "step": 2542
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.593163716860367,
+      "learning_rate": 9.549056584732332e-07,
+      "loss": 0.1677,
+      "step": 2543
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9026525850890124,
+      "learning_rate": 9.54862787807697e-07,
+      "loss": 0.0445,
+      "step": 2544
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7539484755867265,
+      "learning_rate": 9.548198977367997e-07,
+      "loss": 0.3909,
+      "step": 2545
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5651341354415033,
+      "learning_rate": 9.547769882623711e-07,
+      "loss": 0.2107,
+      "step": 2546
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5613177694374492,
+      "learning_rate": 9.54734059386242e-07,
+      "loss": 0.0349,
+      "step": 2547
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7745166693905394,
+      "learning_rate": 9.54691111110244e-07,
+      "loss": 0.1507,
+      "step": 2548
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5372854238568204,
+      "learning_rate": 9.54648143436209e-07,
+      "loss": 0.2093,
+      "step": 2549
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9249320537401026,
+      "learning_rate": 9.546051563659703e-07,
+      "loss": 0.1975,
+      "step": 2550
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7554143629377637,
+      "learning_rate": 9.545621499013618e-07,
+      "loss": 0.1915,
+      "step": 2551
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7802398204245109,
+      "learning_rate": 9.545191240442181e-07,
+      "loss": 0.1643,
+      "step": 2552
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8053293945513413,
+      "learning_rate": 9.54476078796375e-07,
+      "loss": 0.1896,
+      "step": 2553
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6111219177554211,
+      "learning_rate": 9.544330141596687e-07,
+      "loss": 0.1185,
+      "step": 2554
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7438516675839351,
+      "learning_rate": 9.543899301359365e-07,
+      "loss": 0.2293,
+      "step": 2555
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.210898332409302,
+      "learning_rate": 9.543468267270164e-07,
+      "loss": 0.1928,
+      "step": 2556
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5612228283646611,
+      "learning_rate": 9.543037039347473e-07,
+      "loss": 0.2255,
+      "step": 2557
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6662477311234537,
+      "learning_rate": 9.54260561760969e-07,
+      "loss": 0.1428,
+      "step": 2558
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37963751492009234,
+      "learning_rate": 9.54217400207522e-07,
+      "loss": 0.2283,
+      "step": 2559
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.48889752434070105,
+      "learning_rate": 9.541742192762476e-07,
+      "loss": 0.2742,
+      "step": 2560
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7399252688970139,
+      "learning_rate": 9.541310189689879e-07,
+      "loss": 0.3073,
+      "step": 2561
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8568202974158361,
+      "learning_rate": 9.54087799287586e-07,
+      "loss": 0.116,
+      "step": 2562
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.35524417309513245,
+      "learning_rate": 9.540445602338859e-07,
+      "loss": 0.1225,
+      "step": 2563
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8459374834491684,
+      "learning_rate": 9.54001301809732e-07,
+      "loss": 0.2193,
+      "step": 2564
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.36609078980197424,
+      "learning_rate": 9.5395802401697e-07,
+      "loss": 0.0446,
+      "step": 2565
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6045336995181586,
+      "learning_rate": 9.539147268574459e-07,
+      "loss": 0.3463,
+      "step": 2566
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6689902923021831,
+      "learning_rate": 9.538714103330073e-07,
+      "loss": 0.2253,
+      "step": 2567
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9679760657231808,
+      "learning_rate": 9.53828074445502e-07,
+      "loss": 0.0427,
+      "step": 2568
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.11885571743426666,
+      "learning_rate": 9.537847191967785e-07,
+      "loss": 0.0941,
+      "step": 2569
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5651051388610489,
+      "learning_rate": 9.537413445886868e-07,
+      "loss": 0.0448,
+      "step": 2570
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.924577730359375,
+      "learning_rate": 9.53697950623077e-07,
+      "loss": 0.3433,
+      "step": 2571
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1859239450215893,
+      "learning_rate": 9.536545373018009e-07,
+      "loss": 0.1671,
+      "step": 2572
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6339655036031436,
+      "learning_rate": 9.536111046267102e-07,
+      "loss": 0.1357,
+      "step": 2573
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9970807259261836,
+      "learning_rate": 9.535676525996577e-07,
+      "loss": 0.1535,
+      "step": 2574
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.262159878337895,
+      "learning_rate": 9.535241812224975e-07,
+      "loss": 0.2395,
+      "step": 2575
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3263401335027573,
+      "learning_rate": 9.53480690497084e-07,
+      "loss": 0.1853,
+      "step": 2576
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7486868284212916,
+      "learning_rate": 9.534371804252726e-07,
+      "loss": 0.1686,
+      "step": 2577
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8823603113615602,
+      "learning_rate": 9.533936510089197e-07,
+      "loss": 0.3605,
+      "step": 2578
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5844132652817103,
+      "learning_rate": 9.533501022498821e-07,
+      "loss": 0.2721,
+      "step": 2579
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3938624575027053,
+      "learning_rate": 9.533065341500178e-07,
+      "loss": 0.0185,
+      "step": 2580
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0230096021188217,
+      "learning_rate": 9.532629467111855e-07,
+      "loss": 0.206,
+      "step": 2581
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.47886346771279825,
+      "learning_rate": 9.532193399352448e-07,
+      "loss": 0.1239,
+      "step": 2582
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.41974794747755956,
+      "learning_rate": 9.531757138240559e-07,
+      "loss": 0.231,
+      "step": 2583
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4259696823285501,
+      "learning_rate": 9.5313206837948e-07,
+      "loss": 0.3168,
+      "step": 2584
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.21675652071370596,
+      "learning_rate": 9.530884036033793e-07,
+      "loss": 0.1029,
+      "step": 2585
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.163548726262206,
+      "learning_rate": 9.530447194976163e-07,
+      "loss": 0.0652,
+      "step": 2586
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.44003913761455,
+      "learning_rate": 9.53001016064055e-07,
+      "loss": 0.2302,
+      "step": 2587
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.20866023849578044,
+      "learning_rate": 9.529572933045595e-07,
+      "loss": 0.0638,
+      "step": 2588
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7631673046005106,
+      "learning_rate": 9.529135512209955e-07,
+      "loss": 0.3408,
+      "step": 2589
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.371969954717516,
+      "learning_rate": 9.528697898152288e-07,
+      "loss": 0.3842,
+      "step": 2590
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6568255870713744,
+      "learning_rate": 9.528260090891266e-07,
+      "loss": 0.2441,
+      "step": 2591
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.41472603573328204,
+      "learning_rate": 9.527822090445565e-07,
+      "loss": 0.1132,
+      "step": 2592
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8264338215602798,
+      "learning_rate": 9.527383896833872e-07,
+      "loss": 0.3289,
+      "step": 2593
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7378902329751158,
+      "learning_rate": 9.52694551007488e-07,
+      "loss": 0.1971,
+      "step": 2594
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5173460944303994,
+      "learning_rate": 9.526506930187292e-07,
+      "loss": 0.1739,
+      "step": 2595
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8927570683629182,
+      "learning_rate": 9.526068157189819e-07,
+      "loss": 0.4867,
+      "step": 2596
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.522756344053833,
+      "learning_rate": 9.525629191101181e-07,
+      "loss": 0.0821,
+      "step": 2597
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9624025756487392,
+      "learning_rate": 9.525190031940105e-07,
+      "loss": 0.1431,
+      "step": 2598
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9228900074603312,
+      "learning_rate": 9.524750679725323e-07,
+      "loss": 0.3242,
+      "step": 2599
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.31772156343051866,
+      "learning_rate": 9.524311134475582e-07,
+      "loss": 0.2404,
+      "step": 2600
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5661715722613065,
+      "learning_rate": 9.523871396209633e-07,
+      "loss": 0.1122,
+      "step": 2601
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.888294519797739,
+      "learning_rate": 9.523431464946236e-07,
+      "loss": 0.2555,
+      "step": 2602
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4107675325749358,
+      "learning_rate": 9.522991340704161e-07,
+      "loss": 0.3525,
+      "step": 2603
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7403158963458948,
+      "learning_rate": 9.522551023502181e-07,
+      "loss": 0.2662,
+      "step": 2604
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5424851219006864,
+      "learning_rate": 9.522110513359083e-07,
+      "loss": 0.1412,
+      "step": 2605
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4147819856967123,
+      "learning_rate": 9.521669810293661e-07,
+      "loss": 0.0952,
+      "step": 2606
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5717881857910588,
+      "learning_rate": 9.521228914324715e-07,
+      "loss": 0.0613,
+      "step": 2607
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4241907311844593,
+      "learning_rate": 9.520787825471055e-07,
+      "loss": 0.1777,
+      "step": 2608
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2865823410065802,
+      "learning_rate": 9.520346543751496e-07,
+      "loss": 0.0107,
+      "step": 2609
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8866862834291714,
+      "learning_rate": 9.519905069184869e-07,
+      "loss": 0.2322,
+      "step": 2610
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8196263524169246,
+      "learning_rate": 9.519463401790004e-07,
+      "loss": 0.2045,
+      "step": 2611
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7649383541668356,
+      "learning_rate": 9.519021541585748e-07,
+      "loss": 0.3651,
+      "step": 2612
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.44313444995716733,
+      "learning_rate": 9.518579488590946e-07,
+      "loss": 0.3697,
+      "step": 2613
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5828627418819751,
+      "learning_rate": 9.51813724282446e-07,
+      "loss": 0.2626,
+      "step": 2614
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2969889590099513,
+      "learning_rate": 9.517694804305156e-07,
+      "loss": 0.182,
+      "step": 2615
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.20698729443897265,
+      "learning_rate": 9.517252173051911e-07,
+      "loss": 0.004,
+      "step": 2616
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9516054489380347,
+      "learning_rate": 9.516809349083608e-07,
+      "loss": 0.1901,
+      "step": 2617
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3040730187488041,
+      "learning_rate": 9.516366332419137e-07,
+      "loss": 0.07,
+      "step": 2618
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7427736545565709,
+      "learning_rate": 9.515923123077399e-07,
+      "loss": 0.3721,
+      "step": 2619
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4227813936269338,
+      "learning_rate": 9.515479721077303e-07,
+      "loss": 0.1311,
+      "step": 2620
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9890664145268651,
+      "learning_rate": 9.515036126437766e-07,
+      "loss": 0.1963,
+      "step": 2621
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4338410237920062,
+      "learning_rate": 9.514592339177709e-07,
+      "loss": 0.1358,
+      "step": 2622
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6461712191502266,
+      "learning_rate": 9.514148359316069e-07,
+      "loss": 0.1093,
+      "step": 2623
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0781316168167339,
+      "learning_rate": 9.513704186871785e-07,
+      "loss": 0.5929,
+      "step": 2624
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5937921628955494,
+      "learning_rate": 9.513259821863806e-07,
+      "loss": 0.025,
+      "step": 2625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2050441558443006,
+      "learning_rate": 9.512815264311092e-07,
+      "loss": 0.3322,
+      "step": 2626
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6447535671899982,
+      "learning_rate": 9.512370514232606e-07,
+      "loss": 0.2322,
+      "step": 2627
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38982095788275767,
+      "learning_rate": 9.511925571647322e-07,
+      "loss": 0.1749,
+      "step": 2628
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2553585210304645,
+      "learning_rate": 9.511480436574224e-07,
+      "loss": 0.2043,
+      "step": 2629
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6165582812152801,
+      "learning_rate": 9.511035109032301e-07,
+      "loss": 0.0743,
+      "step": 2630
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3624701319014514,
+      "learning_rate": 9.510589589040552e-07,
+      "loss": 0.3819,
+      "step": 2631
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4195544643620065,
+      "learning_rate": 9.510143876617985e-07,
+      "loss": 0.1449,
+      "step": 2632
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2960278562831275,
+      "learning_rate": 9.509697971783612e-07,
+      "loss": 0.0074,
+      "step": 2633
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.45621861203351,
+      "learning_rate": 9.509251874556459e-07,
+      "loss": 0.2407,
+      "step": 2634
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9044462460962452,
+      "learning_rate": 9.508805584955555e-07,
+      "loss": 0.0322,
+      "step": 2635
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6058690813013976,
+      "learning_rate": 9.508359102999941e-07,
+      "loss": 0.2087,
+      "step": 2636
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7613827905763582,
+      "learning_rate": 9.507912428708666e-07,
+      "loss": 0.3585,
+      "step": 2637
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6886394154987239,
+      "learning_rate": 9.507465562100784e-07,
+      "loss": 0.2638,
+      "step": 2638
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7294311137348841,
+      "learning_rate": 9.507018503195361e-07,
+      "loss": 0.1544,
+      "step": 2639
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5800813312282401,
+      "learning_rate": 9.506571252011466e-07,
+      "loss": 0.2576,
+      "step": 2640
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.982939595532959,
+      "learning_rate": 9.506123808568185e-07,
+      "loss": 0.1237,
+      "step": 2641
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.49116126395133736,
+      "learning_rate": 9.505676172884601e-07,
+      "loss": 0.1741,
+      "step": 2642
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8696972031988437,
+      "learning_rate": 9.505228344979817e-07,
+      "loss": 0.2839,
+      "step": 2643
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.45227011206891493,
+      "learning_rate": 9.504780324872932e-07,
+      "loss": 0.1896,
+      "step": 2644
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9564678587797452,
+      "learning_rate": 9.504332112583064e-07,
+      "loss": 0.3192,
+      "step": 2645
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6422774084741891,
+      "learning_rate": 9.503883708129334e-07,
+      "loss": 0.2231,
+      "step": 2646
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9070393905137192,
+      "learning_rate": 9.50343511153087e-07,
+      "loss": 0.2916,
+      "step": 2647
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7973637615785845,
+      "learning_rate": 9.502986322806811e-07,
+      "loss": 0.5046,
+      "step": 2648
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7107042441039871,
+      "learning_rate": 9.502537341976305e-07,
+      "loss": 0.1953,
+      "step": 2649
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6169596807696893,
+      "learning_rate": 9.502088169058503e-07,
+      "loss": 0.1974,
+      "step": 2650
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5224394109098526,
+      "learning_rate": 9.501638804072569e-07,
+      "loss": 0.2402,
+      "step": 2651
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3155133421120215,
+      "learning_rate": 9.501189247037675e-07,
+      "loss": 0.0924,
+      "step": 2652
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4912769084674077,
+      "learning_rate": 9.500739497973e-07,
+      "loss": 0.1181,
+      "step": 2653
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8258919552020955,
+      "learning_rate": 9.500289556897729e-07,
+      "loss": 0.254,
+      "step": 2654
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5399399057606297,
+      "learning_rate": 9.49983942383106e-07,
+      "loss": 0.2062,
+      "step": 2655
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.48393195306922543,
+      "learning_rate": 9.499389098792196e-07,
+      "loss": 0.1076,
+      "step": 2656
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0377411795960845,
+      "learning_rate": 9.498938581800347e-07,
+      "loss": 0.1113,
+      "step": 2657
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4560969694650829,
+      "learning_rate": 9.498487872874733e-07,
+      "loss": 0.3368,
+      "step": 2658
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9862958881707448,
+      "learning_rate": 9.498036972034584e-07,
+      "loss": 0.1763,
+      "step": 2659
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6833198751285896,
+      "learning_rate": 9.497585879299137e-07,
+      "loss": 0.28,
+      "step": 2660
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4888347324512454,
+      "learning_rate": 9.497134594687634e-07,
+      "loss": 0.2217,
+      "step": 2661
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5719202825316656,
+      "learning_rate": 9.496683118219328e-07,
+      "loss": 0.2603,
+      "step": 2662
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7848817116848156,
+      "learning_rate": 9.496231449913482e-07,
+      "loss": 0.0865,
+      "step": 2663
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5583436160757501,
+      "learning_rate": 9.495779589789364e-07,
+      "loss": 0.1155,
+      "step": 2664
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.29229710747176135,
+      "learning_rate": 9.495327537866249e-07,
+      "loss": 0.0897,
+      "step": 2665
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6378875468827189,
+      "learning_rate": 9.494875294163427e-07,
+      "loss": 0.3491,
+      "step": 2666
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7831721155048247,
+      "learning_rate": 9.494422858700187e-07,
+      "loss": 0.2136,
+      "step": 2667
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7366203540518024,
+      "learning_rate": 9.493970231495834e-07,
+      "loss": 0.3171,
+      "step": 2668
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3410124132656613,
+      "learning_rate": 9.493517412569678e-07,
+      "loss": 0.0108,
+      "step": 2669
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0068122108176234,
+      "learning_rate": 9.493064401941034e-07,
+      "loss": 0.1751,
+      "step": 2670
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8529491418660656,
+      "learning_rate": 9.492611199629232e-07,
+      "loss": 0.2523,
+      "step": 2671
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.25242722415960495,
+      "learning_rate": 9.492157805653604e-07,
+      "loss": 0.1797,
+      "step": 2672
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.085744296075026,
+      "learning_rate": 9.491704220033494e-07,
+      "loss": 0.1487,
+      "step": 2673
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5368273019639394,
+      "learning_rate": 9.491250442788252e-07,
+      "loss": 0.1409,
+      "step": 2674
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7625266156549131,
+      "learning_rate": 9.490796473937238e-07,
+      "loss": 0.2027,
+      "step": 2675
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.48853974831983377,
+      "learning_rate": 9.49034231349982e-07,
+      "loss": 0.0527,
+      "step": 2676
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5674460569210614,
+      "learning_rate": 9.48988796149537e-07,
+      "loss": 0.1686,
+      "step": 2677
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3302941832085532,
+      "learning_rate": 9.489433417943275e-07,
+      "loss": 0.2061,
+      "step": 2678
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6617601654386271,
+      "learning_rate": 9.488978682862925e-07,
+      "loss": 0.0978,
+      "step": 2679
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2120206993789706,
+      "learning_rate": 9.488523756273721e-07,
+      "loss": 0.1629,
+      "step": 2680
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.27118500248326494,
+      "learning_rate": 9.488068638195071e-07,
+      "loss": 0.1445,
+      "step": 2681
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2673674372884545,
+      "learning_rate": 9.487613328646389e-07,
+      "loss": 0.1109,
+      "step": 2682
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39638542360387147,
+      "learning_rate": 9.487157827647101e-07,
+      "loss": 0.1716,
+      "step": 2683
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8976279257317973,
+      "learning_rate": 9.486702135216643e-07,
+      "loss": 0.3226,
+      "step": 2684
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8303920226304105,
+      "learning_rate": 9.486246251374449e-07,
+      "loss": 0.253,
+      "step": 2685
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5509054405484801,
+      "learning_rate": 9.485790176139972e-07,
+      "loss": 0.2561,
+      "step": 2686
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4830660775150088,
+      "learning_rate": 9.485333909532669e-07,
+      "loss": 0.1147,
+      "step": 2687
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8118934988535611,
+      "learning_rate": 9.484877451572007e-07,
+      "loss": 0.4417,
+      "step": 2688
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3074240992680236,
+      "learning_rate": 9.484420802277454e-07,
+      "loss": 0.0119,
+      "step": 2689
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4397118420494491,
+      "learning_rate": 9.483963961668496e-07,
+      "loss": 0.3114,
+      "step": 2690
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5599381421398959,
+      "learning_rate": 9.483506929764622e-07,
+      "loss": 0.1553,
+      "step": 2691
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5900923425201272,
+      "learning_rate": 9.483049706585329e-07,
+      "loss": 0.0354,
+      "step": 2692
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7315701933263555,
+      "learning_rate": 9.482592292150122e-07,
+      "loss": 0.1048,
+      "step": 2693
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5995646513358464,
+      "learning_rate": 9.482134686478518e-07,
+      "loss": 0.2526,
+      "step": 2694
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5594556337693576,
+      "learning_rate": 9.481676889590038e-07,
+      "loss": 0.2493,
+      "step": 2695
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.006129048690075,
+      "learning_rate": 9.481218901504214e-07,
+      "loss": 0.4051,
+      "step": 2696
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.46685641387092186,
+      "learning_rate": 9.480760722240582e-07,
+      "loss": 0.2618,
+      "step": 2697
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2050649689990065,
+      "learning_rate": 9.480302351818689e-07,
+      "loss": 0.1889,
+      "step": 2698
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.35983640486842944,
+      "learning_rate": 9.479843790258093e-07,
+      "loss": 0.1236,
+      "step": 2699
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6665101621244433,
+      "learning_rate": 9.479385037578354e-07,
+      "loss": 0.3935,
+      "step": 2700
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7461767823578016,
+      "learning_rate": 9.478926093799045e-07,
+      "loss": 0.0802,
+      "step": 2701
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.729629630723568,
+      "learning_rate": 9.478466958939745e-07,
+      "loss": 0.2092,
+      "step": 2702
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5086910171019614,
+      "learning_rate": 9.478007633020042e-07,
+      "loss": 0.5959,
+      "step": 2703
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.004591934260547,
+      "learning_rate": 9.47754811605953e-07,
+      "loss": 0.2812,
+      "step": 2704
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.46586024315715674,
+      "learning_rate": 9.477088408077816e-07,
+      "loss": 0.1034,
+      "step": 2705
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6870177141639254,
+      "learning_rate": 9.476628509094511e-07,
+      "loss": 0.0949,
+      "step": 2706
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.890797400301904,
+      "learning_rate": 9.476168419129234e-07,
+      "loss": 0.2011,
+      "step": 2707
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4284635485018434,
+      "learning_rate": 9.475708138201612e-07,
+      "loss": 0.0761,
+      "step": 2708
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.49520868939639545,
+      "learning_rate": 9.475247666331285e-07,
+      "loss": 0.1412,
+      "step": 2709
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7249237889122515,
+      "learning_rate": 9.474787003537897e-07,
+      "loss": 0.305,
+      "step": 2710
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3673553451629228,
+      "learning_rate": 9.474326149841099e-07,
+      "loss": 0.1136,
+      "step": 2711
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9513134842408071,
+      "learning_rate": 9.473865105260554e-07,
+      "loss": 0.0823,
+      "step": 2712
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.275082949196564,
+      "learning_rate": 9.47340386981593e-07,
+      "loss": 0.0768,
+      "step": 2713
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38376522592074097,
+      "learning_rate": 9.472942443526903e-07,
+      "loss": 0.016,
+      "step": 2714
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6354252208438675,
+      "learning_rate": 9.472480826413161e-07,
+      "loss": 0.2452,
+      "step": 2715
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7838325564258156,
+      "learning_rate": 9.472019018494395e-07,
+      "loss": 0.2635,
+      "step": 2716
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7046123557223944,
+      "learning_rate": 9.471557019790308e-07,
+      "loss": 0.1751,
+      "step": 2717
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1994700866665963,
+      "learning_rate": 9.471094830320609e-07,
+      "loss": 0.0422,
+      "step": 2718
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0426057348591593,
+      "learning_rate": 9.470632450105018e-07,
+      "loss": 0.1409,
+      "step": 2719
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2426962738025242,
+      "learning_rate": 9.470169879163258e-07,
+      "loss": 0.1847,
+      "step": 2720
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1177294538934748,
+      "learning_rate": 9.469707117515066e-07,
+      "loss": 0.2524,
+      "step": 2721
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.843378487215984,
+      "learning_rate": 9.469244165180183e-07,
+      "loss": 0.1706,
+      "step": 2722
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.526731905595799,
+      "learning_rate": 9.46878102217836e-07,
+      "loss": 0.3975,
+      "step": 2723
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37146034006379797,
+      "learning_rate": 9.468317688529354e-07,
+      "loss": 0.1483,
+      "step": 2724
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1941258775726189,
+      "learning_rate": 9.467854164252934e-07,
+      "loss": 0.0108,
+      "step": 2725
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6248765243759603,
+      "learning_rate": 9.467390449368873e-07,
+      "loss": 0.1815,
+      "step": 2726
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5849445940729799,
+      "learning_rate": 9.466926543896954e-07,
+      "loss": 0.2543,
+      "step": 2727
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6364373563334739,
+      "learning_rate": 9.466462447856971e-07,
+      "loss": 0.1878,
+      "step": 2728
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6116281503984071,
+      "learning_rate": 9.46599816126872e-07,
+      "loss": 0.1425,
+      "step": 2729
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8351472099940584,
+      "learning_rate": 9.465533684152009e-07,
+      "loss": 0.1702,
+      "step": 2730
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.22066066406570445,
+      "learning_rate": 9.465069016526657e-07,
+      "loss": 0.0086,
+      "step": 2731
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5012185405558273,
+      "learning_rate": 9.464604158412483e-07,
+      "loss": 0.3493,
+      "step": 2732
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7384931294052169,
+      "learning_rate": 9.46413910982932e-07,
+      "loss": 0.1253,
+      "step": 2733
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.55046884653624,
+      "learning_rate": 9.463673870797008e-07,
+      "loss": 0.1763,
+      "step": 2734
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6388104129942009,
+      "learning_rate": 9.463208441335398e-07,
+      "loss": 0.3693,
+      "step": 2735
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7641560230791178,
+      "learning_rate": 9.462742821464342e-07,
+      "loss": 0.2299,
+      "step": 2736
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2596875295580591,
+      "learning_rate": 9.462277011203707e-07,
+      "loss": 0.1694,
+      "step": 2737
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.551082981625449,
+      "learning_rate": 9.461811010573364e-07,
+      "loss": 0.0953,
+      "step": 2738
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4609512229823886,
+      "learning_rate": 9.461344819593193e-07,
+      "loss": 0.0984,
+      "step": 2739
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4749796653783331,
+      "learning_rate": 9.460878438283085e-07,
+      "loss": 0.1183,
+      "step": 2740
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7080032973932419,
+      "learning_rate": 9.460411866662935e-07,
+      "loss": 0.3335,
+      "step": 2741
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3570665673311762,
+      "learning_rate": 9.459945104752648e-07,
+      "loss": 0.1678,
+      "step": 2742
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0982488308184704,
+      "learning_rate": 9.459478152572138e-07,
+      "loss": 0.1659,
+      "step": 2743
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7899017343218295,
+      "learning_rate": 9.459011010141324e-07,
+      "loss": 0.1995,
+      "step": 2744
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.87676712175917,
+      "learning_rate": 9.458543677480137e-07,
+      "loss": 0.1066,
+      "step": 2745
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6629090266480245,
+      "learning_rate": 9.458076154608515e-07,
+      "loss": 0.2194,
+      "step": 2746
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6939508591988042,
+      "learning_rate": 9.457608441546401e-07,
+      "loss": 0.2803,
+      "step": 2747
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0121285258968407,
+      "learning_rate": 9.45714053831375e-07,
+      "loss": 0.148,
+      "step": 2748
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7661209179901457,
+      "learning_rate": 9.456672444930524e-07,
+      "loss": 0.0927,
+      "step": 2749
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7991497350903882,
+      "learning_rate": 9.456204161416692e-07,
+      "loss": 0.0553,
+      "step": 2750
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7160695877161593,
+      "learning_rate": 9.455735687792232e-07,
+      "loss": 0.3314,
+      "step": 2751
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0368872036386638,
+      "learning_rate": 9.455267024077131e-07,
+      "loss": 0.1253,
+      "step": 2752
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4879145975404485,
+      "learning_rate": 9.454798170291384e-07,
+      "loss": 0.396,
+      "step": 2753
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8331207912215928,
+      "learning_rate": 9.454329126454991e-07,
+      "loss": 0.0842,
+      "step": 2754
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.47135914673985724,
+      "learning_rate": 9.453859892587962e-07,
+      "loss": 0.2683,
+      "step": 2755
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5011380936496737,
+      "learning_rate": 9.453390468710317e-07,
+      "loss": 0.2232,
+      "step": 2756
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5455151754771657,
+      "learning_rate": 9.452920854842084e-07,
+      "loss": 0.0862,
+      "step": 2757
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4558905703488411,
+      "learning_rate": 9.452451051003294e-07,
+      "loss": 0.2177,
+      "step": 2758
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7678508793696209,
+      "learning_rate": 9.451981057213991e-07,
+      "loss": 0.4681,
+      "step": 2759
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7600235819809289,
+      "learning_rate": 9.451510873494228e-07,
+      "loss": 0.2709,
+      "step": 2760
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2709085011050159,
+      "learning_rate": 9.451040499864061e-07,
+      "loss": 0.34,
+      "step": 2761
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0000470715173306,
+      "learning_rate": 9.45056993634356e-07,
+      "loss": 0.1959,
+      "step": 2762
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0978895279532295,
+      "learning_rate": 9.450099182952797e-07,
+      "loss": 0.2758,
+      "step": 2763
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15437705270089677,
+      "learning_rate": 9.449628239711859e-07,
+      "loss": 0.0788,
+      "step": 2764
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6322614890421141,
+      "learning_rate": 9.449157106640834e-07,
+      "loss": 0.2018,
+      "step": 2765
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4956315422057888,
+      "learning_rate": 9.448685783759825e-07,
+      "loss": 0.1677,
+      "step": 2766
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6298075310166793,
+      "learning_rate": 9.448214271088936e-07,
+      "loss": 0.1066,
+      "step": 2767
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0729077712714175,
+      "learning_rate": 9.447742568648285e-07,
+      "loss": 0.0818,
+      "step": 2768
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.47651090783674244,
+      "learning_rate": 9.447270676457994e-07,
+      "loss": 0.3393,
+      "step": 2769
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5555946853100611,
+      "learning_rate": 9.446798594538194e-07,
+      "loss": 0.356,
+      "step": 2770
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.47069576540041774,
+      "learning_rate": 9.446326322909031e-07,
+      "loss": 0.177,
+      "step": 2771
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.022874544414808,
+      "learning_rate": 9.445853861590646e-07,
+      "loss": 0.1407,
+      "step": 2772
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.76619778994947,
+      "learning_rate": 9.445381210603198e-07,
+      "loss": 0.1177,
+      "step": 2773
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.816355372826185,
+      "learning_rate": 9.444908369966852e-07,
+      "loss": 0.3085,
+      "step": 2774
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0502165711859144,
+      "learning_rate": 9.444435339701779e-07,
+      "loss": 0.1333,
+      "step": 2775
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4496538403505556,
+      "learning_rate": 9.443962119828161e-07,
+      "loss": 0.0114,
+      "step": 2776
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7821882847088305,
+      "learning_rate": 9.443488710366184e-07,
+      "loss": 0.2463,
+      "step": 2777
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2748034457238213,
+      "learning_rate": 9.443015111336048e-07,
+      "loss": 0.4336,
+      "step": 2778
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.46986003249958325,
+      "learning_rate": 9.442541322757953e-07,
+      "loss": 0.281,
+      "step": 2779
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0312067062208974,
+      "learning_rate": 9.442067344652117e-07,
+      "loss": 0.3935,
+      "step": 2780
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8290958117572361,
+      "learning_rate": 9.441593177038758e-07,
+      "loss": 0.098,
+      "step": 2781
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4844868718260467,
+      "learning_rate": 9.441118819938104e-07,
+      "loss": 0.1814,
+      "step": 2782
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6000533724746735,
+      "learning_rate": 9.440644273370395e-07,
+      "loss": 0.0846,
+      "step": 2783
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2138730908567696,
+      "learning_rate": 9.440169537355873e-07,
+      "loss": 0.0233,
+      "step": 2784
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5013900350971062,
+      "learning_rate": 9.439694611914795e-07,
+      "loss": 0.1756,
+      "step": 2785
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.36931333974677744,
+      "learning_rate": 9.439219497067417e-07,
+      "loss": 0.0095,
+      "step": 2786
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3750648838577161,
+      "learning_rate": 9.438744192834013e-07,
+      "loss": 0.1656,
+      "step": 2787
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.605318119090337,
+      "learning_rate": 9.438268699234858e-07,
+      "loss": 0.0886,
+      "step": 2788
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2959328351693167,
+      "learning_rate": 9.437793016290239e-07,
+      "loss": 0.115,
+      "step": 2789
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4474127652051423,
+      "learning_rate": 9.43731714402045e-07,
+      "loss": 0.0102,
+      "step": 2790
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6136794135858379,
+      "learning_rate": 9.436841082445788e-07,
+      "loss": 0.0886,
+      "step": 2791
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1717763396087342,
+      "learning_rate": 9.436364831586569e-07,
+      "loss": 0.3371,
+      "step": 2792
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7635911988360328,
+      "learning_rate": 9.435888391463107e-07,
+      "loss": 0.3099,
+      "step": 2793
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4099248645231432,
+      "learning_rate": 9.435411762095729e-07,
+      "loss": 0.0585,
+      "step": 2794
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.011576269807772,
+      "learning_rate": 9.43493494350477e-07,
+      "loss": 0.1689,
+      "step": 2795
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.22326415266088995,
+      "learning_rate": 9.434457935710569e-07,
+      "loss": 0.0892,
+      "step": 2796
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.19328249836507289,
+      "learning_rate": 9.43398073873348e-07,
+      "loss": 0.0992,
+      "step": 2797
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9261282826568348,
+      "learning_rate": 9.433503352593859e-07,
+      "loss": 0.2951,
+      "step": 2798
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.40628338413841864,
+      "learning_rate": 9.433025777312072e-07,
+      "loss": 0.0931,
+      "step": 2799
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8013568734705643,
+      "learning_rate": 9.432548012908495e-07,
+      "loss": 0.2714,
+      "step": 2800
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7317670146233269,
+      "learning_rate": 9.432070059403507e-07,
+      "loss": 0.1874,
+      "step": 2801
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9686810419542108,
+      "learning_rate": 9.431591916817502e-07,
+      "loss": 0.1488,
+      "step": 2802
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8012002521344556,
+      "learning_rate": 9.431113585170877e-07,
+      "loss": 0.1894,
+      "step": 2803
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8454033223737674,
+      "learning_rate": 9.430635064484038e-07,
+      "loss": 0.2057,
+      "step": 2804
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7072694430681157,
+      "learning_rate": 9.430156354777402e-07,
+      "loss": 0.4607,
+      "step": 2805
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4874392147836718,
+      "learning_rate": 9.42967745607139e-07,
+      "loss": 0.1571,
+      "step": 2806
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5946393715378242,
+      "learning_rate": 9.429198368386433e-07,
+      "loss": 0.0976,
+      "step": 2807
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.51813433087222,
+      "learning_rate": 9.428719091742968e-07,
+      "loss": 0.2874,
+      "step": 2808
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4208942596352635,
+      "learning_rate": 9.428239626161445e-07,
+      "loss": 0.0901,
+      "step": 2809
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.23755054114355956,
+      "learning_rate": 9.427759971662318e-07,
+      "loss": 0.1652,
+      "step": 2810
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8589361478411537,
+      "learning_rate": 9.427280128266049e-07,
+      "loss": 0.1959,
+      "step": 2811
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6991347503127461,
+      "learning_rate": 9.426800095993111e-07,
+      "loss": 0.3735,
+      "step": 2812
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7716760270043876,
+      "learning_rate": 9.426319874863981e-07,
+      "loss": 0.1688,
+      "step": 2813
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7459044544613995,
+      "learning_rate": 9.425839464899145e-07,
+      "loss": 0.5136,
+      "step": 2814
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7709215179212534,
+      "learning_rate": 9.425358866119104e-07,
+      "loss": 0.2338,
+      "step": 2815
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3334461217493869,
+      "learning_rate": 9.424878078544356e-07,
+      "loss": 0.2086,
+      "step": 2816
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7471230427630468,
+      "learning_rate": 9.424397102195414e-07,
+      "loss": 0.2253,
+      "step": 2817
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8213268370779977,
+      "learning_rate": 9.423915937092798e-07,
+      "loss": 0.1502,
+      "step": 2818
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4561214217091379,
+      "learning_rate": 9.423434583257035e-07,
+      "loss": 0.1782,
+      "step": 2819
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.807199616528415,
+      "learning_rate": 9.42295304070866e-07,
+      "loss": 0.154,
+      "step": 2820
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41297631557583814,
+      "learning_rate": 9.422471309468216e-07,
+      "loss": 0.1923,
+      "step": 2821
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5799683306955786,
+      "learning_rate": 9.421989389556258e-07,
+      "loss": 0.2438,
+      "step": 2822
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.772270986558091,
+      "learning_rate": 9.421507280993341e-07,
+      "loss": 0.2805,
+      "step": 2823
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1538887519345769,
+      "learning_rate": 9.421024983800037e-07,
+      "loss": 0.1897,
+      "step": 2824
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3294123485329724,
+      "learning_rate": 9.42054249799692e-07,
+      "loss": 0.1754,
+      "step": 2825
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.43713400313651707,
+      "learning_rate": 9.420059823604571e-07,
+      "loss": 0.0802,
+      "step": 2826
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6594140278897644,
+      "learning_rate": 9.419576960643587e-07,
+      "loss": 0.2665,
+      "step": 2827
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8470930858443656,
+      "learning_rate": 9.419093909134563e-07,
+      "loss": 0.2339,
+      "step": 2828
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4306736376756811,
+      "learning_rate": 9.418610669098113e-07,
+      "loss": 0.1545,
+      "step": 2829
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.365434225331647,
+      "learning_rate": 9.418127240554845e-07,
+      "loss": 0.2879,
+      "step": 2830
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7651480910982791,
+      "learning_rate": 9.417643623525391e-07,
+      "loss": 0.2465,
+      "step": 2831
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4917679686000983,
+      "learning_rate": 9.417159818030378e-07,
+      "loss": 0.0449,
+      "step": 2832
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1656930051141519,
+      "learning_rate": 9.416675824090448e-07,
+      "loss": 0.1094,
+      "step": 2833
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.41184955136228063,
+      "learning_rate": 9.416191641726248e-07,
+      "loss": 0.1345,
+      "step": 2834
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4588694991771305,
+      "learning_rate": 9.415707270958435e-07,
+      "loss": 0.1318,
+      "step": 2835
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.807216627429904,
+      "learning_rate": 9.415222711807673e-07,
+      "loss": 0.3597,
+      "step": 2836
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7407881832879403,
+      "learning_rate": 9.414737964294634e-07,
+      "loss": 0.2336,
+      "step": 2837
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5398464107680441,
+      "learning_rate": 9.41425302844e-07,
+      "loss": 0.1482,
+      "step": 2838
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2362484444063,
+      "learning_rate": 9.413767904264457e-07,
+      "loss": 0.128,
+      "step": 2839
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45437418205698415,
+      "learning_rate": 9.413282591788703e-07,
+      "loss": 0.0624,
+      "step": 2840
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37334653359496567,
+      "learning_rate": 9.412797091033442e-07,
+      "loss": 0.0551,
+      "step": 2841
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9241025056858144,
+      "learning_rate": 9.412311402019387e-07,
+      "loss": 0.0881,
+      "step": 2842
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2378366271539436,
+      "learning_rate": 9.411825524767255e-07,
+      "loss": 0.0359,
+      "step": 2843
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0344821750785687,
+      "learning_rate": 9.411339459297779e-07,
+      "loss": 0.0915,
+      "step": 2844
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9693798456261467,
+      "learning_rate": 9.410853205631693e-07,
+      "loss": 0.3174,
+      "step": 2845
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5421670416631592,
+      "learning_rate": 9.410366763789743e-07,
+      "loss": 0.1069,
+      "step": 2846
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2709530149224765,
+      "learning_rate": 9.409880133792682e-07,
+      "loss": 0.1021,
+      "step": 2847
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6347101893040504,
+      "learning_rate": 9.409393315661268e-07,
+      "loss": 0.1221,
+      "step": 2848
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8951731804134153,
+      "learning_rate": 9.408906309416271e-07,
+      "loss": 0.1671,
+      "step": 2849
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7422000431521963,
+      "learning_rate": 9.40841911507847e-07,
+      "loss": 0.2216,
+      "step": 2850
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.670706390499011,
+      "learning_rate": 9.407931732668645e-07,
+      "loss": 0.185,
+      "step": 2851
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2784967892109064,
+      "learning_rate": 9.407444162207591e-07,
+      "loss": 0.0595,
+      "step": 2852
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6966841610794263,
+      "learning_rate": 9.40695640371611e-07,
+      "loss": 0.4388,
+      "step": 2853
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8689733905563939,
+      "learning_rate": 9.406468457215011e-07,
+      "loss": 0.3751,
+      "step": 2854
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3656774912938305,
+      "learning_rate": 9.405980322725109e-07,
+      "loss": 0.1813,
+      "step": 2855
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6905757753799551,
+      "learning_rate": 9.405492000267228e-07,
+      "loss": 0.2256,
+      "step": 2856
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1243153878301837,
+      "learning_rate": 9.405003489862202e-07,
+      "loss": 0.1498,
+      "step": 2857
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7253835102938415,
+      "learning_rate": 9.404514791530873e-07,
+      "loss": 0.4478,
+      "step": 2858
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4978015300325407,
+      "learning_rate": 9.404025905294088e-07,
+      "loss": 0.1165,
+      "step": 2859
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.44617374946937977,
+      "learning_rate": 9.403536831172706e-07,
+      "loss": 0.0725,
+      "step": 2860
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6314616661282472,
+      "learning_rate": 9.40304756918759e-07,
+      "loss": 0.1408,
+      "step": 2861
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6965625633951027,
+      "learning_rate": 9.402558119359614e-07,
+      "loss": 0.1106,
+      "step": 2862
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.22985774636408715,
+      "learning_rate": 9.402068481709657e-07,
+      "loss": 0.1237,
+      "step": 2863
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6809065562402292,
+      "learning_rate": 9.40157865625861e-07,
+      "loss": 0.2483,
+      "step": 2864
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9062160343423031,
+      "learning_rate": 9.401088643027369e-07,
+      "loss": 0.196,
+      "step": 2865
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7100489542755926,
+      "learning_rate": 9.400598442036839e-07,
+      "loss": 0.063,
+      "step": 2866
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.42267334696169123,
+      "learning_rate": 9.400108053307934e-07,
+      "loss": 0.0933,
+      "step": 2867
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.22334642641603122,
+      "learning_rate": 9.399617476861573e-07,
+      "loss": 0.2095,
+      "step": 2868
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2243792781462794,
+      "learning_rate": 9.399126712718687e-07,
+      "loss": 0.1065,
+      "step": 2869
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.362080305206458,
+      "learning_rate": 9.398635760900211e-07,
+      "loss": 0.0415,
+      "step": 2870
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8801455382905071,
+      "learning_rate": 9.398144621427093e-07,
+      "loss": 0.2932,
+      "step": 2871
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0103986879678948,
+      "learning_rate": 9.397653294320282e-07,
+      "loss": 0.2368,
+      "step": 2872
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.625343372049108,
+      "learning_rate": 9.397161779600742e-07,
+      "loss": 0.216,
+      "step": 2873
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1910556190819925,
+      "learning_rate": 9.396670077289441e-07,
+      "loss": 0.1636,
+      "step": 2874
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5439027466683719,
+      "learning_rate": 9.396178187407356e-07,
+      "loss": 0.3946,
+      "step": 2875
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3236032792196466,
+      "learning_rate": 9.395686109975473e-07,
+      "loss": 0.1851,
+      "step": 2876
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2515487957717811,
+      "learning_rate": 9.395193845014784e-07,
+      "loss": 0.0908,
+      "step": 2877
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8810352395602445,
+      "learning_rate": 9.39470139254629e-07,
+      "loss": 0.2022,
+      "step": 2878
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6101326355028606,
+      "learning_rate": 9.394208752590997e-07,
+      "loss": 0.0564,
+      "step": 2879
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8622326528124292,
+      "learning_rate": 9.393715925169929e-07,
+      "loss": 0.1853,
+      "step": 2880
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6068834151155335,
+      "learning_rate": 9.393222910304106e-07,
+      "loss": 0.3243,
+      "step": 2881
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7504937712945394,
+      "learning_rate": 9.392729708014562e-07,
+      "loss": 0.1388,
+      "step": 2882
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8035627430314166,
+      "learning_rate": 9.392236318322337e-07,
+      "loss": 0.2444,
+      "step": 2883
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5813276585105368,
+      "learning_rate": 9.391742741248483e-07,
+      "loss": 0.0145,
+      "step": 2884
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3399477342637993,
+      "learning_rate": 9.391248976814054e-07,
+      "loss": 0.3238,
+      "step": 2885
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.35618492077932135,
+      "learning_rate": 9.390755025040118e-07,
+      "loss": 0.1506,
+      "step": 2886
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.10978232209715,
+      "learning_rate": 9.390260885947745e-07,
+      "loss": 0.1978,
+      "step": 2887
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9966430862469495,
+      "learning_rate": 9.389766559558017e-07,
+      "loss": 0.0761,
+      "step": 2888
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5628107443022811,
+      "learning_rate": 9.389272045892023e-07,
+      "loss": 0.1672,
+      "step": 2889
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5324227945428179,
+      "learning_rate": 9.38877734497086e-07,
+      "loss": 0.098,
+      "step": 2890
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1450986980950073,
+      "learning_rate": 9.388282456815632e-07,
+      "loss": 0.0692,
+      "step": 2891
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2870732688804849,
+      "learning_rate": 9.387787381447454e-07,
+      "loss": 0.1587,
+      "step": 2892
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1026915337376895,
+      "learning_rate": 9.387292118887444e-07,
+      "loss": 0.2455,
+      "step": 2893
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6398918224849218,
+      "learning_rate": 9.386796669156735e-07,
+      "loss": 0.3308,
+      "step": 2894
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9118360693252334,
+      "learning_rate": 9.386301032276461e-07,
+      "loss": 0.1218,
+      "step": 2895
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.44660226248487206,
+      "learning_rate": 9.385805208267766e-07,
+      "loss": 0.0959,
+      "step": 2896
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.05321455559004,
+      "learning_rate": 9.385309197151805e-07,
+      "loss": 0.1418,
+      "step": 2897
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6948435045112532,
+      "learning_rate": 9.384812998949739e-07,
+      "loss": 0.2146,
+      "step": 2898
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45948361273294575,
+      "learning_rate": 9.384316613682735e-07,
+      "loss": 0.2599,
+      "step": 2899
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3149086661546606,
+      "learning_rate": 9.38382004137197e-07,
+      "loss": 0.129,
+      "step": 2900
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6401148842503128,
+      "learning_rate": 9.383323282038631e-07,
+      "loss": 0.2047,
+      "step": 2901
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1647592754413674,
+      "learning_rate": 9.382826335703908e-07,
+      "loss": 0.1428,
+      "step": 2902
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.349040834073399,
+      "learning_rate": 9.382329202389003e-07,
+      "loss": 0.2031,
+      "step": 2903
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.80803777930169,
+      "learning_rate": 9.381831882115126e-07,
+      "loss": 0.0875,
+      "step": 2904
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6901241070061314,
+      "learning_rate": 9.381334374903491e-07,
+      "loss": 0.3223,
+      "step": 2905
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.44340982598308587,
+      "learning_rate": 9.380836680775324e-07,
+      "loss": 0.1118,
+      "step": 2906
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.36681370962138576,
+      "learning_rate": 9.380338799751858e-07,
+      "loss": 0.3782,
+      "step": 2907
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6694382343922336,
+      "learning_rate": 9.379840731854334e-07,
+      "loss": 0.3132,
+      "step": 2908
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7581315571969631,
+      "learning_rate": 9.379342477103998e-07,
+      "loss": 0.1568,
+      "step": 2909
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.41683816538272617,
+      "learning_rate": 9.37884403552211e-07,
+      "loss": 0.2433,
+      "step": 2910
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0928002264705523,
+      "learning_rate": 9.378345407129931e-07,
+      "loss": 0.356,
+      "step": 2911
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6464702572252623,
+      "learning_rate": 9.377846591948737e-07,
+      "loss": 0.2987,
+      "step": 2912
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.213944566539613,
+      "learning_rate": 9.377347589999806e-07,
+      "loss": 0.4899,
+      "step": 2913
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2569801615400901,
+      "learning_rate": 9.376848401304428e-07,
+      "loss": 0.215,
+      "step": 2914
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7564381394785538,
+      "learning_rate": 9.376349025883899e-07,
+      "loss": 0.1616,
+      "step": 2915
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9610078231646705,
+      "learning_rate": 9.375849463759522e-07,
+      "loss": 0.1232,
+      "step": 2916
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3820691631100241,
+      "learning_rate": 9.375349714952609e-07,
+      "loss": 0.0886,
+      "step": 2917
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5951588990461749,
+      "learning_rate": 9.374849779484484e-07,
+      "loss": 0.2954,
+      "step": 2918
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.44345965442842933,
+      "learning_rate": 9.374349657376472e-07,
+      "loss": 0.2337,
+      "step": 2919
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.47022348751950854,
+      "learning_rate": 9.373849348649909e-07,
+      "loss": 0.0834,
+      "step": 2920
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5549102258857699,
+      "learning_rate": 9.373348853326142e-07,
+      "loss": 0.3269,
+      "step": 2921
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3390363419600777,
+      "learning_rate": 9.372848171426522e-07,
+      "loss": 0.0601,
+      "step": 2922
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4553715406375695,
+      "learning_rate": 9.372347302972407e-07,
+      "loss": 0.1892,
+      "step": 2923
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.315082537773244,
+      "learning_rate": 9.371846247985166e-07,
+      "loss": 0.1059,
+      "step": 2924
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7620149541053305,
+      "learning_rate": 9.371345006486176e-07,
+      "loss": 0.1488,
+      "step": 2925
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4819073686687152,
+      "learning_rate": 9.37084357849682e-07,
+      "loss": 0.2861,
+      "step": 2926
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7497531244440875,
+      "learning_rate": 9.370341964038492e-07,
+      "loss": 0.1649,
+      "step": 2927
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9647720741092505,
+      "learning_rate": 9.36984016313259e-07,
+      "loss": 0.4131,
+      "step": 2928
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7008518156276032,
+      "learning_rate": 9.369338175800521e-07,
+      "loss": 0.3531,
+      "step": 2929
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5904076191224823,
+      "learning_rate": 9.368836002063703e-07,
+      "loss": 0.2638,
+      "step": 2930
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1396464800581088,
+      "learning_rate": 9.368333641943558e-07,
+      "loss": 0.0155,
+      "step": 2931
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2158797865005662,
+      "learning_rate": 9.367831095461518e-07,
+      "loss": 0.4174,
+      "step": 2932
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.42457405895374317,
+      "learning_rate": 9.367328362639024e-07,
+      "loss": 0.0814,
+      "step": 2933
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1301786172208941,
+      "learning_rate": 9.366825443497522e-07,
+      "loss": 0.008,
+      "step": 2934
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9093203437447465,
+      "learning_rate": 9.366322338058469e-07,
+      "loss": 0.4459,
+      "step": 2935
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.49447847109671217,
+      "learning_rate": 9.365819046343328e-07,
+      "loss": 0.1887,
+      "step": 2936
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0850813480109034,
+      "learning_rate": 9.365315568373568e-07,
+      "loss": 0.0585,
+      "step": 2937
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4610127694846558,
+      "learning_rate": 9.364811904170672e-07,
+      "loss": 0.0876,
+      "step": 2938
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2660699869751306,
+      "learning_rate": 9.364308053756126e-07,
+      "loss": 0.1479,
+      "step": 2939
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5483732666481342,
+      "learning_rate": 9.363804017151424e-07,
+      "loss": 0.1685,
+      "step": 2940
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3343442458788976,
+      "learning_rate": 9.363299794378071e-07,
+      "loss": 0.0201,
+      "step": 2941
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.001151416510769,
+      "learning_rate": 9.362795385457578e-07,
+      "loss": 0.2791,
+      "step": 2942
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8030334382420369,
+      "learning_rate": 9.362290790411463e-07,
+      "loss": 0.1531,
+      "step": 2943
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2317266383697126,
+      "learning_rate": 9.361786009261252e-07,
+      "loss": 0.0073,
+      "step": 2944
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.047789694645446,
+      "learning_rate": 9.361281042028484e-07,
+      "loss": 0.2145,
+      "step": 2945
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4685516692379248,
+      "learning_rate": 9.360775888734697e-07,
+      "loss": 0.3019,
+      "step": 2946
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7773936351369717,
+      "learning_rate": 9.360270549401445e-07,
+      "loss": 0.3539,
+      "step": 2947
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5734200535640486,
+      "learning_rate": 9.359765024050288e-07,
+      "loss": 0.1606,
+      "step": 2948
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7189535865509294,
+      "learning_rate": 9.35925931270279e-07,
+      "loss": 0.4715,
+      "step": 2949
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3679596216667496,
+      "learning_rate": 9.358753415380527e-07,
+      "loss": 0.0498,
+      "step": 2950
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1985638110274803,
+      "learning_rate": 9.35824733210508e-07,
+      "loss": 0.2713,
+      "step": 2951
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.41398293651477885,
+      "learning_rate": 9.35774106289804e-07,
+      "loss": 0.2053,
+      "step": 2952
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6099238968783113,
+      "learning_rate": 9.357234607781008e-07,
+      "loss": 0.1821,
+      "step": 2953
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8709531253950371,
+      "learning_rate": 9.356727966775587e-07,
+      "loss": 0.3359,
+      "step": 2954
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2278086119851056,
+      "learning_rate": 9.356221139903393e-07,
+      "loss": 0.005,
+      "step": 2955
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9272032589655799,
+      "learning_rate": 9.355714127186048e-07,
+      "loss": 0.3826,
+      "step": 2956
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0236446162888828,
+      "learning_rate": 9.355206928645183e-07,
+      "loss": 0.2564,
+      "step": 2957
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8801309550772775,
+      "learning_rate": 9.354699544302435e-07,
+      "loss": 0.272,
+      "step": 2958
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7152614180118375,
+      "learning_rate": 9.354191974179451e-07,
+      "loss": 0.1935,
+      "step": 2959
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.242140104161178,
+      "learning_rate": 9.353684218297884e-07,
+      "loss": 0.1313,
+      "step": 2960
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4786856065967177,
+      "learning_rate": 9.353176276679395e-07,
+      "loss": 0.1123,
+      "step": 2961
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6804098926691428,
+      "learning_rate": 9.352668149345657e-07,
+      "loss": 0.3683,
+      "step": 2962
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3810212011607888,
+      "learning_rate": 9.352159836318345e-07,
+      "loss": 0.3455,
+      "step": 2963
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5196253006025056,
+      "learning_rate": 9.351651337619144e-07,
+      "loss": 0.2195,
+      "step": 2964
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7831762124216888,
+      "learning_rate": 9.351142653269752e-07,
+      "loss": 0.1341,
+      "step": 2965
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.401533870177438,
+      "learning_rate": 9.350633783291866e-07,
+      "loss": 0.0186,
+      "step": 2966
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2830055046461243,
+      "learning_rate": 9.350124727707196e-07,
+      "loss": 0.0125,
+      "step": 2967
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9275444900338184,
+      "learning_rate": 9.349615486537461e-07,
+      "loss": 0.1766,
+      "step": 2968
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4813644088477882,
+      "learning_rate": 9.349106059804386e-07,
+      "loss": 0.2933,
+      "step": 2969
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5543308751412848,
+      "learning_rate": 9.348596447529702e-07,
+      "loss": 0.3231,
+      "step": 2970
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3809371773300341,
+      "learning_rate": 9.348086649735154e-07,
+      "loss": 0.086,
+      "step": 2971
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1614202413974644,
+      "learning_rate": 9.347576666442487e-07,
+      "loss": 0.1066,
+      "step": 2972
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5684829679157605,
+      "learning_rate": 9.347066497673461e-07,
+      "loss": 0.0861,
+      "step": 2973
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8791000751341507,
+      "learning_rate": 9.34655614344984e-07,
+      "loss": 0.2266,
+      "step": 2974
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5369128656478885,
+      "learning_rate": 9.346045603793394e-07,
+      "loss": 0.1987,
+      "step": 2975
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2612040497881984,
+      "learning_rate": 9.345534878725907e-07,
+      "loss": 0.0356,
+      "step": 2976
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5417937297154061,
+      "learning_rate": 9.345023968269167e-07,
+      "loss": 0.1972,
+      "step": 2977
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.427665547698021,
+      "learning_rate": 9.344512872444969e-07,
+      "loss": 0.2863,
+      "step": 2978
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0737672201175752,
+      "learning_rate": 9.344001591275119e-07,
+      "loss": 0.1526,
+      "step": 2979
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7145115557551948,
+      "learning_rate": 9.343490124781428e-07,
+      "loss": 0.3481,
+      "step": 2980
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.119609581908684,
+      "learning_rate": 9.342978472985718e-07,
+      "loss": 0.2666,
+      "step": 2981
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7464592130899599,
+      "learning_rate": 9.342466635909815e-07,
+      "loss": 0.2481,
+      "step": 2982
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.12397808819221302,
+      "learning_rate": 9.341954613575555e-07,
+      "loss": 0.0045,
+      "step": 2983
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.40409144562835825,
+      "learning_rate": 9.341442406004784e-07,
+      "loss": 0.1173,
+      "step": 2984
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3297218341450486,
+      "learning_rate": 9.340930013219352e-07,
+      "loss": 0.0489,
+      "step": 2985
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.382178387832423,
+      "learning_rate": 9.340417435241119e-07,
+      "loss": 0.0356,
+      "step": 2986
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7806082011581832,
+      "learning_rate": 9.339904672091953e-07,
+      "loss": 0.1423,
+      "step": 2987
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5177359123813741,
+      "learning_rate": 9.33939172379373e-07,
+      "loss": 0.2016,
+      "step": 2988
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.31598636718736156,
+      "learning_rate": 9.338878590368333e-07,
+      "loss": 0.076,
+      "step": 2989
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3549764100792365,
+      "learning_rate": 9.338365271837654e-07,
+      "loss": 0.1827,
+      "step": 2990
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1013235744070637,
+      "learning_rate": 9.337851768223588e-07,
+      "loss": 0.3258,
+      "step": 2991
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.772870087027144,
+      "learning_rate": 9.337338079548048e-07,
+      "loss": 0.4289,
+      "step": 2992
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5294229312283212,
+      "learning_rate": 9.336824205832947e-07,
+      "loss": 0.2537,
+      "step": 2993
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1287496183325993,
+      "learning_rate": 9.336310147100205e-07,
+      "loss": 0.1337,
+      "step": 2994
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.44638174383621354,
+      "learning_rate": 9.335795903371755e-07,
+      "loss": 0.2678,
+      "step": 2995
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.593454638131262,
+      "learning_rate": 9.335281474669538e-07,
+      "loss": 0.2518,
+      "step": 2996
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.44976684179108467,
+      "learning_rate": 9.334766861015496e-07,
+      "loss": 0.0079,
+      "step": 2997
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4337280332740164,
+      "learning_rate": 9.334252062431587e-07,
+      "loss": 0.3132,
+      "step": 2998
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2867560830291846,
+      "learning_rate": 9.333737078939772e-07,
+      "loss": 0.1556,
+      "step": 2999
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4448942958304109,
+      "learning_rate": 9.333221910562022e-07,
+      "loss": 0.1751,
+      "step": 3000
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.263005784438112,
+      "learning_rate": 9.332706557320314e-07,
+      "loss": 0.1779,
+      "step": 3001
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5466353098286592,
+      "learning_rate": 9.332191019236632e-07,
+      "loss": 0.099,
+      "step": 3002
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7976615097556736,
+      "learning_rate": 9.331675296332975e-07,
+      "loss": 0.1148,
+      "step": 3003
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8784737078696657,
+      "learning_rate": 9.331159388631341e-07,
+      "loss": 0.2069,
+      "step": 3004
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.535056162166755,
+      "learning_rate": 9.330643296153742e-07,
+      "loss": 0.1783,
+      "step": 3005
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.239662280674292,
+      "learning_rate": 9.330127018922193e-07,
+      "loss": 0.1772,
+      "step": 3006
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4329142584792107,
+      "learning_rate": 9.329610556958722e-07,
+      "loss": 0.2907,
+      "step": 3007
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.28893600436723366,
+      "learning_rate": 9.32909391028536e-07,
+      "loss": 0.1061,
+      "step": 3008
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6270169813375313,
+      "learning_rate": 9.32857707892415e-07,
+      "loss": 0.1671,
+      "step": 3009
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7369538297638233,
+      "learning_rate": 9.328060062897138e-07,
+      "loss": 0.0758,
+      "step": 3010
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4241868162913618,
+      "learning_rate": 9.327542862226386e-07,
+      "loss": 0.1742,
+      "step": 3011
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.34388789875246983,
+      "learning_rate": 9.327025476933954e-07,
+      "loss": 0.0318,
+      "step": 3012
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5254738570119284,
+      "learning_rate": 9.326507907041918e-07,
+      "loss": 0.067,
+      "step": 3013
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3090281030578117,
+      "learning_rate": 9.325990152572358e-07,
+      "loss": 0.1806,
+      "step": 3014
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.573799180637348,
+      "learning_rate": 9.32547221354736e-07,
+      "loss": 0.1102,
+      "step": 3015
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5302728596888309,
+      "learning_rate": 9.324954089989023e-07,
+      "loss": 0.2121,
+      "step": 3016
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.33753031745126294,
+      "learning_rate": 9.324435781919449e-07,
+      "loss": 0.1485,
+      "step": 3017
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.998994753179129,
+      "learning_rate": 9.323917289360753e-07,
+      "loss": 0.2142,
+      "step": 3018
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.19605758441521368,
+      "learning_rate": 9.323398612335054e-07,
+      "loss": 0.0065,
+      "step": 3019
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4283434558037134,
+      "learning_rate": 9.322879750864476e-07,
+      "loss": 0.0639,
+      "step": 3020
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8714663833200217,
+      "learning_rate": 9.32236070497116e-07,
+      "loss": 0.2954,
+      "step": 3021
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.858049201538456,
+      "learning_rate": 9.321841474677247e-07,
+      "loss": 0.1448,
+      "step": 3022
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1389143844340983,
+      "learning_rate": 9.321322060004888e-07,
+      "loss": 0.3223,
+      "step": 3023
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5234024329695265,
+      "learning_rate": 9.320802460976245e-07,
+      "loss": 0.1825,
+      "step": 3024
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9722075998628446,
+      "learning_rate": 9.32028267761348e-07,
+      "loss": 0.2889,
+      "step": 3025
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9437832253730604,
+      "learning_rate": 9.319762709938775e-07,
+      "loss": 0.3222,
+      "step": 3026
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.830519395355223,
+      "learning_rate": 9.319242557974305e-07,
+      "loss": 0.3216,
+      "step": 3027
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.46865160802583294,
+      "learning_rate": 9.318722221742267e-07,
+      "loss": 0.1342,
+      "step": 3028
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7704922069197175,
+      "learning_rate": 9.318201701264857e-07,
+      "loss": 0.2531,
+      "step": 3029
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9057929614247713,
+      "learning_rate": 9.317680996564281e-07,
+      "loss": 0.1031,
+      "step": 3030
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6183863786282269,
+      "learning_rate": 9.317160107662754e-07,
+      "loss": 0.0579,
+      "step": 3031
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6514179057007335,
+      "learning_rate": 9.316639034582498e-07,
+      "loss": 0.1904,
+      "step": 3032
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2777188197898017,
+      "learning_rate": 9.316117777345746e-07,
+      "loss": 0.036,
+      "step": 3033
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7732503085369049,
+      "learning_rate": 9.315596335974731e-07,
+      "loss": 0.4658,
+      "step": 3034
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8356075156372184,
+      "learning_rate": 9.3150747104917e-07,
+      "loss": 0.2359,
+      "step": 3035
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4573641176035874,
+      "learning_rate": 9.314552900918908e-07,
+      "loss": 0.0818,
+      "step": 3036
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.085807918386808,
+      "learning_rate": 9.314030907278618e-07,
+      "loss": 0.4449,
+      "step": 3037
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8291466502959902,
+      "learning_rate": 9.313508729593096e-07,
+      "loss": 0.2132,
+      "step": 3038
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2470427402668394,
+      "learning_rate": 9.312986367884619e-07,
+      "loss": 0.1351,
+      "step": 3039
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8357623921356458,
+      "learning_rate": 9.312463822175474e-07,
+      "loss": 0.2655,
+      "step": 3040
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6289347384014446,
+      "learning_rate": 9.311941092487954e-07,
+      "loss": 0.1375,
+      "step": 3041
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6852238162700425,
+      "learning_rate": 9.31141817884436e-07,
+      "loss": 0.1956,
+      "step": 3042
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2071504666592814,
+      "learning_rate": 9.310895081266996e-07,
+      "loss": 0.4316,
+      "step": 3043
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.672865138314787,
+      "learning_rate": 9.310371799778184e-07,
+      "loss": 0.1289,
+      "step": 3044
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.15820142374764,
+      "learning_rate": 9.309848334400245e-07,
+      "loss": 0.1639,
+      "step": 3045
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5501107913660684,
+      "learning_rate": 9.309324685155513e-07,
+      "loss": 0.2278,
+      "step": 3046
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8376405818747217,
+      "learning_rate": 9.308800852066328e-07,
+      "loss": 0.2784,
+      "step": 3047
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8444680283046364,
+      "learning_rate": 9.308276835155036e-07,
+      "loss": 0.1992,
+      "step": 3048
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9159911788834041,
+      "learning_rate": 9.307752634443992e-07,
+      "loss": 0.3426,
+      "step": 3049
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2841327552971487,
+      "learning_rate": 9.307228249955563e-07,
+      "loss": 0.2586,
+      "step": 3050
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3766795409358213,
+      "learning_rate": 9.306703681712118e-07,
+      "loss": 0.2657,
+      "step": 3051
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6124945721716994,
+      "learning_rate": 9.306178929736037e-07,
+      "loss": 0.1604,
+      "step": 3052
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0815748384024368,
+      "learning_rate": 9.305653994049705e-07,
+      "loss": 0.2442,
+      "step": 3053
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2882148876130204,
+      "learning_rate": 9.305128874675519e-07,
+      "loss": 0.1155,
+      "step": 3054
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3607427777409394,
+      "learning_rate": 9.304603571635879e-07,
+      "loss": 0.2395,
+      "step": 3055
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6693896201291227,
+      "learning_rate": 9.3040780849532e-07,
+      "loss": 0.2456,
+      "step": 3056
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.31469024781178107,
+      "learning_rate": 9.303552414649896e-07,
+      "loss": 0.1758,
+      "step": 3057
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4505317171746773,
+      "learning_rate": 9.303026560748395e-07,
+      "loss": 0.297,
+      "step": 3058
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6597112278091493,
+      "learning_rate": 9.302500523271131e-07,
+      "loss": 0.1577,
+      "step": 3059
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7323757246455531,
+      "learning_rate": 9.301974302240545e-07,
+      "loss": 0.2976,
+      "step": 3060
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7159980576059044,
+      "learning_rate": 9.301447897679087e-07,
+      "loss": 0.2212,
+      "step": 3061
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6263325445076474,
+      "learning_rate": 9.300921309609215e-07,
+      "loss": 0.2819,
+      "step": 3062
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7504126780496241,
+      "learning_rate": 9.300394538053394e-07,
+      "loss": 0.203,
+      "step": 3063
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8346360508181327,
+      "learning_rate": 9.299867583034098e-07,
+      "loss": 0.2596,
+      "step": 3064
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.296066955293461,
+      "learning_rate": 9.299340444573807e-07,
+      "loss": 0.1294,
+      "step": 3065
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7019518256264955,
+      "learning_rate": 9.298813122695009e-07,
+      "loss": 0.1725,
+      "step": 3066
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7518499734295849,
+      "learning_rate": 9.298285617420202e-07,
+      "loss": 0.1102,
+      "step": 3067
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3364157984802234,
+      "learning_rate": 9.29775792877189e-07,
+      "loss": 0.1319,
+      "step": 3068
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.33961466118762706,
+      "learning_rate": 9.297230056772585e-07,
+      "loss": 0.191,
+      "step": 3069
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5594391802486758,
+      "learning_rate": 9.296702001444807e-07,
+      "loss": 0.1917,
+      "step": 3070
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2721794436919729,
+      "learning_rate": 9.296173762811084e-07,
+      "loss": 0.3695,
+      "step": 3071
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9533258107642867,
+      "learning_rate": 9.295645340893952e-07,
+      "loss": 0.2842,
+      "step": 3072
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.073891511275848,
+      "learning_rate": 9.295116735715955e-07,
+      "loss": 0.518,
+      "step": 3073
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9572885740850663,
+      "learning_rate": 9.294587947299644e-07,
+      "loss": 0.0734,
+      "step": 3074
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2986012575020074,
+      "learning_rate": 9.294058975667575e-07,
+      "loss": 0.3546,
+      "step": 3075
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4913653469206503,
+      "learning_rate": 9.293529820842322e-07,
+      "loss": 0.1909,
+      "step": 3076
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7482298590180433,
+      "learning_rate": 9.293000482846453e-07,
+      "loss": 0.141,
+      "step": 3077
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4274984032690888,
+      "learning_rate": 9.292470961702555e-07,
+      "loss": 0.273,
+      "step": 3078
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0737864233379126,
+      "learning_rate": 9.291941257433217e-07,
+      "loss": 0.1208,
+      "step": 3079
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6277167426847763,
+      "learning_rate": 9.291411370061036e-07,
+      "loss": 0.314,
+      "step": 3080
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.490536857820418,
+      "learning_rate": 9.29088129960862e-07,
+      "loss": 0.1106,
+      "step": 3081
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9631473008227158,
+      "learning_rate": 9.290351046098581e-07,
+      "loss": 0.1021,
+      "step": 3082
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.040663390524604,
+      "learning_rate": 9.289820609553542e-07,
+      "loss": 0.0639,
+      "step": 3083
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5977638524386644,
+      "learning_rate": 9.289289989996132e-07,
+      "loss": 0.2086,
+      "step": 3084
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7715634833014068,
+      "learning_rate": 9.288759187448989e-07,
+      "loss": 0.1186,
+      "step": 3085
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0831515958975575,
+      "learning_rate": 9.288228201934758e-07,
+      "loss": 0.1367,
+      "step": 3086
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.59423851730006,
+      "learning_rate": 9.287697033476091e-07,
+      "loss": 0.4256,
+      "step": 3087
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4208760634114546,
+      "learning_rate": 9.287165682095649e-07,
+      "loss": 0.3848,
+      "step": 3088
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.23706709793319297,
+      "learning_rate": 9.286634147816102e-07,
+      "loss": 0.2232,
+      "step": 3089
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3404176398453622,
+      "learning_rate": 9.286102430660123e-07,
+      "loss": 0.2283,
+      "step": 3090
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1211225236398437,
+      "learning_rate": 9.285570530650399e-07,
+      "loss": 0.2886,
+      "step": 3091
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8640493538302704,
+      "learning_rate": 9.285038447809621e-07,
+      "loss": 0.1125,
+      "step": 3092
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7734970842558655,
+      "learning_rate": 9.284506182160489e-07,
+      "loss": 0.2469,
+      "step": 3093
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6817783569138928,
+      "learning_rate": 9.283973733725709e-07,
+      "loss": 0.0149,
+      "step": 3094
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.746164700972667,
+      "learning_rate": 9.283441102528e-07,
+      "loss": 0.4098,
+      "step": 3095
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6777464618768618,
+      "learning_rate": 9.282908288590082e-07,
+      "loss": 0.1893,
+      "step": 3096
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.631082316411457,
+      "learning_rate": 9.282375291934685e-07,
+      "loss": 0.1531,
+      "step": 3097
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4404346136544487,
+      "learning_rate": 9.281842112584552e-07,
+      "loss": 0.0854,
+      "step": 3098
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6649865624253605,
+      "learning_rate": 9.281308750562425e-07,
+      "loss": 0.1809,
+      "step": 3099
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.553830471140561,
+      "learning_rate": 9.280775205891062e-07,
+      "loss": 0.2709,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.787758787758307,
+      "learning_rate": 9.280241478593221e-07,
+      "loss": 0.31,
+      "step": 3101
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36627677398165226,
+      "learning_rate": 9.279707568691676e-07,
+      "loss": 0.2286,
+      "step": 3102
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5283829688432854,
+      "learning_rate": 9.279173476209202e-07,
+      "loss": 0.3185,
+      "step": 3103
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6095925885429754,
+      "learning_rate": 9.278639201168585e-07,
+      "loss": 0.0636,
+      "step": 3104
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4672601309115711,
+      "learning_rate": 9.27810474359262e-07,
+      "loss": 0.2262,
+      "step": 3105
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8983970054704082,
+      "learning_rate": 9.277570103504104e-07,
+      "loss": 0.1639,
+      "step": 3106
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6342052380139972,
+      "learning_rate": 9.277035280925852e-07,
+      "loss": 0.1156,
+      "step": 3107
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0159746958773266,
+      "learning_rate": 9.276500275880675e-07,
+      "loss": 0.2254,
+      "step": 3108
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.431654174276687,
+      "learning_rate": 9.275965088391397e-07,
+      "loss": 0.2284,
+      "step": 3109
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1935745659077637,
+      "learning_rate": 9.275429718480856e-07,
+      "loss": 0.1168,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4555984205316003,
+      "learning_rate": 9.274894166171887e-07,
+      "loss": 0.06,
+      "step": 3111
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1012201002196862,
+      "learning_rate": 9.274358431487339e-07,
+      "loss": 0.2859,
+      "step": 3112
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9332877898438722,
+      "learning_rate": 9.27382251445007e-07,
+      "loss": 0.2569,
+      "step": 3113
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7874397460911885,
+      "learning_rate": 9.273286415082939e-07,
+      "loss": 0.3878,
+      "step": 3114
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7959578824051317,
+      "learning_rate": 9.272750133408819e-07,
+      "loss": 0.1595,
+      "step": 3115
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.46285371181947127,
+      "learning_rate": 9.27221366945059e-07,
+      "loss": 0.1919,
+      "step": 3116
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1482722745697806,
+      "learning_rate": 9.271677023231137e-07,
+      "loss": 0.2325,
+      "step": 3117
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7778395195759555,
+      "learning_rate": 9.271140194773355e-07,
+      "loss": 0.3524,
+      "step": 3118
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5775770200927671,
+      "learning_rate": 9.270603184100148e-07,
+      "loss": 0.2144,
+      "step": 3119
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.857440018690515,
+      "learning_rate": 9.270065991234421e-07,
+      "loss": 0.2713,
+      "step": 3120
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5705712224451414,
+      "learning_rate": 9.269528616199097e-07,
+      "loss": 0.2801,
+      "step": 3121
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4706964326756469,
+      "learning_rate": 9.2689910590171e-07,
+      "loss": 0.0459,
+      "step": 3122
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.42521057349481917,
+      "learning_rate": 9.268453319711362e-07,
+      "loss": 0.1089,
+      "step": 3123
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.40657016370441595,
+      "learning_rate": 9.267915398304823e-07,
+      "loss": 0.1521,
+      "step": 3124
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.46261893384349917,
+      "learning_rate": 9.267377294820435e-07,
+      "loss": 0.1424,
+      "step": 3125
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6833032160213107,
+      "learning_rate": 9.266839009281153e-07,
+      "loss": 0.1854,
+      "step": 3126
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7736416060711386,
+      "learning_rate": 9.266300541709942e-07,
+      "loss": 0.3294,
+      "step": 3127
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2219178648628992,
+      "learning_rate": 9.265761892129773e-07,
+      "loss": 0.1536,
+      "step": 3128
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6033314381868828,
+      "learning_rate": 9.265223060563626e-07,
+      "loss": 0.1082,
+      "step": 3129
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.262137011317028,
+      "learning_rate": 9.26468404703449e-07,
+      "loss": 0.1726,
+      "step": 3130
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7195111645442679,
+      "learning_rate": 9.264144851565359e-07,
+      "loss": 0.1617,
+      "step": 3131
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3230029545712064,
+      "learning_rate": 9.263605474179237e-07,
+      "loss": 0.0923,
+      "step": 3132
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.956611115674006,
+      "learning_rate": 9.263065914899133e-07,
+      "loss": 0.3059,
+      "step": 3133
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.359438984188222,
+      "learning_rate": 9.262526173748069e-07,
+      "loss": 0.0781,
+      "step": 3134
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36141643563992937,
+      "learning_rate": 9.261986250749067e-07,
+      "loss": 0.1543,
+      "step": 3135
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9103019185795369,
+      "learning_rate": 9.261446145925167e-07,
+      "loss": 0.2738,
+      "step": 3136
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3150390424059855,
+      "learning_rate": 9.260905859299407e-07,
+      "loss": 0.3519,
+      "step": 3137
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4051852584242612,
+      "learning_rate": 9.260365390894837e-07,
+      "loss": 0.0811,
+      "step": 3138
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8460199513689234,
+      "learning_rate": 9.259824740734516e-07,
+      "loss": 0.122,
+      "step": 3139
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9071104525547069,
+      "learning_rate": 9.259283908841506e-07,
+      "loss": 0.3743,
+      "step": 3140
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9524789406700434,
+      "learning_rate": 9.258742895238885e-07,
+      "loss": 0.1818,
+      "step": 3141
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6664822300195525,
+      "learning_rate": 9.25820169994973e-07,
+      "loss": 0.1822,
+      "step": 3142
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6172938015588753,
+      "learning_rate": 9.25766032299713e-07,
+      "loss": 0.2441,
+      "step": 3143
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.849651525952206,
+      "learning_rate": 9.257118764404181e-07,
+      "loss": 0.107,
+      "step": 3144
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7729040587341335,
+      "learning_rate": 9.256577024193989e-07,
+      "loss": 0.3767,
+      "step": 3145
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6564250188192128,
+      "learning_rate": 9.256035102389663e-07,
+      "loss": 0.0797,
+      "step": 3146
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4430327980532014,
+      "learning_rate": 9.255492999014324e-07,
+      "loss": 0.2038,
+      "step": 3147
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6259749299365427,
+      "learning_rate": 9.254950714091101e-07,
+      "loss": 0.0472,
+      "step": 3148
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.575732959471976,
+      "learning_rate": 9.254408247643125e-07,
+      "loss": 0.123,
+      "step": 3149
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7782952435832244,
+      "learning_rate": 9.253865599693543e-07,
+      "loss": 0.1398,
+      "step": 3150
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5640252414464564,
+      "learning_rate": 9.253322770265501e-07,
+      "loss": 0.1967,
+      "step": 3151
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6524550108989349,
+      "learning_rate": 9.252779759382158e-07,
+      "loss": 0.268,
+      "step": 3152
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.640403865390597,
+      "learning_rate": 9.252236567066685e-07,
+      "loss": 0.2419,
+      "step": 3153
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1765964313625707,
+      "learning_rate": 9.251693193342249e-07,
+      "loss": 0.3511,
+      "step": 3154
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7874003331630535,
+      "learning_rate": 9.251149638232037e-07,
+      "loss": 0.1543,
+      "step": 3155
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1031380268003386,
+      "learning_rate": 9.250605901759233e-07,
+      "loss": 0.2093,
+      "step": 3156
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7965048234082887,
+      "learning_rate": 9.250061983947038e-07,
+      "loss": 0.3321,
+      "step": 3157
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6196023978433362,
+      "learning_rate": 9.249517884818654e-07,
+      "loss": 0.4197,
+      "step": 3158
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5654349449389893,
+      "learning_rate": 9.248973604397295e-07,
+      "loss": 0.1719,
+      "step": 3159
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.698012012240679,
+      "learning_rate": 9.248429142706181e-07,
+      "loss": 0.2616,
+      "step": 3160
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2680825715026603,
+      "learning_rate": 9.247884499768539e-07,
+      "loss": 0.1239,
+      "step": 3161
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4076732361684476,
+      "learning_rate": 9.247339675607605e-07,
+      "loss": 0.1892,
+      "step": 3162
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3428337471960096,
+      "learning_rate": 9.246794670246623e-07,
+      "loss": 0.0122,
+      "step": 3163
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6622795978962026,
+      "learning_rate": 9.246249483708842e-07,
+      "loss": 0.4124,
+      "step": 3164
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.43485167127355784,
+      "learning_rate": 9.24570411601752e-07,
+      "loss": 0.1117,
+      "step": 3165
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5284284378374157,
+      "learning_rate": 9.24515856719593e-07,
+      "loss": 0.1605,
+      "step": 3166
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.451271690731251,
+      "learning_rate": 9.244612837267338e-07,
+      "loss": 0.1461,
+      "step": 3167
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.41283869920204125,
+      "learning_rate": 9.244066926255031e-07,
+      "loss": 0.0468,
+      "step": 3168
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.859544709261444,
+      "learning_rate": 9.243520834182297e-07,
+      "loss": 0.0915,
+      "step": 3169
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5412777735384318,
+      "learning_rate": 9.242974561072436e-07,
+      "loss": 0.1003,
+      "step": 3170
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8852762174299448,
+      "learning_rate": 9.242428106948748e-07,
+      "loss": 0.035,
+      "step": 3171
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.46427628358900325,
+      "learning_rate": 9.241881471834549e-07,
+      "loss": 0.2438,
+      "step": 3172
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3303918204521397,
+      "learning_rate": 9.24133465575316e-07,
+      "loss": 0.1165,
+      "step": 3173
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3878750718711695,
+      "learning_rate": 9.240787658727909e-07,
+      "loss": 0.3216,
+      "step": 3174
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5824463727444047,
+      "learning_rate": 9.240240480782129e-07,
+      "loss": 0.1372,
+      "step": 3175
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4037450558744446,
+      "learning_rate": 9.239693121939168e-07,
+      "loss": 0.1651,
+      "step": 3176
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8191030154457623,
+      "learning_rate": 9.239145582222376e-07,
+      "loss": 0.2761,
+      "step": 3177
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4330613171027677,
+      "learning_rate": 9.238597861655111e-07,
+      "loss": 0.1141,
+      "step": 3178
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9178433176483896,
+      "learning_rate": 9.23804996026074e-07,
+      "loss": 0.2528,
+      "step": 3179
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8536592757608799,
+      "learning_rate": 9.237501878062638e-07,
+      "loss": 0.4177,
+      "step": 3180
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5320097823585402,
+      "learning_rate": 9.236953615084189e-07,
+      "loss": 0.3115,
+      "step": 3181
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5953625757265195,
+      "learning_rate": 9.23640517134878e-07,
+      "loss": 0.1737,
+      "step": 3182
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8007362504841291,
+      "learning_rate": 9.23585654687981e-07,
+      "loss": 0.0391,
+      "step": 3183
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8458625569678068,
+      "learning_rate": 9.235307741700685e-07,
+      "loss": 0.0691,
+      "step": 3184
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7845374611852582,
+      "learning_rate": 9.234758755834818e-07,
+      "loss": 0.4652,
+      "step": 3185
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9736760960263428,
+      "learning_rate": 9.234209589305629e-07,
+      "loss": 0.1573,
+      "step": 3186
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2100880068248985,
+      "learning_rate": 9.233660242136548e-07,
+      "loss": 0.2671,
+      "step": 3187
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8863558069875754,
+      "learning_rate": 9.233110714351009e-07,
+      "loss": 0.285,
+      "step": 3188
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.387485326428505,
+      "learning_rate": 9.232561005972459e-07,
+      "loss": 0.1589,
+      "step": 3189
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7961383788030786,
+      "learning_rate": 9.232011117024348e-07,
+      "loss": 0.2707,
+      "step": 3190
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9879379142928602,
+      "learning_rate": 9.231461047530133e-07,
+      "loss": 0.2961,
+      "step": 3191
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7928095490159558,
+      "learning_rate": 9.230910797513285e-07,
+      "loss": 0.143,
+      "step": 3192
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7905290172441409,
+      "learning_rate": 9.230360366997277e-07,
+      "loss": 0.3047,
+      "step": 3193
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.171167431695269,
+      "learning_rate": 9.229809756005592e-07,
+      "loss": 0.0411,
+      "step": 3194
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7363608158765037,
+      "learning_rate": 9.229258964561719e-07,
+      "loss": 0.2254,
+      "step": 3195
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.054144945273629,
+      "learning_rate": 9.228707992689157e-07,
+      "loss": 0.2566,
+      "step": 3196
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1728080823178616,
+      "learning_rate": 9.228156840411411e-07,
+      "loss": 0.0789,
+      "step": 3197
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4382911752143824,
+      "learning_rate": 9.227605507751997e-07,
+      "loss": 0.0144,
+      "step": 3198
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4021783476774723,
+      "learning_rate": 9.227053994734431e-07,
+      "loss": 0.4258,
+      "step": 3199
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.469696850201351,
+      "learning_rate": 9.226502301382244e-07,
+      "loss": 0.0046,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7784926479436306,
+      "learning_rate": 9.225950427718974e-07,
+      "loss": 0.187,
+      "step": 3201
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8830903774871972,
+      "learning_rate": 9.225398373768163e-07,
+      "loss": 0.2908,
+      "step": 3202
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7850095466288786,
+      "learning_rate": 9.224846139553362e-07,
+      "loss": 0.2986,
+      "step": 3203
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7171573984108145,
+      "learning_rate": 9.224293725098132e-07,
+      "loss": 0.136,
+      "step": 3204
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7147338692888996,
+      "learning_rate": 9.223741130426041e-07,
+      "loss": 0.2694,
+      "step": 3205
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5449525920143408,
+      "learning_rate": 9.223188355560663e-07,
+      "loss": 0.1357,
+      "step": 3206
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8683597603208748,
+      "learning_rate": 9.222635400525578e-07,
+      "loss": 0.0655,
+      "step": 3207
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7005967370533396,
+      "learning_rate": 9.22208226534438e-07,
+      "loss": 0.1175,
+      "step": 3208
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.43738091256188927,
+      "learning_rate": 9.221528950040663e-07,
+      "loss": 0.1073,
+      "step": 3209
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8581921996171491,
+      "learning_rate": 9.220975454638036e-07,
+      "loss": 0.081,
+      "step": 3210
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1351383388479153,
+      "learning_rate": 9.220421779160111e-07,
+      "loss": 0.0515,
+      "step": 3211
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.650205733208947,
+      "learning_rate": 9.219867923630509e-07,
+      "loss": 0.1263,
+      "step": 3212
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5094960839844335,
+      "learning_rate": 9.219313888072859e-07,
+      "loss": 0.1935,
+      "step": 3213
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3775566066303138,
+      "learning_rate": 9.218759672510795e-07,
+      "loss": 0.29,
+      "step": 3214
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5839181438936653,
+      "learning_rate": 9.218205276967963e-07,
+      "loss": 0.121,
+      "step": 3215
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.619965388979965,
+      "learning_rate": 9.217650701468015e-07,
+      "loss": 0.2328,
+      "step": 3216
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4594208800221896,
+      "learning_rate": 9.217095946034609e-07,
+      "loss": 0.1759,
+      "step": 3217
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.894595064160595,
+      "learning_rate": 9.216541010691412e-07,
+      "loss": 0.0834,
+      "step": 3218
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6950540322671729,
+      "learning_rate": 9.215985895462101e-07,
+      "loss": 0.0271,
+      "step": 3219
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7352186234647778,
+      "learning_rate": 9.215430600370356e-07,
+      "loss": 0.1694,
+      "step": 3220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.44343233299035845,
+      "learning_rate": 9.214875125439865e-07,
+      "loss": 0.1934,
+      "step": 3221
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4432183421572362,
+      "learning_rate": 9.21431947069433e-07,
+      "loss": 0.1698,
+      "step": 3222
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.48105287720482787,
+      "learning_rate": 9.213763636157454e-07,
+      "loss": 0.1607,
+      "step": 3223
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.31192982752598486,
+      "learning_rate": 9.213207621852952e-07,
+      "loss": 0.0758,
+      "step": 3224
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6726849577678,
+      "learning_rate": 9.212651427804543e-07,
+      "loss": 0.1775,
+      "step": 3225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6263180529191299,
+      "learning_rate": 9.212095054035955e-07,
+      "loss": 0.0706,
+      "step": 3226
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4635604648015689,
+      "learning_rate": 9.211538500570923e-07,
+      "loss": 0.1287,
+      "step": 3227
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8474045210307591,
+      "learning_rate": 9.210981767433195e-07,
+      "loss": 0.3819,
+      "step": 3228
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6874972538717036,
+      "learning_rate": 9.210424854646519e-07,
+      "loss": 0.2192,
+      "step": 3229
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7691094876067304,
+      "learning_rate": 9.209867762234653e-07,
+      "loss": 0.022,
+      "step": 3230
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.696244205564151,
+      "learning_rate": 9.209310490221367e-07,
+      "loss": 0.3037,
+      "step": 3231
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8153250419759999,
+      "learning_rate": 9.208753038630434e-07,
+      "loss": 0.1564,
+      "step": 3232
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9541435679121205,
+      "learning_rate": 9.208195407485634e-07,
+      "loss": 0.0973,
+      "step": 3233
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8509587442576104,
+      "learning_rate": 9.20763759681076e-07,
+      "loss": 0.1268,
+      "step": 3234
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2920801299379249,
+      "learning_rate": 9.207079606629606e-07,
+      "loss": 0.1542,
+      "step": 3235
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1822533011617806,
+      "learning_rate": 9.206521436965981e-07,
+      "loss": 0.1604,
+      "step": 3236
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0059318658786853,
+      "learning_rate": 9.205963087843693e-07,
+      "loss": 0.2971,
+      "step": 3237
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6283855641188356,
+      "learning_rate": 9.205404559286567e-07,
+      "loss": 0.2701,
+      "step": 3238
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6352247159513633,
+      "learning_rate": 9.204845851318427e-07,
+      "loss": 0.2549,
+      "step": 3239
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0524122518502468,
+      "learning_rate": 9.204286963963111e-07,
+      "loss": 0.3879,
+      "step": 3240
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9223261899788481,
+      "learning_rate": 9.20372789724446e-07,
+      "loss": 0.1511,
+      "step": 3241
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0375883721197088,
+      "learning_rate": 9.203168651186329e-07,
+      "loss": 0.1093,
+      "step": 3242
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6249258115181283,
+      "learning_rate": 9.202609225812571e-07,
+      "loss": 0.1495,
+      "step": 3243
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3723623613833438,
+      "learning_rate": 9.202049621147055e-07,
+      "loss": 0.2153,
+      "step": 3244
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.116965781580051,
+      "learning_rate": 9.201489837213658e-07,
+      "loss": 0.1362,
+      "step": 3245
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5384575646385588,
+      "learning_rate": 9.200929874036257e-07,
+      "loss": 0.2366,
+      "step": 3246
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0716566854155367,
+      "learning_rate": 9.200369731638741e-07,
+      "loss": 0.0693,
+      "step": 3247
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6249185693333995,
+      "learning_rate": 9.19980941004501e-07,
+      "loss": 0.1311,
+      "step": 3248
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7468108907428338,
+      "learning_rate": 9.199248909278967e-07,
+      "loss": 0.1688,
+      "step": 3249
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3635267021327973,
+      "learning_rate": 9.198688229364524e-07,
+      "loss": 0.0944,
+      "step": 3250
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.49514635687869285,
+      "learning_rate": 9.198127370325601e-07,
+      "loss": 0.1282,
+      "step": 3251
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5079375544928111,
+      "learning_rate": 9.197566332186124e-07,
+      "loss": 0.106,
+      "step": 3252
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4813235463590242,
+      "learning_rate": 9.197005114970029e-07,
+      "loss": 0.0962,
+      "step": 3253
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4454324149510626,
+      "learning_rate": 9.19644371870126e-07,
+      "loss": 0.2286,
+      "step": 3254
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4247545785031421,
+      "learning_rate": 9.195882143403766e-07,
+      "loss": 0.0118,
+      "step": 3255
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6139334715147486,
+      "learning_rate": 9.195320389101504e-07,
+      "loss": 0.3141,
+      "step": 3256
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8679100346942271,
+      "learning_rate": 9.194758455818441e-07,
+      "loss": 0.1987,
+      "step": 3257
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7583339385851732,
+      "learning_rate": 9.194196343578549e-07,
+      "loss": 0.4295,
+      "step": 3258
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5185716101109296,
+      "learning_rate": 9.193634052405811e-07,
+      "loss": 0.2579,
+      "step": 3259
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2887558954245897,
+      "learning_rate": 9.193071582324213e-07,
+      "loss": 0.0536,
+      "step": 3260
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7758585411009997,
+      "learning_rate": 9.192508933357752e-07,
+      "loss": 0.171,
+      "step": 3261
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2598361492938699,
+      "learning_rate": 9.191946105530432e-07,
+      "loss": 0.0823,
+      "step": 3262
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5276822194378585,
+      "learning_rate": 9.191383098866265e-07,
+      "loss": 0.1623,
+      "step": 3263
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6953044661041045,
+      "learning_rate": 9.190819913389269e-07,
+      "loss": 0.2203,
+      "step": 3264
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.27018655248726875,
+      "learning_rate": 9.190256549123471e-07,
+      "loss": 0.2548,
+      "step": 3265
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6395899427747753,
+      "learning_rate": 9.189693006092905e-07,
+      "loss": 0.0718,
+      "step": 3266
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.22490555307165516,
+      "learning_rate": 9.189129284321614e-07,
+      "loss": 0.0889,
+      "step": 3267
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9837568311853337,
+      "learning_rate": 9.188565383833647e-07,
+      "loss": 0.1677,
+      "step": 3268
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6026477433004154,
+      "learning_rate": 9.188001304653058e-07,
+      "loss": 0.1218,
+      "step": 3269
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8915853470145135,
+      "learning_rate": 9.187437046803915e-07,
+      "loss": 0.1409,
+      "step": 3270
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9289081429824383,
+      "learning_rate": 9.18687261031029e-07,
+      "loss": 0.2263,
+      "step": 3271
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7102800021580643,
+      "learning_rate": 9.186307995196263e-07,
+      "loss": 0.0423,
+      "step": 3272
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40979456304112044,
+      "learning_rate": 9.185743201485922e-07,
+      "loss": 0.2591,
+      "step": 3273
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8633246845478875,
+      "learning_rate": 9.185178229203361e-07,
+      "loss": 0.1852,
+      "step": 3274
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.36422496372365826,
+      "learning_rate": 9.184613078372685e-07,
+      "loss": 0.2195,
+      "step": 3275
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6511353010145003,
+      "learning_rate": 9.184047749018e-07,
+      "loss": 0.1936,
+      "step": 3276
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6264697924714663,
+      "learning_rate": 9.183482241163429e-07,
+      "loss": 0.1882,
+      "step": 3277
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6652060262657263,
+      "learning_rate": 9.182916554833096e-07,
+      "loss": 0.2467,
+      "step": 3278
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.013861147834192,
+      "learning_rate": 9.182350690051132e-07,
+      "loss": 0.2543,
+      "step": 3279
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5113679897931713,
+      "learning_rate": 9.181784646841683e-07,
+      "loss": 0.317,
+      "step": 3280
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5882751125353862,
+      "learning_rate": 9.181218425228893e-07,
+      "loss": 0.1704,
+      "step": 3281
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9396655086336284,
+      "learning_rate": 9.180652025236919e-07,
+      "loss": 0.1283,
+      "step": 3282
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40541933418651066,
+      "learning_rate": 9.180085446889927e-07,
+      "loss": 0.1214,
+      "step": 3283
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1760734602911802,
+      "learning_rate": 9.179518690212085e-07,
+      "loss": 0.1954,
+      "step": 3284
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.46856952626879744,
+      "learning_rate": 9.178951755227575e-07,
+      "loss": 0.0292,
+      "step": 3285
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7802914955954653,
+      "learning_rate": 9.178384641960582e-07,
+      "loss": 0.4202,
+      "step": 3286
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7315321948269884,
+      "learning_rate": 9.1778173504353e-07,
+      "loss": 0.1565,
+      "step": 3287
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.222393850481576,
+      "learning_rate": 9.177249880675933e-07,
+      "loss": 0.2781,
+      "step": 3288
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.666436708168778,
+      "learning_rate": 9.176682232706687e-07,
+      "loss": 0.1222,
+      "step": 3289
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9280795230892156,
+      "learning_rate": 9.176114406551783e-07,
+      "loss": 0.1178,
+      "step": 3290
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7398973842010642,
+      "learning_rate": 9.175546402235441e-07,
+      "loss": 0.326,
+      "step": 3291
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4088669314826049,
+      "learning_rate": 9.174978219781897e-07,
+      "loss": 0.2255,
+      "step": 3292
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.175157809384668,
+      "learning_rate": 9.174409859215388e-07,
+      "loss": 0.0047,
+      "step": 3293
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9253171979955276,
+      "learning_rate": 9.173841320560164e-07,
+      "loss": 0.0849,
+      "step": 3294
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7178912350948362,
+      "learning_rate": 9.173272603840478e-07,
+      "loss": 0.115,
+      "step": 3295
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8153411201239142,
+      "learning_rate": 9.172703709080593e-07,
+      "loss": 0.1246,
+      "step": 3296
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5726353685004575,
+      "learning_rate": 9.172134636304782e-07,
+      "loss": 0.37,
+      "step": 3297
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.460500779942402,
+      "learning_rate": 9.171565385537318e-07,
+      "loss": 0.1719,
+      "step": 3298
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2420851980112937,
+      "learning_rate": 9.17099595680249e-07,
+      "loss": 0.182,
+      "step": 3299
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4513015773792623,
+      "learning_rate": 9.170426350124589e-07,
+      "loss": 0.2616,
+      "step": 3300
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5483890191093747,
+      "learning_rate": 9.169856565527916e-07,
+      "loss": 0.0853,
+      "step": 3301
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3191579094376202,
+      "learning_rate": 9.169286603036781e-07,
+      "loss": 0.2658,
+      "step": 3302
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.49489582655847625,
+      "learning_rate": 9.168716462675497e-07,
+      "loss": 0.2081,
+      "step": 3303
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40434010307233975,
+      "learning_rate": 9.168146144468388e-07,
+      "loss": 0.1118,
+      "step": 3304
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.965669223985352,
+      "learning_rate": 9.167575648439787e-07,
+      "loss": 0.0482,
+      "step": 3305
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8037598988165185,
+      "learning_rate": 9.16700497461403e-07,
+      "loss": 0.1939,
+      "step": 3306
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0228184358035761,
+      "learning_rate": 9.166434123015462e-07,
+      "loss": 0.1571,
+      "step": 3307
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.106800100492122,
+      "learning_rate": 9.165863093668442e-07,
+      "loss": 0.2343,
+      "step": 3308
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.13701452775901876,
+      "learning_rate": 9.165291886597328e-07,
+      "loss": 0.0035,
+      "step": 3309
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6820776536663483,
+      "learning_rate": 9.164720501826488e-07,
+      "loss": 0.1915,
+      "step": 3310
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5055141620505854,
+      "learning_rate": 9.164148939380299e-07,
+      "loss": 0.3251,
+      "step": 3311
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.730238558757734,
+      "learning_rate": 9.163577199283144e-07,
+      "loss": 0.261,
+      "step": 3312
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3908591679499098,
+      "learning_rate": 9.163005281559418e-07,
+      "loss": 0.2514,
+      "step": 3313
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7861487704535053,
+      "learning_rate": 9.162433186233516e-07,
+      "loss": 0.2354,
+      "step": 3314
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6252351727346455,
+      "learning_rate": 9.161860913329848e-07,
+      "loss": 0.144,
+      "step": 3315
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.939100655055246,
+      "learning_rate": 9.161288462872827e-07,
+      "loss": 0.2447,
+      "step": 3316
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9484435047301384,
+      "learning_rate": 9.160715834886876e-07,
+      "loss": 0.3138,
+      "step": 3317
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9739724864529796,
+      "learning_rate": 9.160143029396421e-07,
+      "loss": 0.1434,
+      "step": 3318
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4506480021370867,
+      "learning_rate": 9.159570046425902e-07,
+      "loss": 0.0983,
+      "step": 3319
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5178633961922393,
+      "learning_rate": 9.158996885999763e-07,
+      "loss": 0.1196,
+      "step": 3320
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9579539480681765,
+      "learning_rate": 9.158423548142458e-07,
+      "loss": 0.2286,
+      "step": 3321
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3888085819064977,
+      "learning_rate": 9.157850032878445e-07,
+      "loss": 0.0631,
+      "step": 3322
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.845659596007509,
+      "learning_rate": 9.157276340232189e-07,
+      "loss": 0.167,
+      "step": 3323
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3534485994896798,
+      "learning_rate": 9.156702470228168e-07,
+      "loss": 0.0901,
+      "step": 3324
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1605635089106716,
+      "learning_rate": 9.156128422890864e-07,
+      "loss": 0.3227,
+      "step": 3325
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7428022722660884,
+      "learning_rate": 9.155554198244766e-07,
+      "loss": 0.1184,
+      "step": 3326
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2918648195664704,
+      "learning_rate": 9.154979796314373e-07,
+      "loss": 0.3346,
+      "step": 3327
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5856853294500018,
+      "learning_rate": 9.15440521712419e-07,
+      "loss": 0.1636,
+      "step": 3328
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9437189009731898,
+      "learning_rate": 9.153830460698727e-07,
+      "loss": 0.0998,
+      "step": 3329
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.11880129973324605,
+      "learning_rate": 9.153255527062508e-07,
+      "loss": 0.0034,
+      "step": 3330
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7106187310767957,
+      "learning_rate": 9.152680416240058e-07,
+      "loss": 0.1795,
+      "step": 3331
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.165237470034804,
+      "learning_rate": 9.152105128255913e-07,
+      "loss": 0.1602,
+      "step": 3332
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5943122879158438,
+      "learning_rate": 9.151529663134618e-07,
+      "loss": 0.2253,
+      "step": 3333
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6323695057504283,
+      "learning_rate": 9.150954020900721e-07,
+      "loss": 0.27,
+      "step": 3334
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6057181664144572,
+      "learning_rate": 9.150378201578783e-07,
+      "loss": 0.0806,
+      "step": 3335
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8036668462639066,
+      "learning_rate": 9.149802205193364e-07,
+      "loss": 0.3558,
+      "step": 3336
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9407285413340083,
+      "learning_rate": 9.149226031769044e-07,
+      "loss": 0.5969,
+      "step": 3337
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0538561578320227,
+      "learning_rate": 9.148649681330399e-07,
+      "loss": 0.2665,
+      "step": 3338
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9601589805852963,
+      "learning_rate": 9.148073153902018e-07,
+      "loss": 0.1061,
+      "step": 3339
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2809480204308987,
+      "learning_rate": 9.1474964495085e-07,
+      "loss": 0.3411,
+      "step": 3340
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9292083215413531,
+      "learning_rate": 9.146919568174443e-07,
+      "loss": 0.4065,
+      "step": 3341
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8696332294107295,
+      "learning_rate": 9.146342509924463e-07,
+      "loss": 0.1685,
+      "step": 3342
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8419173746837841,
+      "learning_rate": 9.145765274783176e-07,
+      "loss": 0.3281,
+      "step": 3343
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7068294963354437,
+      "learning_rate": 9.145187862775208e-07,
+      "loss": 0.3983,
+      "step": 3344
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7314959557750793,
+      "learning_rate": 9.144610273925193e-07,
+      "loss": 0.1848,
+      "step": 3345
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.33549357986216816,
+      "learning_rate": 9.144032508257772e-07,
+      "loss": 0.0108,
+      "step": 3346
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2287606074263027,
+      "learning_rate": 9.143454565797593e-07,
+      "loss": 0.3093,
+      "step": 3347
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5120194785272824,
+      "learning_rate": 9.142876446569314e-07,
+      "loss": 0.1189,
+      "step": 3348
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7351605139084882,
+      "learning_rate": 9.142298150597597e-07,
+      "loss": 0.2479,
+      "step": 3349
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8778220062242337,
+      "learning_rate": 9.141719677907114e-07,
+      "loss": 0.2634,
+      "step": 3350
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6896967732039392,
+      "learning_rate": 9.141141028522544e-07,
+      "loss": 0.4148,
+      "step": 3351
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.14780695521541093,
+      "learning_rate": 9.140562202468571e-07,
+      "loss": 0.0727,
+      "step": 3352
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.33412376814886113,
+      "learning_rate": 9.139983199769894e-07,
+      "loss": 0.1369,
+      "step": 3353
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4228120604743642,
+      "learning_rate": 9.139404020451209e-07,
+      "loss": 0.3021,
+      "step": 3354
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5890829771220137,
+      "learning_rate": 9.138824664537228e-07,
+      "loss": 0.3531,
+      "step": 3355
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9781033684600996,
+      "learning_rate": 9.138245132052667e-07,
+      "loss": 0.1932,
+      "step": 3356
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.224919100610762,
+      "learning_rate": 9.13766542302225e-07,
+      "loss": 0.1376,
+      "step": 3357
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6696608362091178,
+      "learning_rate": 9.137085537470706e-07,
+      "loss": 0.3244,
+      "step": 3358
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.43231195722747895,
+      "learning_rate": 9.136505475422778e-07,
+      "loss": 0.0095,
+      "step": 3359
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6223939786173917,
+      "learning_rate": 9.135925236903212e-07,
+      "loss": 0.1022,
+      "step": 3360
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5430507895496046,
+      "learning_rate": 9.135344821936759e-07,
+      "loss": 0.3908,
+      "step": 3361
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4061619624480858,
+      "learning_rate": 9.134764230548184e-07,
+      "loss": 0.0202,
+      "step": 3362
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7312764284057695,
+      "learning_rate": 9.134183462762255e-07,
+      "loss": 0.132,
+      "step": 3363
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4608353146168241,
+      "learning_rate": 9.133602518603749e-07,
+      "loss": 0.4335,
+      "step": 3364
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.726778267690586,
+      "learning_rate": 9.133021398097449e-07,
+      "loss": 0.3721,
+      "step": 3365
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3008911729782962,
+      "learning_rate": 9.132440101268149e-07,
+      "loss": 0.0331,
+      "step": 3366
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.34111277979521704,
+      "learning_rate": 9.131858628140647e-07,
+      "loss": 0.1128,
+      "step": 3367
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6935948489725566,
+      "learning_rate": 9.131276978739748e-07,
+      "loss": 0.0844,
+      "step": 3368
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2863030783735738,
+      "learning_rate": 9.130695153090271e-07,
+      "loss": 0.1082,
+      "step": 3369
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28620158719811806,
+      "learning_rate": 9.130113151217033e-07,
+      "loss": 0.0241,
+      "step": 3370
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8812706385404825,
+      "learning_rate": 9.129530973144866e-07,
+      "loss": 0.2179,
+      "step": 3371
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.56321103348772,
+      "learning_rate": 9.128948618898606e-07,
+      "loss": 0.3122,
+      "step": 3372
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7809217075713873,
+      "learning_rate": 9.128366088503098e-07,
+      "loss": 0.2553,
+      "step": 3373
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.487430715921176,
+      "learning_rate": 9.127783381983194e-07,
+      "loss": 0.1502,
+      "step": 3374
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5128932989123394,
+      "learning_rate": 9.127200499363752e-07,
+      "loss": 0.0713,
+      "step": 3375
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.686968120630709,
+      "learning_rate": 9.126617440669641e-07,
+      "loss": 0.2624,
+      "step": 3376
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9774532076092192,
+      "learning_rate": 9.126034205925733e-07,
+      "loss": 0.1044,
+      "step": 3377
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7031551533349818,
+      "learning_rate": 9.125450795156912e-07,
+      "loss": 0.0799,
+      "step": 3378
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.550485547328557,
+      "learning_rate": 9.124867208388067e-07,
+      "loss": 0.2822,
+      "step": 3379
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6038386615340078,
+      "learning_rate": 9.124283445644097e-07,
+      "loss": 0.1177,
+      "step": 3380
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3482277354928197,
+      "learning_rate": 9.123699506949901e-07,
+      "loss": 0.0413,
+      "step": 3381
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5948768393364658,
+      "learning_rate": 9.123115392330396e-07,
+      "loss": 0.1652,
+      "step": 3382
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3894779516719231,
+      "learning_rate": 9.1225311018105e-07,
+      "loss": 0.1263,
+      "step": 3383
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4785026138141128,
+      "learning_rate": 9.121946635415139e-07,
+      "loss": 0.0096,
+      "step": 3384
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8184672825827085,
+      "learning_rate": 9.121361993169249e-07,
+      "loss": 0.0957,
+      "step": 3385
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.246811980644127,
+      "learning_rate": 9.120777175097771e-07,
+      "loss": 0.2403,
+      "step": 3386
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9640966324191959,
+      "learning_rate": 9.120192181225656e-07,
+      "loss": 0.3179,
+      "step": 3387
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6895051692302739,
+      "learning_rate": 9.119607011577859e-07,
+      "loss": 0.2984,
+      "step": 3388
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.29280048192975544,
+      "learning_rate": 9.119021666179345e-07,
+      "loss": 0.134,
+      "step": 3389
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.137237951547494,
+      "learning_rate": 9.118436145055089e-07,
+      "loss": 0.1653,
+      "step": 3390
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5688838425370352,
+      "learning_rate": 9.117850448230065e-07,
+      "loss": 0.1128,
+      "step": 3391
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4400530005215225,
+      "learning_rate": 9.117264575729265e-07,
+      "loss": 0.0096,
+      "step": 3392
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7937961745154153,
+      "learning_rate": 9.116678527577679e-07,
+      "loss": 0.1149,
+      "step": 3393
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2148334514226202,
+      "learning_rate": 9.116092303800314e-07,
+      "loss": 0.1107,
+      "step": 3394
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9967046538924188,
+      "learning_rate": 9.115505904422175e-07,
+      "loss": 0.4086,
+      "step": 3395
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4439013094625506,
+      "learning_rate": 9.114919329468282e-07,
+      "loss": 0.2029,
+      "step": 3396
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5850941191041946,
+      "learning_rate": 9.114332578963657e-07,
+      "loss": 0.0204,
+      "step": 3397
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5596567682767649,
+      "learning_rate": 9.113745652933336e-07,
+      "loss": 0.2591,
+      "step": 3398
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7153097443238435,
+      "learning_rate": 9.113158551402354e-07,
+      "loss": 0.3206,
+      "step": 3399
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.660737198000805,
+      "learning_rate": 9.11257127439576e-07,
+      "loss": 0.3275,
+      "step": 3400
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8420471598318644,
+      "learning_rate": 9.111983821938607e-07,
+      "loss": 0.2957,
+      "step": 3401
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3974637733227828,
+      "learning_rate": 9.11139619405596e-07,
+      "loss": 0.1292,
+      "step": 3402
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8618460911519606,
+      "learning_rate": 9.110808390772885e-07,
+      "loss": 0.2779,
+      "step": 3403
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5857774826771328,
+      "learning_rate": 9.110220412114461e-07,
+      "loss": 0.2278,
+      "step": 3404
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1973854569132587,
+      "learning_rate": 9.10963225810577e-07,
+      "loss": 0.1789,
+      "step": 3405
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6026320213695966,
+      "learning_rate": 9.109043928771909e-07,
+      "loss": 0.1854,
+      "step": 3406
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7815164386452087,
+      "learning_rate": 9.10845542413797e-07,
+      "loss": 0.2431,
+      "step": 3407
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4577216473190088,
+      "learning_rate": 9.107866744229066e-07,
+      "loss": 0.1924,
+      "step": 3408
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.40027900877199785,
+      "learning_rate": 9.107277889070309e-07,
+      "loss": 0.0155,
+      "step": 3409
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.293075533621946,
+      "learning_rate": 9.106688858686819e-07,
+      "loss": 0.1404,
+      "step": 3410
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0836027098191086,
+      "learning_rate": 9.106099653103727e-07,
+      "loss": 0.2441,
+      "step": 3411
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.23668753584277077,
+      "learning_rate": 9.10551027234617e-07,
+      "loss": 0.1216,
+      "step": 3412
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.58709382814804,
+      "learning_rate": 9.104920716439293e-07,
+      "loss": 0.2864,
+      "step": 3413
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5794583746193465,
+      "learning_rate": 9.104330985408244e-07,
+      "loss": 0.2686,
+      "step": 3414
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8895225246835184,
+      "learning_rate": 9.103741079278186e-07,
+      "loss": 0.5283,
+      "step": 3415
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.39283921122768195,
+      "learning_rate": 9.103150998074283e-07,
+      "loss": 0.2204,
+      "step": 3416
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.804370593211932,
+      "learning_rate": 9.10256074182171e-07,
+      "loss": 0.1024,
+      "step": 3417
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7469095195242345,
+      "learning_rate": 9.101970310545649e-07,
+      "loss": 0.1825,
+      "step": 3418
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4301125149502763,
+      "learning_rate": 9.101379704271288e-07,
+      "loss": 0.1813,
+      "step": 3419
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.43787974520146306,
+      "learning_rate": 9.100788923023826e-07,
+      "loss": 0.1324,
+      "step": 3420
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9814626283804289,
+      "learning_rate": 9.100197966828462e-07,
+      "loss": 0.1023,
+      "step": 3421
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6053470594817225,
+      "learning_rate": 9.099606835710413e-07,
+      "loss": 0.305,
+      "step": 3422
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.43098791540108977,
+      "learning_rate": 9.099015529694893e-07,
+      "loss": 0.0927,
+      "step": 3423
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.30846510323918225,
+      "learning_rate": 9.098424048807131e-07,
+      "loss": 0.2379,
+      "step": 3424
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3731493271743987,
+      "learning_rate": 9.097832393072362e-07,
+      "loss": 0.1818,
+      "step": 3425
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6252877550256376,
+      "learning_rate": 9.097240562515824e-07,
+      "loss": 0.1592,
+      "step": 3426
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5703630961706992,
+      "learning_rate": 9.096648557162767e-07,
+      "loss": 0.3112,
+      "step": 3427
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0771455845190558,
+      "learning_rate": 9.096056377038448e-07,
+      "loss": 0.2474,
+      "step": 3428
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6758815883228315,
+      "learning_rate": 9.095464022168129e-07,
+      "loss": 0.2477,
+      "step": 3429
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3800595338507114,
+      "learning_rate": 9.094871492577081e-07,
+      "loss": 0.2124,
+      "step": 3430
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6086154417384398,
+      "learning_rate": 9.094278788290586e-07,
+      "loss": 0.142,
+      "step": 3431
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8861271522221894,
+      "learning_rate": 9.093685909333925e-07,
+      "loss": 0.081,
+      "step": 3432
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5888344856273661,
+      "learning_rate": 9.093092855732395e-07,
+      "loss": 0.5064,
+      "step": 3433
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5248938488932281,
+      "learning_rate": 9.092499627511295e-07,
+      "loss": 0.0358,
+      "step": 3434
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6760893053749472,
+      "learning_rate": 9.091906224695935e-07,
+      "loss": 0.1522,
+      "step": 3435
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5565931103970779,
+      "learning_rate": 9.091312647311629e-07,
+      "loss": 0.2553,
+      "step": 3436
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.43821355945024576,
+      "learning_rate": 9.0907188953837e-07,
+      "loss": 0.115,
+      "step": 3437
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0487219325921577,
+      "learning_rate": 9.09012496893748e-07,
+      "loss": 0.2093,
+      "step": 3438
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.710696025931875,
+      "learning_rate": 9.089530867998307e-07,
+      "loss": 0.1013,
+      "step": 3439
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.76015426839559,
+      "learning_rate": 9.088936592591528e-07,
+      "loss": 0.2028,
+      "step": 3440
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6190455024758854,
+      "learning_rate": 9.088342142742492e-07,
+      "loss": 0.3572,
+      "step": 3441
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0968735196658128,
+      "learning_rate": 9.087747518476561e-07,
+      "loss": 0.0332,
+      "step": 3442
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5591905620825228,
+      "learning_rate": 9.087152719819104e-07,
+      "loss": 0.2026,
+      "step": 3443
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8116329837785369,
+      "learning_rate": 9.086557746795497e-07,
+      "loss": 0.2132,
+      "step": 3444
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5818075691140604,
+      "learning_rate": 9.085962599431121e-07,
+      "loss": 0.2607,
+      "step": 3445
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7708689610124019,
+      "learning_rate": 9.085367277751366e-07,
+      "loss": 0.29,
+      "step": 3446
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8386997305596612,
+      "learning_rate": 9.084771781781631e-07,
+      "loss": 0.0975,
+      "step": 3447
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2847712370763844,
+      "learning_rate": 9.08417611154732e-07,
+      "loss": 0.0138,
+      "step": 3448
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5567018486648099,
+      "learning_rate": 9.083580267073846e-07,
+      "loss": 0.2348,
+      "step": 3449
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.14978507688318166,
+      "learning_rate": 9.082984248386629e-07,
+      "loss": 0.0757,
+      "step": 3450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8480235839216598,
+      "learning_rate": 9.082388055511096e-07,
+      "loss": 0.1654,
+      "step": 3451
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.758476965822796,
+      "learning_rate": 9.081791688472682e-07,
+      "loss": 0.0173,
+      "step": 3452
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.28901291714694527,
+      "learning_rate": 9.081195147296829e-07,
+      "loss": 0.1958,
+      "step": 3453
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5632435502419064,
+      "learning_rate": 9.080598432008986e-07,
+      "loss": 0.1931,
+      "step": 3454
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.20853929637542068,
+      "learning_rate": 9.080001542634612e-07,
+      "loss": 0.0078,
+      "step": 3455
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8147606927698112,
+      "learning_rate": 9.079404479199168e-07,
+      "loss": 0.2416,
+      "step": 3456
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5896403903375568,
+      "learning_rate": 9.078807241728131e-07,
+      "loss": 0.0463,
+      "step": 3457
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6193532850598485,
+      "learning_rate": 9.078209830246977e-07,
+      "loss": 0.3053,
+      "step": 3458
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3673911368542017,
+      "learning_rate": 9.077612244781195e-07,
+      "loss": 0.095,
+      "step": 3459
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4741756122471617,
+      "learning_rate": 9.077014485356274e-07,
+      "loss": 0.2309,
+      "step": 3460
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6886654317580722,
+      "learning_rate": 9.076416551997721e-07,
+      "loss": 0.1479,
+      "step": 3461
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3884602630498089,
+      "learning_rate": 9.075818444731044e-07,
+      "loss": 0.0143,
+      "step": 3462
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9800609006038236,
+      "learning_rate": 9.075220163581758e-07,
+      "loss": 0.2561,
+      "step": 3463
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.38942150556075106,
+      "learning_rate": 9.074621708575387e-07,
+      "loss": 0.103,
+      "step": 3464
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3611265364250646,
+      "learning_rate": 9.074023079737465e-07,
+      "loss": 0.2226,
+      "step": 3465
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8001913346072398,
+      "learning_rate": 9.073424277093527e-07,
+      "loss": 0.4161,
+      "step": 3466
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.44584243740872953,
+      "learning_rate": 9.07282530066912e-07,
+      "loss": 0.1643,
+      "step": 3467
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.44721985428880234,
+      "learning_rate": 9.0722261504898e-07,
+      "loss": 0.0276,
+      "step": 3468
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5476807711359044,
+      "learning_rate": 9.071626826581124e-07,
+      "loss": 0.3451,
+      "step": 3469
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4448469967113406,
+      "learning_rate": 9.071027328968665e-07,
+      "loss": 0.2502,
+      "step": 3470
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9780446987211733,
+      "learning_rate": 9.070427657677994e-07,
+      "loss": 0.1259,
+      "step": 3471
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6524682311126996,
+      "learning_rate": 9.069827812734698e-07,
+      "loss": 0.1111,
+      "step": 3472
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8261236053892801,
+      "learning_rate": 9.069227794164366e-07,
+      "loss": 0.273,
+      "step": 3473
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8456024771978315,
+      "learning_rate": 9.068627601992598e-07,
+      "loss": 0.1448,
+      "step": 3474
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4122556009766779,
+      "learning_rate": 9.068027236244995e-07,
+      "loss": 0.2236,
+      "step": 3475
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8702978327038665,
+      "learning_rate": 9.067426696947174e-07,
+      "loss": 0.3429,
+      "step": 3476
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6658536681051026,
+      "learning_rate": 9.066825984124751e-07,
+      "loss": 0.2262,
+      "step": 3477
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8015228154440659,
+      "learning_rate": 9.066225097803358e-07,
+      "loss": 0.1998,
+      "step": 3478
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.674366712686864,
+      "learning_rate": 9.065624038008628e-07,
+      "loss": 0.1484,
+      "step": 3479
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6267862760521721,
+      "learning_rate": 9.065022804766204e-07,
+      "loss": 0.107,
+      "step": 3480
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0302001191660715,
+      "learning_rate": 9.064421398101733e-07,
+      "loss": 0.3326,
+      "step": 3481
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7957291490691056,
+      "learning_rate": 9.063819818040878e-07,
+      "loss": 0.1871,
+      "step": 3482
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1249923368632135,
+      "learning_rate": 9.063218064609299e-07,
+      "loss": 0.2669,
+      "step": 3483
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7913429341548515,
+      "learning_rate": 9.062616137832668e-07,
+      "loss": 0.4369,
+      "step": 3484
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4232030837204642,
+      "learning_rate": 9.062014037736667e-07,
+      "loss": 0.3205,
+      "step": 3485
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5464013784120936,
+      "learning_rate": 9.061411764346982e-07,
+      "loss": 0.0086,
+      "step": 3486
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.47319299397048015,
+      "learning_rate": 9.060809317689306e-07,
+      "loss": 0.2368,
+      "step": 3487
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.461021949249322,
+      "learning_rate": 9.060206697789341e-07,
+      "loss": 0.1098,
+      "step": 3488
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.50489660755318,
+      "learning_rate": 9.059603904672797e-07,
+      "loss": 0.1486,
+      "step": 3489
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7062686015839136,
+      "learning_rate": 9.059000938365388e-07,
+      "loss": 0.378,
+      "step": 3490
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2318923336010885,
+      "learning_rate": 9.058397798892841e-07,
+      "loss": 0.0119,
+      "step": 3491
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.119039946903587,
+      "learning_rate": 9.057794486280885e-07,
+      "loss": 0.1205,
+      "step": 3492
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1552504432905428,
+      "learning_rate": 9.057191000555259e-07,
+      "loss": 0.2216,
+      "step": 3493
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.18240345717577552,
+      "learning_rate": 9.056587341741708e-07,
+      "loss": 0.0044,
+      "step": 3494
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.920239819563848,
+      "learning_rate": 9.055983509865988e-07,
+      "loss": 0.1902,
+      "step": 3495
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5578636167243833,
+      "learning_rate": 9.055379504953857e-07,
+      "loss": 0.2409,
+      "step": 3496
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.32673939571085087,
+      "learning_rate": 9.054775327031084e-07,
+      "loss": 0.1857,
+      "step": 3497
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3054597557355878,
+      "learning_rate": 9.054170976123444e-07,
+      "loss": 0.074,
+      "step": 3498
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.47660839361366525,
+      "learning_rate": 9.05356645225672e-07,
+      "loss": 0.1878,
+      "step": 3499
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8488780011047519,
+      "learning_rate": 9.052961755456704e-07,
+      "loss": 0.2275,
+      "step": 3500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4759906315349322,
+      "learning_rate": 9.052356885749191e-07,
+      "loss": 0.2249,
+      "step": 3501
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.18884537303613233,
+      "learning_rate": 9.051751843159987e-07,
+      "loss": 0.1303,
+      "step": 3502
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8426818040273766,
+      "learning_rate": 9.051146627714905e-07,
+      "loss": 0.188,
+      "step": 3503
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1731863384220949,
+      "learning_rate": 9.050541239439763e-07,
+      "loss": 0.2124,
+      "step": 3504
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3368678322200815,
+      "learning_rate": 9.04993567836039e-07,
+      "loss": 0.4364,
+      "step": 3505
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.621984000338872,
+      "learning_rate": 9.049329944502619e-07,
+      "loss": 0.3331,
+      "step": 3506
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0557024485074031,
+      "learning_rate": 9.048724037892293e-07,
+      "loss": 0.2421,
+      "step": 3507
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.034919908349202,
+      "learning_rate": 9.04811795855526e-07,
+      "loss": 0.216,
+      "step": 3508
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5535103811207018,
+      "learning_rate": 9.047511706517377e-07,
+      "loss": 0.2032,
+      "step": 3509
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.422120351200221,
+      "learning_rate": 9.046905281804509e-07,
+      "loss": 0.2738,
+      "step": 3510
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.001188790890657,
+      "learning_rate": 9.046298684442525e-07,
+      "loss": 0.3171,
+      "step": 3511
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7747187881056723,
+      "learning_rate": 9.045691914457305e-07,
+      "loss": 0.0981,
+      "step": 3512
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.279504743833741,
+      "learning_rate": 9.045084971874737e-07,
+      "loss": 0.0469,
+      "step": 3513
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5824157956468737,
+      "learning_rate": 9.044477856720712e-07,
+      "loss": 0.2563,
+      "step": 3514
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0540488650606123,
+      "learning_rate": 9.04387056902113e-07,
+      "loss": 0.199,
+      "step": 3515
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0224820944299369,
+      "learning_rate": 9.043263108801901e-07,
+      "loss": 0.3534,
+      "step": 3516
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.480155624685507,
+      "learning_rate": 9.04265547608894e-07,
+      "loss": 0.0166,
+      "step": 3517
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1695620230090724,
+      "learning_rate": 9.042047670908169e-07,
+      "loss": 0.1602,
+      "step": 3518
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.616620524514955,
+      "learning_rate": 9.041439693285519e-07,
+      "loss": 0.1087,
+      "step": 3519
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0588946185105392,
+      "learning_rate": 9.040831543246928e-07,
+      "loss": 0.1376,
+      "step": 3520
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.43301704960439397,
+      "learning_rate": 9.040223220818339e-07,
+      "loss": 0.1515,
+      "step": 3521
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9372290561504759,
+      "learning_rate": 9.039614726025706e-07,
+      "loss": 0.0296,
+      "step": 3522
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7290055521646502,
+      "learning_rate": 9.039006058894988e-07,
+      "loss": 0.1478,
+      "step": 3523
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6009524441824383,
+      "learning_rate": 9.038397219452154e-07,
+      "loss": 0.0492,
+      "step": 3524
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3510990732077877,
+      "learning_rate": 9.037788207723174e-07,
+      "loss": 0.1141,
+      "step": 3525
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1877054520282975,
+      "learning_rate": 9.037179023734034e-07,
+      "loss": 0.0938,
+      "step": 3526
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.38774478221254344,
+      "learning_rate": 9.036569667510719e-07,
+      "loss": 0.2048,
+      "step": 3527
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2650309446210396,
+      "learning_rate": 9.035960139079229e-07,
+      "loss": 0.1019,
+      "step": 3528
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4097990888608176,
+      "learning_rate": 9.035350438465566e-07,
+      "loss": 0.0862,
+      "step": 3529
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4170593866752147,
+      "learning_rate": 9.034740565695741e-07,
+      "loss": 0.0864,
+      "step": 3530
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.49179527875071183,
+      "learning_rate": 9.034130520795773e-07,
+      "loss": 0.0727,
+      "step": 3531
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4844474975087014,
+      "learning_rate": 9.033520303791686e-07,
+      "loss": 0.1681,
+      "step": 3532
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7568547366344485,
+      "learning_rate": 9.032909914709516e-07,
+      "loss": 0.065,
+      "step": 3533
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.439624171337634,
+      "learning_rate": 9.032299353575301e-07,
+      "loss": 0.1562,
+      "step": 3534
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.40402549015647976,
+      "learning_rate": 9.03168862041509e-07,
+      "loss": 0.108,
+      "step": 3535
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7998233890169933,
+      "learning_rate": 9.031077715254937e-07,
+      "loss": 0.2988,
+      "step": 3536
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3299364680259563,
+      "learning_rate": 9.030466638120907e-07,
+      "loss": 0.1662,
+      "step": 3537
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.33398480048584933,
+      "learning_rate": 9.029855389039067e-07,
+      "loss": 0.0746,
+      "step": 3538
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5046925690501163,
+      "learning_rate": 9.029243968035495e-07,
+      "loss": 0.1415,
+      "step": 3539
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0899228427771,
+      "learning_rate": 9.028632375136276e-07,
+      "loss": 0.2744,
+      "step": 3540
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4820328028191452,
+      "learning_rate": 9.028020610367499e-07,
+      "loss": 0.0907,
+      "step": 3541
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6686129981117492,
+      "learning_rate": 9.027408673755268e-07,
+      "loss": 0.1155,
+      "step": 3542
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7304402378877142,
+      "learning_rate": 9.026796565325687e-07,
+      "loss": 0.0507,
+      "step": 3543
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6368166231070489,
+      "learning_rate": 9.026184285104867e-07,
+      "loss": 0.3022,
+      "step": 3544
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6703179137627501,
+      "learning_rate": 9.025571833118935e-07,
+      "loss": 0.2356,
+      "step": 3545
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8085542338999495,
+      "learning_rate": 9.024959209394014e-07,
+      "loss": 0.2126,
+      "step": 3546
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.228714239918408,
+      "learning_rate": 9.024346413956243e-07,
+      "loss": 0.2395,
+      "step": 3547
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0296238013379133,
+      "learning_rate": 9.023733446831762e-07,
+      "loss": 0.1334,
+      "step": 3548
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8240975558791944,
+      "learning_rate": 9.023120308046726e-07,
+      "loss": 0.2793,
+      "step": 3549
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5901663841476886,
+      "learning_rate": 9.022506997627289e-07,
+      "loss": 0.1835,
+      "step": 3550
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4576929666443724,
+      "learning_rate": 9.021893515599618e-07,
+      "loss": 0.107,
+      "step": 3551
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7028228227971861,
+      "learning_rate": 9.021279861989884e-07,
+      "loss": 0.1563,
+      "step": 3552
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.49652258624001755,
+      "learning_rate": 9.020666036824267e-07,
+      "loss": 0.1915,
+      "step": 3553
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0911481865292731,
+      "learning_rate": 9.020052040128955e-07,
+      "loss": 0.2879,
+      "step": 3554
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9767828769964186,
+      "learning_rate": 9.019437871930143e-07,
+      "loss": 0.2459,
+      "step": 3555
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6713058897409105,
+      "learning_rate": 9.018823532254028e-07,
+      "loss": 0.1628,
+      "step": 3556
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8181149095416154,
+      "learning_rate": 9.018209021126824e-07,
+      "loss": 0.1507,
+      "step": 3557
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2652378181965303,
+      "learning_rate": 9.017594338574745e-07,
+      "loss": 0.2984,
+      "step": 3558
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5574580406275796,
+      "learning_rate": 9.016979484624017e-07,
+      "loss": 0.2797,
+      "step": 3559
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.085960464776571,
+      "learning_rate": 9.016364459300867e-07,
+      "loss": 0.0731,
+      "step": 3560
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9820920115828107,
+      "learning_rate": 9.015749262631535e-07,
+      "loss": 0.1038,
+      "step": 3561
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7786666678562189,
+      "learning_rate": 9.015133894642268e-07,
+      "loss": 0.1636,
+      "step": 3562
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8077732798249364,
+      "learning_rate": 9.014518355359318e-07,
+      "loss": 0.2149,
+      "step": 3563
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8379422447571268,
+      "learning_rate": 9.013902644808944e-07,
+      "loss": 0.0319,
+      "step": 3564
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8020231600160077,
+      "learning_rate": 9.013286763017414e-07,
+      "loss": 0.0149,
+      "step": 3565
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7451250982089039,
+      "learning_rate": 9.012670710011003e-07,
+      "loss": 0.1621,
+      "step": 3566
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5775504911964549,
+      "learning_rate": 9.012054485815993e-07,
+      "loss": 0.0363,
+      "step": 3567
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.36901694160006543,
+      "learning_rate": 9.011438090458674e-07,
+      "loss": 0.1155,
+      "step": 3568
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4530151377449247,
+      "learning_rate": 9.010821523965342e-07,
+      "loss": 0.4444,
+      "step": 3569
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7453481219095509,
+      "learning_rate": 9.0102047863623e-07,
+      "loss": 0.401,
+      "step": 3570
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6608854831066162,
+      "learning_rate": 9.009587877675862e-07,
+      "loss": 0.0626,
+      "step": 3571
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6146155111136735,
+      "learning_rate": 9.008970797932343e-07,
+      "loss": 0.1994,
+      "step": 3572
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6827877979600937,
+      "learning_rate": 9.008353547158072e-07,
+      "loss": 0.1298,
+      "step": 3573
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.44509956643235676,
+      "learning_rate": 9.00773612537938e-07,
+      "loss": 0.0597,
+      "step": 3574
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9247401941630845,
+      "learning_rate": 9.007118532622608e-07,
+      "loss": 0.3066,
+      "step": 3575
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3500663477620438,
+      "learning_rate": 9.006500768914106e-07,
+      "loss": 0.2817,
+      "step": 3576
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.147320025553286,
+      "learning_rate": 9.005882834280226e-07,
+      "loss": 0.335,
+      "step": 3577
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5504171426043073,
+      "learning_rate": 9.005264728747331e-07,
+      "loss": 0.3613,
+      "step": 3578
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8790360380964415,
+      "learning_rate": 9.004646452341792e-07,
+      "loss": 0.0743,
+      "step": 3579
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3644271116628688,
+      "learning_rate": 9.004028005089985e-07,
+      "loss": 0.1868,
+      "step": 3580
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.34683762899779425,
+      "learning_rate": 9.003409387018293e-07,
+      "loss": 0.0939,
+      "step": 3581
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7087190263976482,
+      "learning_rate": 9.00279059815311e-07,
+      "loss": 0.1915,
+      "step": 3582
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9889991485407398,
+      "learning_rate": 9.002171638520833e-07,
+      "loss": 0.2909,
+      "step": 3583
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6306306038822274,
+      "learning_rate": 9.001552508147868e-07,
+      "loss": 0.3875,
+      "step": 3584
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8414396902575082,
+      "learning_rate": 9.000933207060629e-07,
+      "loss": 0.2494,
+      "step": 3585
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3442777423239455,
+      "learning_rate": 9.000313735285537e-07,
+      "loss": 0.241,
+      "step": 3586
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7755224462971616,
+      "learning_rate": 8.999694092849019e-07,
+      "loss": 0.0217,
+      "step": 3587
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7497411582538355,
+      "learning_rate": 8.999074279777511e-07,
+      "loss": 0.269,
+      "step": 3588
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9658463537603914,
+      "learning_rate": 8.998454296097456e-07,
+      "loss": 0.216,
+      "step": 3589
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.156068956355357,
+      "learning_rate": 8.997834141835301e-07,
+      "loss": 0.0136,
+      "step": 3590
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0337821522188728,
+      "learning_rate": 8.997213817017506e-07,
+      "loss": 0.1946,
+      "step": 3591
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5773588168066646,
+      "learning_rate": 8.996593321670533e-07,
+      "loss": 0.203,
+      "step": 3592
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5244756314694453,
+      "learning_rate": 8.995972655820856e-07,
+      "loss": 0.2456,
+      "step": 3593
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4893468963190216,
+      "learning_rate": 8.995351819494952e-07,
+      "loss": 0.0508,
+      "step": 3594
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8466957374097075,
+      "learning_rate": 8.994730812719307e-07,
+      "loss": 0.1263,
+      "step": 3595
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7415694023113186,
+      "learning_rate": 8.994109635520416e-07,
+      "loss": 0.2457,
+      "step": 3596
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7099841326077181,
+      "learning_rate": 8.99348828792478e-07,
+      "loss": 0.0197,
+      "step": 3597
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.38683694772595284,
+      "learning_rate": 8.992866769958904e-07,
+      "loss": 0.1075,
+      "step": 3598
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.17439063080097897,
+      "learning_rate": 8.992245081649304e-07,
+      "loss": 0.0243,
+      "step": 3599
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.970067514021893,
+      "learning_rate": 8.991623223022505e-07,
+      "loss": 0.1047,
+      "step": 3600
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4837965208476653,
+      "learning_rate": 8.991001194105034e-07,
+      "loss": 0.0782,
+      "step": 3601
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9435958016207453,
+      "learning_rate": 8.990378994923431e-07,
+      "loss": 0.2051,
+      "step": 3602
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2429304708913325,
+      "learning_rate": 8.989756625504237e-07,
+      "loss": 0.1081,
+      "step": 3603
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.055464231568157,
+      "learning_rate": 8.989134085874006e-07,
+      "loss": 0.2532,
+      "step": 3604
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.569220773678353,
+      "learning_rate": 8.988511376059295e-07,
+      "loss": 0.1961,
+      "step": 3605
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.24562876287549157,
+      "learning_rate": 8.98788849608667e-07,
+      "loss": 0.0949,
+      "step": 3606
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5056491008363143,
+      "learning_rate": 8.987265445982706e-07,
+      "loss": 0.1763,
+      "step": 3607
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9208834173216598,
+      "learning_rate": 8.986642225773984e-07,
+      "loss": 0.3181,
+      "step": 3608
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5927033106464928,
+      "learning_rate": 8.986018835487088e-07,
+      "loss": 0.1896,
+      "step": 3609
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8667824774310198,
+      "learning_rate": 8.985395275148618e-07,
+      "loss": 0.1819,
+      "step": 3610
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5911908766796049,
+      "learning_rate": 8.984771544785172e-07,
+      "loss": 0.4034,
+      "step": 3611
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.730668819028163,
+      "learning_rate": 8.984147644423361e-07,
+      "loss": 0.1424,
+      "step": 3612
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5961925230821833,
+      "learning_rate": 8.983523574089805e-07,
+      "loss": 0.3239,
+      "step": 3613
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3782501076675327,
+      "learning_rate": 8.982899333811123e-07,
+      "loss": 0.2761,
+      "step": 3614
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.225164008271313,
+      "learning_rate": 8.98227492361395e-07,
+      "loss": 0.1838,
+      "step": 3615
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6473365017567108,
+      "learning_rate": 8.981650343524923e-07,
+      "loss": 0.0759,
+      "step": 3616
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7565623573857385,
+      "learning_rate": 8.981025593570689e-07,
+      "loss": 0.2515,
+      "step": 3617
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4702202017001563,
+      "learning_rate": 8.980400673777899e-07,
+      "loss": 0.0947,
+      "step": 3618
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9656628407236097,
+      "learning_rate": 8.979775584173215e-07,
+      "loss": 0.2115,
+      "step": 3619
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.860858477890622,
+      "learning_rate": 8.979150324783304e-07,
+      "loss": 0.0973,
+      "step": 3620
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0181446681122626,
+      "learning_rate": 8.978524895634842e-07,
+      "loss": 0.1194,
+      "step": 3621
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.53271215351201,
+      "learning_rate": 8.97789929675451e-07,
+      "loss": 0.4167,
+      "step": 3622
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6576937400855337,
+      "learning_rate": 8.977273528168995e-07,
+      "loss": 0.2884,
+      "step": 3623
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4513117586626686,
+      "learning_rate": 8.976647589905e-07,
+      "loss": 0.1806,
+      "step": 3624
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8760306204582416,
+      "learning_rate": 8.976021481989222e-07,
+      "loss": 0.2436,
+      "step": 3625
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5021420846358876,
+      "learning_rate": 8.975395204448375e-07,
+      "loss": 0.2186,
+      "step": 3626
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6725821463708842,
+      "learning_rate": 8.974768757309178e-07,
+      "loss": 0.234,
+      "step": 3627
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5985002495509291,
+      "learning_rate": 8.974142140598355e-07,
+      "loss": 0.2526,
+      "step": 3628
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.07129142280051656,
+      "learning_rate": 8.97351535434264e-07,
+      "loss": 0.003,
+      "step": 3629
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7747541248706957,
+      "learning_rate": 8.972888398568771e-07,
+      "loss": 0.0597,
+      "step": 3630
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.42375668303332786,
+      "learning_rate": 8.972261273303496e-07,
+      "loss": 0.1236,
+      "step": 3631
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.20468387464282548,
+      "learning_rate": 8.971633978573572e-07,
+      "loss": 0.151,
+      "step": 3632
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.650357402418647,
+      "learning_rate": 8.971006514405757e-07,
+      "loss": 0.1957,
+      "step": 3633
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5187919697079516,
+      "learning_rate": 8.970378880826821e-07,
+      "loss": 0.1054,
+      "step": 3634
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0042025818428884,
+      "learning_rate": 8.969751077863541e-07,
+      "loss": 0.2455,
+      "step": 3635
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1524725878424387,
+      "learning_rate": 8.969123105542701e-07,
+      "loss": 0.1395,
+      "step": 3636
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5317960185564282,
+      "learning_rate": 8.968494963891088e-07,
+      "loss": 0.1611,
+      "step": 3637
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2729045166459671,
+      "learning_rate": 8.967866652935505e-07,
+      "loss": 0.19,
+      "step": 3638
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7719217778106593,
+      "learning_rate": 8.967238172702752e-07,
+      "loss": 0.2054,
+      "step": 3639
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0165043542325558,
+      "learning_rate": 8.966609523219644e-07,
+      "loss": 0.1063,
+      "step": 3640
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9964120352545338,
+      "learning_rate": 8.965980704513001e-07,
+      "loss": 0.4258,
+      "step": 3641
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5789623426836484,
+      "learning_rate": 8.965351716609646e-07,
+      "loss": 0.2288,
+      "step": 3642
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4845448513474563,
+      "learning_rate": 8.964722559536417e-07,
+      "loss": 0.1012,
+      "step": 3643
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6260612248673366,
+      "learning_rate": 8.964093233320154e-07,
+      "loss": 0.1403,
+      "step": 3644
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7093553891900518,
+      "learning_rate": 8.963463737987705e-07,
+      "loss": 0.1666,
+      "step": 3645
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9548331172676473,
+      "learning_rate": 8.962834073565923e-07,
+      "loss": 0.1197,
+      "step": 3646
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6320170371849005,
+      "learning_rate": 8.962204240081675e-07,
+      "loss": 0.2755,
+      "step": 3647
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8136771418377878,
+      "learning_rate": 8.961574237561829e-07,
+      "loss": 0.178,
+      "step": 3648
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6243822576557541,
+      "learning_rate": 8.960944066033262e-07,
+      "loss": 0.336,
+      "step": 3649
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5325262107975081,
+      "learning_rate": 8.960313725522859e-07,
+      "loss": 0.131,
+      "step": 3650
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.18395353413330576,
+      "learning_rate": 8.959683216057511e-07,
+      "loss": 0.1292,
+      "step": 3651
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9914847779468045,
+      "learning_rate": 8.959052537664117e-07,
+      "loss": 0.1831,
+      "step": 3652
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.38738104396809164,
+      "learning_rate": 8.958421690369583e-07,
+      "loss": 0.1754,
+      "step": 3653
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.836668169740809,
+      "learning_rate": 8.957790674200822e-07,
+      "loss": 0.2014,
+      "step": 3654
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6888123401638223,
+      "learning_rate": 8.957159489184756e-07,
+      "loss": 0.2075,
+      "step": 3655
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.40301539098331646,
+      "learning_rate": 8.956528135348309e-07,
+      "loss": 0.049,
+      "step": 3656
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1883763894185228,
+      "learning_rate": 8.955896612718419e-07,
+      "loss": 0.0882,
+      "step": 3657
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.41368478323780855,
+      "learning_rate": 8.955264921322028e-07,
+      "loss": 0.273,
+      "step": 3658
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8492349494450537,
+      "learning_rate": 8.954633061186085e-07,
+      "loss": 0.1135,
+      "step": 3659
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9084133800749307,
+      "learning_rate": 8.954001032337544e-07,
+      "loss": 0.237,
+      "step": 3660
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9855625340864884,
+      "learning_rate": 8.953368834803371e-07,
+      "loss": 0.2454,
+      "step": 3661
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5264867254320804,
+      "learning_rate": 8.952736468610537e-07,
+      "loss": 0.1048,
+      "step": 3662
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3548027213828942,
+      "learning_rate": 8.952103933786018e-07,
+      "loss": 0.1395,
+      "step": 3663
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.20752820812496128,
+      "learning_rate": 8.951471230356802e-07,
+      "loss": 0.0762,
+      "step": 3664
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2104146494183132,
+      "learning_rate": 8.950838358349879e-07,
+      "loss": 0.0713,
+      "step": 3665
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5097759774422576,
+      "learning_rate": 8.950205317792248e-07,
+      "loss": 0.149,
+      "step": 3666
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8002322322693497,
+      "learning_rate": 8.949572108710919e-07,
+      "loss": 0.1094,
+      "step": 3667
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7567734818145712,
+      "learning_rate": 8.948938731132905e-07,
+      "loss": 0.1855,
+      "step": 3668
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5485931173642535,
+      "learning_rate": 8.948305185085224e-07,
+      "loss": 0.1999,
+      "step": 3669
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7758814921664122,
+      "learning_rate": 8.947671470594909e-07,
+      "loss": 0.253,
+      "step": 3670
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.548740515899794,
+      "learning_rate": 8.947037587688991e-07,
+      "loss": 0.1348,
+      "step": 3671
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8562468637051345,
+      "learning_rate": 8.946403536394517e-07,
+      "loss": 0.2659,
+      "step": 3672
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5972999892965324,
+      "learning_rate": 8.945769316738534e-07,
+      "loss": 0.3071,
+      "step": 3673
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7637563834327782,
+      "learning_rate": 8.945134928748099e-07,
+      "loss": 0.2423,
+      "step": 3674
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.724880431718904,
+      "learning_rate": 8.944500372450279e-07,
+      "loss": 0.0644,
+      "step": 3675
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.29369274739420687,
+      "learning_rate": 8.943865647872142e-07,
+      "loss": 0.0833,
+      "step": 3676
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5174022107305446,
+      "learning_rate": 8.943230755040769e-07,
+      "loss": 0.0989,
+      "step": 3677
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3386383112516127,
+      "learning_rate": 8.942595693983246e-07,
+      "loss": 0.1398,
+      "step": 3678
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2894587813608773,
+      "learning_rate": 8.941960464726664e-07,
+      "loss": 0.2248,
+      "step": 3679
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9263771756047618,
+      "learning_rate": 8.941325067298125e-07,
+      "loss": 0.1979,
+      "step": 3680
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8364227497770703,
+      "learning_rate": 8.940689501724736e-07,
+      "loss": 0.2238,
+      "step": 3681
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6979653931635937,
+      "learning_rate": 8.940053768033608e-07,
+      "loss": 0.1571,
+      "step": 3682
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.11305024212531636,
+      "learning_rate": 8.93941786625187e-07,
+      "loss": 0.0059,
+      "step": 3683
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5546405488659987,
+      "learning_rate": 8.938781796406645e-07,
+      "loss": 0.4754,
+      "step": 3684
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8013374007740004,
+      "learning_rate": 8.93814555852507e-07,
+      "loss": 0.223,
+      "step": 3685
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.36380398795993935,
+      "learning_rate": 8.937509152634288e-07,
+      "loss": 0.1786,
+      "step": 3686
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.30294934011358565,
+      "learning_rate": 8.936872578761452e-07,
+      "loss": 0.0062,
+      "step": 3687
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.113872932046635,
+      "learning_rate": 8.936235836933716e-07,
+      "loss": 0.1251,
+      "step": 3688
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.927337177241909,
+      "learning_rate": 8.935598927178247e-07,
+      "loss": 0.198,
+      "step": 3689
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9345968878431729,
+      "learning_rate": 8.934961849522218e-07,
+      "loss": 0.1778,
+      "step": 3690
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7948996619172269,
+      "learning_rate": 8.934324603992803e-07,
+      "loss": 0.0913,
+      "step": 3691
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8982354655223829,
+      "learning_rate": 8.933687190617194e-07,
+      "loss": 0.1007,
+      "step": 3692
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.72415753707952,
+      "learning_rate": 8.933049609422581e-07,
+      "loss": 0.3315,
+      "step": 3693
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.506559311550516,
+      "learning_rate": 8.932411860436165e-07,
+      "loss": 0.1811,
+      "step": 3694
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5399864366168942,
+      "learning_rate": 8.931773943685155e-07,
+      "loss": 0.0679,
+      "step": 3695
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5290569932581717,
+      "learning_rate": 8.931135859196762e-07,
+      "loss": 0.2001,
+      "step": 3696
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.683327435889844,
+      "learning_rate": 8.930497606998213e-07,
+      "loss": 0.2632,
+      "step": 3697
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5840598372568698,
+      "learning_rate": 8.929859187116734e-07,
+      "loss": 0.1184,
+      "step": 3698
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5773771090751436,
+      "learning_rate": 8.929220599579562e-07,
+      "loss": 0.1974,
+      "step": 3699
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7974491427480053,
+      "learning_rate": 8.92858184441394e-07,
+      "loss": 0.1526,
+      "step": 3700
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.87679290499142,
+      "learning_rate": 8.92794292164712e-07,
+      "loss": 0.1365,
+      "step": 3701
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.44484177771603206,
+      "learning_rate": 8.927303831306358e-07,
+      "loss": 0.0774,
+      "step": 3702
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.819815447823791,
+      "learning_rate": 8.926664573418922e-07,
+      "loss": 0.177,
+      "step": 3703
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9231950718037863,
+      "learning_rate": 8.92602514801208e-07,
+      "loss": 0.2868,
+      "step": 3704
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.20792229694152603,
+      "learning_rate": 8.925385555113111e-07,
+      "loss": 0.0374,
+      "step": 3705
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6290562676820869,
+      "learning_rate": 8.924745794749307e-07,
+      "loss": 0.1801,
+      "step": 3706
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8703829044456655,
+      "learning_rate": 8.924105866947955e-07,
+      "loss": 0.1889,
+      "step": 3707
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.38764591570315005,
+      "learning_rate": 8.923465771736359e-07,
+      "loss": 0.3859,
+      "step": 3708
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7098724276435888,
+      "learning_rate": 8.922825509141827e-07,
+      "loss": 0.2385,
+      "step": 3709
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.38401416793688486,
+      "learning_rate": 8.922185079191671e-07,
+      "loss": 0.0083,
+      "step": 3710
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0987334480298931,
+      "learning_rate": 8.921544481913217e-07,
+      "loss": 0.3234,
+      "step": 3711
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5960848589499238,
+      "learning_rate": 8.920903717333789e-07,
+      "loss": 0.1536,
+      "step": 3712
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7386602970647814,
+      "learning_rate": 8.92026278548073e-07,
+      "loss": 0.2815,
+      "step": 3713
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.31023055252201687,
+      "learning_rate": 8.919621686381378e-07,
+      "loss": 0.0413,
+      "step": 3714
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0399407344066338,
+      "learning_rate": 8.918980420063086e-07,
+      "loss": 0.1246,
+      "step": 3715
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2737335065949902,
+      "learning_rate": 8.918338986553211e-07,
+      "loss": 0.0326,
+      "step": 3716
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.296067112610091,
+      "learning_rate": 8.917697385879117e-07,
+      "loss": 0.3811,
+      "step": 3717
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.794927169010393,
+      "learning_rate": 8.917055618068178e-07,
+      "loss": 0.2224,
+      "step": 3718
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4387865933496512,
+      "learning_rate": 8.916413683147772e-07,
+      "loss": 0.1924,
+      "step": 3719
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5285970623958716,
+      "learning_rate": 8.915771581145285e-07,
+      "loss": 0.1733,
+      "step": 3720
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6351532285632555,
+      "learning_rate": 8.915129312088112e-07,
+      "loss": 0.0195,
+      "step": 3721
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.42340552146390065,
+      "learning_rate": 8.914486876003649e-07,
+      "loss": 0.1661,
+      "step": 3722
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8304222957107547,
+      "learning_rate": 8.913844272919309e-07,
+      "loss": 0.3068,
+      "step": 3723
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5835541018828425,
+      "learning_rate": 8.913201502862504e-07,
+      "loss": 0.1241,
+      "step": 3724
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0814322182376175,
+      "learning_rate": 8.912558565860657e-07,
+      "loss": 0.1055,
+      "step": 3725
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5133578413313381,
+      "learning_rate": 8.911915461941196e-07,
+      "loss": 0.2432,
+      "step": 3726
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7605543377735867,
+      "learning_rate": 8.911272191131559e-07,
+      "loss": 0.016,
+      "step": 3727
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6770768411894138,
+      "learning_rate": 8.910628753459184e-07,
+      "loss": 0.1394,
+      "step": 3728
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.40152719956764704,
+      "learning_rate": 8.909985148951528e-07,
+      "loss": 0.0609,
+      "step": 3729
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7574561410672707,
+      "learning_rate": 8.909341377636044e-07,
+      "loss": 0.0764,
+      "step": 3730
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7746749103878682,
+      "learning_rate": 8.908697439540198e-07,
+      "loss": 0.0302,
+      "step": 3731
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6624940135187026,
+      "learning_rate": 8.908053334691463e-07,
+      "loss": 0.0921,
+      "step": 3732
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.328521198911137,
+      "learning_rate": 8.907409063117317e-07,
+      "loss": 0.2881,
+      "step": 3733
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9408310520518564,
+      "learning_rate": 8.906764624845244e-07,
+      "loss": 0.0253,
+      "step": 3734
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.829655235082486,
+      "learning_rate": 8.906120019902739e-07,
+      "loss": 0.0305,
+      "step": 3735
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5149580058749262,
+      "learning_rate": 8.905475248317302e-07,
+      "loss": 0.3189,
+      "step": 3736
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16660177430372192,
+      "learning_rate": 8.904830310116439e-07,
+      "loss": 0.0031,
+      "step": 3737
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3968282733334314,
+      "learning_rate": 8.904185205327666e-07,
+      "loss": 0.1126,
+      "step": 3738
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4412258546287834,
+      "learning_rate": 8.903539933978504e-07,
+      "loss": 0.2691,
+      "step": 3739
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4155714819587609,
+      "learning_rate": 8.902894496096481e-07,
+      "loss": 0.1032,
+      "step": 3740
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6939420353806862,
+      "learning_rate": 8.902248891709132e-07,
+      "loss": 0.1945,
+      "step": 3741
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.31966914646427064,
+      "learning_rate": 8.901603120844003e-07,
+      "loss": 0.108,
+      "step": 3742
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7902264693969758,
+      "learning_rate": 8.900957183528639e-07,
+      "loss": 0.3815,
+      "step": 3743
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0140197309363976,
+      "learning_rate": 8.900311079790601e-07,
+      "loss": 0.1647,
+      "step": 3744
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3103666139597173,
+      "learning_rate": 8.899664809657453e-07,
+      "loss": 0.0161,
+      "step": 3745
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9515611665852027,
+      "learning_rate": 8.899018373156763e-07,
+      "loss": 0.1369,
+      "step": 3746
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5987886609029958,
+      "learning_rate": 8.898371770316111e-07,
+      "loss": 0.2235,
+      "step": 3747
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6756562521449005,
+      "learning_rate": 8.897725001163083e-07,
+      "loss": 0.058,
+      "step": 3748
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.31828381595214017,
+      "learning_rate": 8.897078065725272e-07,
+      "loss": 0.2259,
+      "step": 3749
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7350619858446041,
+      "learning_rate": 8.896430964030277e-07,
+      "loss": 0.0763,
+      "step": 3750
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3275744319752052,
+      "learning_rate": 8.895783696105703e-07,
+      "loss": 0.0234,
+      "step": 3751
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9536052494238427,
+      "learning_rate": 8.895136261979166e-07,
+      "loss": 0.0572,
+      "step": 3752
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6724851710180773,
+      "learning_rate": 8.894488661678285e-07,
+      "loss": 0.0616,
+      "step": 3753
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2867622074934325,
+      "learning_rate": 8.893840895230689e-07,
+      "loss": 0.0287,
+      "step": 3754
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.579933443117816,
+      "learning_rate": 8.893192962664012e-07,
+      "loss": 0.0795,
+      "step": 3755
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7035797738352091,
+      "learning_rate": 8.892544864005898e-07,
+      "loss": 0.035,
+      "step": 3756
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.42956073102948233,
+      "learning_rate": 8.891896599283994e-07,
+      "loss": 0.3451,
+      "step": 3757
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8172843223769409,
+      "learning_rate": 8.891248168525957e-07,
+      "loss": 0.1576,
+      "step": 3758
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8120322099943349,
+      "learning_rate": 8.890599571759454e-07,
+      "loss": 0.4472,
+      "step": 3759
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6618354528280581,
+      "learning_rate": 8.88995080901215e-07,
+      "loss": 0.1247,
+      "step": 3760
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9775245784699352,
+      "learning_rate": 8.889301880311724e-07,
+      "loss": 0.1023,
+      "step": 3761
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.555236487849536,
+      "learning_rate": 8.888652785685861e-07,
+      "loss": 0.1841,
+      "step": 3762
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3382029675723734,
+      "learning_rate": 8.888003525162256e-07,
+      "loss": 0.1968,
+      "step": 3763
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8481347461852061,
+      "learning_rate": 8.887354098768602e-07,
+      "loss": 0.0487,
+      "step": 3764
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8186472691297958,
+      "learning_rate": 8.886704506532609e-07,
+      "loss": 0.1845,
+      "step": 3765
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5869573020823922,
+      "learning_rate": 8.886054748481988e-07,
+      "loss": 0.1403,
+      "step": 3766
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4056982892023109,
+      "learning_rate": 8.885404824644459e-07,
+      "loss": 0.1476,
+      "step": 3767
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8142308314391018,
+      "learning_rate": 8.88475473504775e-07,
+      "loss": 0.2135,
+      "step": 3768
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7653636598760375,
+      "learning_rate": 8.884104479719594e-07,
+      "loss": 0.1933,
+      "step": 3769
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6010367539071723,
+      "learning_rate": 8.883454058687734e-07,
+      "loss": 0.1709,
+      "step": 3770
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8750845813632222,
+      "learning_rate": 8.882803471979916e-07,
+      "loss": 0.2474,
+      "step": 3771
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9290887108245592,
+      "learning_rate": 8.882152719623898e-07,
+      "loss": 0.1797,
+      "step": 3772
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1949534146437264,
+      "learning_rate": 8.881501801647439e-07,
+      "loss": 0.0995,
+      "step": 3773
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6126499073367744,
+      "learning_rate": 8.880850718078312e-07,
+      "loss": 0.3397,
+      "step": 3774
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.892068876882014,
+      "learning_rate": 8.880199468944291e-07,
+      "loss": 0.1426,
+      "step": 3775
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6419376933464616,
+      "learning_rate": 8.87954805427316e-07,
+      "loss": 0.2155,
+      "step": 3776
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.20500359466811582,
+      "learning_rate": 8.878896474092712e-07,
+      "loss": 0.0904,
+      "step": 3777
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5216916846480123,
+      "learning_rate": 8.878244728430742e-07,
+      "loss": 0.1431,
+      "step": 3778
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3122812443037244,
+      "learning_rate": 8.877592817315054e-07,
+      "loss": 0.1966,
+      "step": 3779
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7561955810362375,
+      "learning_rate": 8.876940740773463e-07,
+      "loss": 0.4138,
+      "step": 3780
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.1604726993973071,
+      "learning_rate": 8.876288498833786e-07,
+      "loss": 0.3771,
+      "step": 3781
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7917105456265333,
+      "learning_rate": 8.87563609152385e-07,
+      "loss": 0.0606,
+      "step": 3782
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3562561210111628,
+      "learning_rate": 8.874983518871486e-07,
+      "loss": 0.2973,
+      "step": 3783
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5272699290899061,
+      "learning_rate": 8.874330780904537e-07,
+      "loss": 0.1819,
+      "step": 3784
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5617773042352198,
+      "learning_rate": 8.873677877650847e-07,
+      "loss": 0.2149,
+      "step": 3785
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7927284887821636,
+      "learning_rate": 8.873024809138272e-07,
+      "loss": 0.2816,
+      "step": 3786
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.127433478609849,
+      "learning_rate": 8.872371575394674e-07,
+      "loss": 0.3469,
+      "step": 3787
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8124809502485485,
+      "learning_rate": 8.87171817644792e-07,
+      "loss": 0.1047,
+      "step": 3788
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6961022059067764,
+      "learning_rate": 8.871064612325885e-07,
+      "loss": 0.2903,
+      "step": 3789
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8602851144623159,
+      "learning_rate": 8.870410883056451e-07,
+      "loss": 0.026,
+      "step": 3790
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9576690687775362,
+      "learning_rate": 8.869756988667508e-07,
+      "loss": 0.2327,
+      "step": 3791
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.38831604542025516,
+      "learning_rate": 8.869102929186953e-07,
+      "loss": 0.1492,
+      "step": 3792
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4334388429938352,
+      "learning_rate": 8.868448704642691e-07,
+      "loss": 0.267,
+      "step": 3793
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.055976164987003,
+      "learning_rate": 8.867794315062629e-07,
+      "loss": 0.3208,
+      "step": 3794
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5403932677561274,
+      "learning_rate": 8.867139760474687e-07,
+      "loss": 0.1477,
+      "step": 3795
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.21410656598523806,
+      "learning_rate": 8.866485040906788e-07,
+      "loss": 0.0974,
+      "step": 3796
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9468380997808159,
+      "learning_rate": 8.865830156386866e-07,
+      "loss": 0.2684,
+      "step": 3797
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9282333660947354,
+      "learning_rate": 8.865175106942857e-07,
+      "loss": 0.3556,
+      "step": 3798
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.33742281295306603,
+      "learning_rate": 8.864519892602709e-07,
+      "loss": 0.0425,
+      "step": 3799
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2564181620645343,
+      "learning_rate": 8.863864513394372e-07,
+      "loss": 0.227,
+      "step": 3800
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.023836021967167408,
+      "learning_rate": 8.863208969345809e-07,
+      "loss": 0.0004,
+      "step": 3801
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4932632983783014,
+      "learning_rate": 8.862553260484984e-07,
+      "loss": 0.0972,
+      "step": 3802
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8612358583508158,
+      "learning_rate": 8.861897386839874e-07,
+      "loss": 0.2424,
+      "step": 3803
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8281799329071436,
+      "learning_rate": 8.861241348438457e-07,
+      "loss": 0.2149,
+      "step": 3804
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3504376151949932,
+      "learning_rate": 8.860585145308722e-07,
+      "loss": 0.0925,
+      "step": 3805
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7966390297010175,
+      "learning_rate": 8.859928777478664e-07,
+      "loss": 0.2819,
+      "step": 3806
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8976778103953185,
+      "learning_rate": 8.859272244976286e-07,
+      "loss": 0.0704,
+      "step": 3807
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6317080723498039,
+      "learning_rate": 8.858615547829594e-07,
+      "loss": 0.1435,
+      "step": 3808
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.633199222329602,
+      "learning_rate": 8.857958686066607e-07,
+      "loss": 0.2054,
+      "step": 3809
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2357002728651416,
+      "learning_rate": 8.857301659715347e-07,
+      "loss": 0.1984,
+      "step": 3810
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4089302300435955,
+      "learning_rate": 8.856644468803845e-07,
+      "loss": 0.3983,
+      "step": 3811
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4135613464162196,
+      "learning_rate": 8.855987113360134e-07,
+      "loss": 0.176,
+      "step": 3812
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.37719396902233837,
+      "learning_rate": 8.855329593412264e-07,
+      "loss": 0.0489,
+      "step": 3813
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8875365007259433,
+      "learning_rate": 8.854671908988283e-07,
+      "loss": 0.1269,
+      "step": 3814
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0181689889722372,
+      "learning_rate": 8.854014060116249e-07,
+      "loss": 0.2733,
+      "step": 3815
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5273144421567434,
+      "learning_rate": 8.853356046824228e-07,
+      "loss": 0.1581,
+      "step": 3816
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9671347036238022,
+      "learning_rate": 8.852697869140292e-07,
+      "loss": 0.1124,
+      "step": 3817
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7650637101924961,
+      "learning_rate": 8.85203952709252e-07,
+      "loss": 0.2541,
+      "step": 3818
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4237115447194473,
+      "learning_rate": 8.851381020708998e-07,
+      "loss": 0.0422,
+      "step": 3819
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6233864538524351,
+      "learning_rate": 8.850722350017818e-07,
+      "loss": 0.0644,
+      "step": 3820
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.428489971635069,
+      "learning_rate": 8.850063515047083e-07,
+      "loss": 0.183,
+      "step": 3821
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8953283052132391,
+      "learning_rate": 8.8494045158249e-07,
+      "loss": 0.3399,
+      "step": 3822
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6535801662496468,
+      "learning_rate": 8.848745352379381e-07,
+      "loss": 0.1782,
+      "step": 3823
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.49997927614318116,
+      "learning_rate": 8.848086024738648e-07,
+      "loss": 0.1637,
+      "step": 3824
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6410418209206888,
+      "learning_rate": 8.847426532930829e-07,
+      "loss": 0.1569,
+      "step": 3825
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9459033036201963,
+      "learning_rate": 8.846766876984061e-07,
+      "loss": 0.195,
+      "step": 3826
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6702740291675603,
+      "learning_rate": 8.846107056926484e-07,
+      "loss": 0.4062,
+      "step": 3827
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4524134624553657,
+      "learning_rate": 8.84544707278625e-07,
+      "loss": 0.1911,
+      "step": 3828
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9767117423338574,
+      "learning_rate": 8.844786924591512e-07,
+      "loss": 0.0574,
+      "step": 3829
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.881510543980618,
+      "learning_rate": 8.844126612370435e-07,
+      "loss": 0.0707,
+      "step": 3830
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7627701634454506,
+      "learning_rate": 8.84346613615119e-07,
+      "loss": 0.057,
+      "step": 3831
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2173122088258248,
+      "learning_rate": 8.842805495961952e-07,
+      "loss": 0.002,
+      "step": 3832
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.580379117277505,
+      "learning_rate": 8.842144691830906e-07,
+      "loss": 0.1342,
+      "step": 3833
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.120180075118133,
+      "learning_rate": 8.841483723786246e-07,
+      "loss": 0.2354,
+      "step": 3834
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4914165517828563,
+      "learning_rate": 8.840822591856167e-07,
+      "loss": 0.4501,
+      "step": 3835
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0371190255382727,
+      "learning_rate": 8.840161296068876e-07,
+      "loss": 0.2488,
+      "step": 3836
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5973299600775365,
+      "learning_rate": 8.839499836452582e-07,
+      "loss": 0.3872,
+      "step": 3837
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.395174487770906,
+      "learning_rate": 8.838838213035509e-07,
+      "loss": 0.1156,
+      "step": 3838
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.356218367510238,
+      "learning_rate": 8.83817642584588e-07,
+      "loss": 0.1328,
+      "step": 3839
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0232113436227914,
+      "learning_rate": 8.837514474911929e-07,
+      "loss": 0.0872,
+      "step": 3840
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6967371947473936,
+      "learning_rate": 8.836852360261895e-07,
+      "loss": 0.1095,
+      "step": 3841
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6899814270235796,
+      "learning_rate": 8.836190081924027e-07,
+      "loss": 0.0583,
+      "step": 3842
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.47253467787743975,
+      "learning_rate": 8.835527639926579e-07,
+      "loss": 0.1843,
+      "step": 3843
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6822384251805668,
+      "learning_rate": 8.834865034297812e-07,
+      "loss": 0.2154,
+      "step": 3844
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5155496044580115,
+      "learning_rate": 8.834202265065993e-07,
+      "loss": 0.0798,
+      "step": 3845
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6233805862568065,
+      "learning_rate": 8.833539332259396e-07,
+      "loss": 0.2562,
+      "step": 3846
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.745255601651051,
+      "learning_rate": 8.832876235906306e-07,
+      "loss": 0.1589,
+      "step": 3847
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.18299751295070765,
+      "learning_rate": 8.832212976035012e-07,
+      "loss": 0.0722,
+      "step": 3848
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.37131939273704917,
+      "learning_rate": 8.831549552673807e-07,
+      "loss": 0.0921,
+      "step": 3849
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0402267070631617,
+      "learning_rate": 8.830885965850998e-07,
+      "loss": 0.1206,
+      "step": 3850
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8733116472389192,
+      "learning_rate": 8.83022221559489e-07,
+      "loss": 0.3168,
+      "step": 3851
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.158788411492282,
+      "learning_rate": 8.829558301933804e-07,
+      "loss": 0.1649,
+      "step": 3852
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7594693834484741,
+      "learning_rate": 8.828894224896062e-07,
+      "loss": 0.3117,
+      "step": 3853
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.00607259640321,
+      "learning_rate": 8.828229984509996e-07,
+      "loss": 0.3921,
+      "step": 3854
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3712145253287933,
+      "learning_rate": 8.827565580803941e-07,
+      "loss": 0.1097,
+      "step": 3855
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5518431096903318,
+      "learning_rate": 8.826901013806245e-07,
+      "loss": 0.0852,
+      "step": 3856
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.260101596858923,
+      "learning_rate": 8.826236283545259e-07,
+      "loss": 0.2245,
+      "step": 3857
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9216122789470282,
+      "learning_rate": 8.825571390049343e-07,
+      "loss": 0.141,
+      "step": 3858
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6574770388498139,
+      "learning_rate": 8.824906333346858e-07,
+      "loss": 0.1655,
+      "step": 3859
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4896930052088437,
+      "learning_rate": 8.824241113466182e-07,
+      "loss": 0.129,
+      "step": 3860
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.045338539757439,
+      "learning_rate": 8.823575730435693e-07,
+      "loss": 0.1246,
+      "step": 3861
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5777459068356137,
+      "learning_rate": 8.822910184283776e-07,
+      "loss": 0.2317,
+      "step": 3862
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7493117669883436,
+      "learning_rate": 8.822244475038825e-07,
+      "loss": 0.2838,
+      "step": 3863
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7158063307446045,
+      "learning_rate": 8.821578602729241e-07,
+      "loss": 0.0901,
+      "step": 3864
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9323696081863153,
+      "learning_rate": 8.820912567383432e-07,
+      "loss": 0.4398,
+      "step": 3865
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5992314213267559,
+      "learning_rate": 8.820246369029812e-07,
+      "loss": 0.106,
+      "step": 3866
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5060150979246018,
+      "learning_rate": 8.819580007696802e-07,
+      "loss": 0.1455,
+      "step": 3867
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1236957189997694,
+      "learning_rate": 8.818913483412831e-07,
+      "loss": 0.3267,
+      "step": 3868
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7399667473326889,
+      "learning_rate": 8.818246796206332e-07,
+      "loss": 0.0643,
+      "step": 3869
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5064265662293846,
+      "learning_rate": 8.817579946105751e-07,
+      "loss": 0.148,
+      "step": 3870
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6474326731435445,
+      "learning_rate": 8.816912933139535e-07,
+      "loss": 0.2105,
+      "step": 3871
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8501109047346274,
+      "learning_rate": 8.81624575733614e-07,
+      "loss": 0.3672,
+      "step": 3872
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7323724092358175,
+      "learning_rate": 8.81557841872403e-07,
+      "loss": 0.317,
+      "step": 3873
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2862806210495028,
+      "learning_rate": 8.814910917331673e-07,
+      "loss": 0.0497,
+      "step": 3874
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3478016951604009,
+      "learning_rate": 8.814243253187548e-07,
+      "loss": 0.0718,
+      "step": 3875
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.38831128658614433,
+      "learning_rate": 8.813575426320139e-07,
+      "loss": 0.1864,
+      "step": 3876
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9050264415368248,
+      "learning_rate": 8.812907436757935e-07,
+      "loss": 0.2218,
+      "step": 3877
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9137963310217068,
+      "learning_rate": 8.812239284529435e-07,
+      "loss": 0.0833,
+      "step": 3878
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2065487505746293,
+      "learning_rate": 8.811570969663144e-07,
+      "loss": 0.1434,
+      "step": 3879
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5826485797134789,
+      "learning_rate": 8.810902492187573e-07,
+      "loss": 0.1305,
+      "step": 3880
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5433145356257765,
+      "learning_rate": 8.81023385213124e-07,
+      "loss": 0.1828,
+      "step": 3881
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8426871437234196,
+      "learning_rate": 8.809565049522671e-07,
+      "loss": 0.2539,
+      "step": 3882
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6029357850617248,
+      "learning_rate": 8.8088960843904e-07,
+      "loss": 0.082,
+      "step": 3883
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8689748220497369,
+      "learning_rate": 8.808226956762967e-07,
+      "loss": 0.3289,
+      "step": 3884
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5100034663399725,
+      "learning_rate": 8.807557666668914e-07,
+      "loss": 0.3517,
+      "step": 3885
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9071223821693056,
+      "learning_rate": 8.806888214136798e-07,
+      "loss": 0.1622,
+      "step": 3886
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.25651385324206816,
+      "learning_rate": 8.806218599195177e-07,
+      "loss": 0.1169,
+      "step": 3887
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6205044009298561,
+      "learning_rate": 8.805548821872619e-07,
+      "loss": 0.3588,
+      "step": 3888
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2051808882190857,
+      "learning_rate": 8.8048788821977e-07,
+      "loss": 0.0084,
+      "step": 3889
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5313437251721552,
+      "learning_rate": 8.804208780198997e-07,
+      "loss": 0.156,
+      "step": 3890
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5608721803888229,
+      "learning_rate": 8.8035385159051e-07,
+      "loss": 0.1737,
+      "step": 3891
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.114285425950021,
+      "learning_rate": 8.802868089344605e-07,
+      "loss": 0.0337,
+      "step": 3892
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1108144278333296,
+      "learning_rate": 8.802197500546112e-07,
+      "loss": 0.2733,
+      "step": 3893
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6756606510327932,
+      "learning_rate": 8.801526749538231e-07,
+      "loss": 0.0114,
+      "step": 3894
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7827544871261352,
+      "learning_rate": 8.800855836349577e-07,
+      "loss": 0.1278,
+      "step": 3895
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8810978403334839,
+      "learning_rate": 8.800184761008771e-07,
+      "loss": 0.2304,
+      "step": 3896
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0625358915168617,
+      "learning_rate": 8.799513523544444e-07,
+      "loss": 0.0853,
+      "step": 3897
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8936459555487971,
+      "learning_rate": 8.798842123985233e-07,
+      "loss": 0.1544,
+      "step": 3898
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.27459104503815157,
+      "learning_rate": 8.798170562359779e-07,
+      "loss": 0.1053,
+      "step": 3899
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0829617560998321,
+      "learning_rate": 8.797498838696735e-07,
+      "loss": 0.3077,
+      "step": 3900
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4402056171559821,
+      "learning_rate": 8.796826953024756e-07,
+      "loss": 0.1567,
+      "step": 3901
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.625065524314885,
+      "learning_rate": 8.796154905372506e-07,
+      "loss": 0.1401,
+      "step": 3902
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.09889685619163836,
+      "learning_rate": 8.795482695768656e-07,
+      "loss": 0.006,
+      "step": 3903
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.40167191233341654,
+      "learning_rate": 8.794810324241886e-07,
+      "loss": 0.1823,
+      "step": 3904
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9797198670885936,
+      "learning_rate": 8.794137790820879e-07,
+      "loss": 0.0842,
+      "step": 3905
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6342464441119157,
+      "learning_rate": 8.793465095534327e-07,
+      "loss": 0.0054,
+      "step": 3906
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8674351748185777,
+      "learning_rate": 8.792792238410926e-07,
+      "loss": 0.2681,
+      "step": 3907
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8127929286039683,
+      "learning_rate": 8.792119219479386e-07,
+      "loss": 0.3621,
+      "step": 3908
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.998942057555394,
+      "learning_rate": 8.791446038768415e-07,
+      "loss": 0.1985,
+      "step": 3909
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.733145838965004,
+      "learning_rate": 8.790772696306736e-07,
+      "loss": 0.0689,
+      "step": 3910
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8314182062716361,
+      "learning_rate": 8.790099192123073e-07,
+      "loss": 0.3592,
+      "step": 3911
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.429362481326908,
+      "learning_rate": 8.789425526246159e-07,
+      "loss": 0.0961,
+      "step": 3912
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5162858269850434,
+      "learning_rate": 8.788751698704734e-07,
+      "loss": 0.0957,
+      "step": 3913
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.778633990635694,
+      "learning_rate": 8.788077709527546e-07,
+      "loss": 0.1796,
+      "step": 3914
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2358032788236904,
+      "learning_rate": 8.787403558743347e-07,
+      "loss": 0.0903,
+      "step": 3915
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6581315004821515,
+      "learning_rate": 8.786729246380899e-07,
+      "loss": 0.006,
+      "step": 3916
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9809002158680535,
+      "learning_rate": 8.786054772468968e-07,
+      "loss": 0.1373,
+      "step": 3917
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0868837241043616,
+      "learning_rate": 8.785380137036331e-07,
+      "loss": 0.3455,
+      "step": 3918
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.33965775754026445,
+      "learning_rate": 8.784705340111767e-07,
+      "loss": 0.1726,
+      "step": 3919
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3881610503078679,
+      "learning_rate": 8.784030381724066e-07,
+      "loss": 0.1477,
+      "step": 3920
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2779562197772847,
+      "learning_rate": 8.783355261902021e-07,
+      "loss": 0.3554,
+      "step": 3921
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.46884013910490463,
+      "learning_rate": 8.782679980674436e-07,
+      "loss": 0.1424,
+      "step": 3922
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4142878280564671,
+      "learning_rate": 8.782004538070118e-07,
+      "loss": 0.049,
+      "step": 3923
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.435630693591566,
+      "learning_rate": 8.781328934117885e-07,
+      "loss": 0.1453,
+      "step": 3924
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7895980903877988,
+      "learning_rate": 8.780653168846556e-07,
+      "loss": 0.1049,
+      "step": 3925
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.078528448372595,
+      "learning_rate": 8.779977242284964e-07,
+      "loss": 0.1944,
+      "step": 3926
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3513130449104446,
+      "learning_rate": 8.779301154461945e-07,
+      "loss": 0.0848,
+      "step": 3927
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6341401280500651,
+      "learning_rate": 8.778624905406339e-07,
+      "loss": 0.0727,
+      "step": 3928
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1356196801368217,
+      "learning_rate": 8.777948495147e-07,
+      "loss": 0.0199,
+      "step": 3929
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.036735641930418,
+      "learning_rate": 8.777271923712783e-07,
+      "loss": 0.1751,
+      "step": 3930
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9435075619147448,
+      "learning_rate": 8.776595191132553e-07,
+      "loss": 0.3591,
+      "step": 3931
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6932852345179985,
+      "learning_rate": 8.775918297435181e-07,
+      "loss": 0.2179,
+      "step": 3932
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6888035519656425,
+      "learning_rate": 8.775241242649543e-07,
+      "loss": 0.0518,
+      "step": 3933
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5971998529801377,
+      "learning_rate": 8.774564026804525e-07,
+      "loss": 0.231,
+      "step": 3934
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1419428339338078,
+      "learning_rate": 8.773886649929017e-07,
+      "loss": 0.1401,
+      "step": 3935
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.627408617573541,
+      "learning_rate": 8.773209112051918e-07,
+      "loss": 0.1987,
+      "step": 3936
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9453981933126301,
+      "learning_rate": 8.772531413202133e-07,
+      "loss": 0.2172,
+      "step": 3937
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3180858625553908,
+      "learning_rate": 8.771853553408575e-07,
+      "loss": 0.1285,
+      "step": 3938
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3865235113430008,
+      "learning_rate": 8.771175532700162e-07,
+      "loss": 0.158,
+      "step": 3939
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9173031682855154,
+      "learning_rate": 8.770497351105819e-07,
+      "loss": 0.2518,
+      "step": 3940
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6388942602289808,
+      "learning_rate": 8.76981900865448e-07,
+      "loss": 0.1793,
+      "step": 3941
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5680970316244751,
+      "learning_rate": 8.769140505375083e-07,
+      "loss": 0.1978,
+      "step": 3942
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9833493507997556,
+      "learning_rate": 8.768461841296577e-07,
+      "loss": 0.2289,
+      "step": 3943
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8079163924290765,
+      "learning_rate": 8.767783016447912e-07,
+      "loss": 0.0793,
+      "step": 3944
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5906476527055019,
+      "learning_rate": 8.767104030858049e-07,
+      "loss": 0.1805,
+      "step": 3945
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6493009863255066,
+      "learning_rate": 8.766424884555956e-07,
+      "loss": 0.2852,
+      "step": 3946
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8273038894585243,
+      "learning_rate": 8.765745577570606e-07,
+      "loss": 0.2025,
+      "step": 3947
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.44501179601301144,
+      "learning_rate": 8.765066109930979e-07,
+      "loss": 0.1043,
+      "step": 3948
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5612978662180629,
+      "learning_rate": 8.764386481666062e-07,
+      "loss": 0.0897,
+      "step": 3949
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6603623168005797,
+      "learning_rate": 8.763706692804852e-07,
+      "loss": 0.3171,
+      "step": 3950
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1833506083656713,
+      "learning_rate": 8.763026743376347e-07,
+      "loss": 0.264,
+      "step": 3951
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.456151883137029,
+      "learning_rate": 8.762346633409559e-07,
+      "loss": 0.2565,
+      "step": 3952
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9147030031313685,
+      "learning_rate": 8.761666362933497e-07,
+      "loss": 0.2614,
+      "step": 3953
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6258556487063822,
+      "learning_rate": 8.760985931977189e-07,
+      "loss": 0.3442,
+      "step": 3954
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5065709907473496,
+      "learning_rate": 8.76030534056966e-07,
+      "loss": 0.3571,
+      "step": 3955
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7141061511668348,
+      "learning_rate": 8.759624588739945e-07,
+      "loss": 0.0737,
+      "step": 3956
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8226379117225698,
+      "learning_rate": 8.758943676517088e-07,
+      "loss": 0.1746,
+      "step": 3957
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.564638146772833,
+      "learning_rate": 8.758262603930137e-07,
+      "loss": 0.1079,
+      "step": 3958
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.27651579401960985,
+      "learning_rate": 8.757581371008148e-07,
+      "loss": 0.1224,
+      "step": 3959
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3603241503001796,
+      "learning_rate": 8.756899977780185e-07,
+      "loss": 0.1384,
+      "step": 3960
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7128421237886637,
+      "learning_rate": 8.756218424275315e-07,
+      "loss": 0.3432,
+      "step": 3961
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5427821502755281,
+      "learning_rate": 8.755536710522617e-07,
+      "loss": 0.109,
+      "step": 3962
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8949546193401502,
+      "learning_rate": 8.754854836551173e-07,
+      "loss": 0.1083,
+      "step": 3963
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7523697441707426,
+      "learning_rate": 8.754172802390074e-07,
+      "loss": 0.2896,
+      "step": 3964
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.470811108919015,
+      "learning_rate": 8.753490608068416e-07,
+      "loss": 0.204,
+      "step": 3965
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.859132997973097,
+      "learning_rate": 8.752808253615305e-07,
+      "loss": 0.1744,
+      "step": 3966
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7903062271790547,
+      "learning_rate": 8.752125739059848e-07,
+      "loss": 0.0353,
+      "step": 3967
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6433512424106262,
+      "learning_rate": 8.751443064431164e-07,
+      "loss": 0.3293,
+      "step": 3968
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9272328463302985,
+      "learning_rate": 8.750760229758378e-07,
+      "loss": 0.0907,
+      "step": 3969
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0245838538405079,
+      "learning_rate": 8.750077235070621e-07,
+      "loss": 0.3167,
+      "step": 3970
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6949014063101808,
+      "learning_rate": 8.74939408039703e-07,
+      "loss": 0.3784,
+      "step": 3971
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16563874919046356,
+      "learning_rate": 8.748710765766752e-07,
+      "loss": 0.0028,
+      "step": 3972
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2013521468154433,
+      "learning_rate": 8.748027291208934e-07,
+      "loss": 0.0904,
+      "step": 3973
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6037748857188424,
+      "learning_rate": 8.747343656752739e-07,
+      "loss": 0.4494,
+      "step": 3974
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3484933915549889,
+      "learning_rate": 8.74665986242733e-07,
+      "loss": 0.3483,
+      "step": 3975
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8444821672923177,
+      "learning_rate": 8.74597590826188e-07,
+      "loss": 0.1801,
+      "step": 3976
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5820251833319969,
+      "learning_rate": 8.745291794285568e-07,
+      "loss": 0.1341,
+      "step": 3977
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5458408601507707,
+      "learning_rate": 8.744607520527577e-07,
+      "loss": 0.1478,
+      "step": 3978
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5041824124099171,
+      "learning_rate": 8.743923087017102e-07,
+      "loss": 0.2239,
+      "step": 3979
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6601179844297378,
+      "learning_rate": 8.743238493783343e-07,
+      "loss": 0.3004,
+      "step": 3980
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8275809964913583,
+      "learning_rate": 8.742553740855505e-07,
+      "loss": 0.2218,
+      "step": 3981
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3065065175720807,
+      "learning_rate": 8.741868828262802e-07,
+      "loss": 0.145,
+      "step": 3982
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8819701262901135,
+      "learning_rate": 8.741183756034449e-07,
+      "loss": 0.2202,
+      "step": 3983
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0010176950752636,
+      "learning_rate": 8.74049852419968e-07,
+      "loss": 0.3801,
+      "step": 3984
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5025045189329125,
+      "learning_rate": 8.739813132787723e-07,
+      "loss": 0.008,
+      "step": 3985
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6235355600080691,
+      "learning_rate": 8.73912758182782e-07,
+      "loss": 0.1102,
+      "step": 3986
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9359451485156095,
+      "learning_rate": 8.738441871349219e-07,
+      "loss": 0.1621,
+      "step": 3987
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9931543106272482,
+      "learning_rate": 8.737756001381169e-07,
+      "loss": 0.1736,
+      "step": 3988
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.48887747494279876,
+      "learning_rate": 8.737069971952937e-07,
+      "loss": 0.3141,
+      "step": 3989
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3276454607633839,
+      "learning_rate": 8.736383783093787e-07,
+      "loss": 0.4351,
+      "step": 3990
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6281135719184645,
+      "learning_rate": 8.735697434832993e-07,
+      "loss": 0.1919,
+      "step": 3991
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9004611249581926,
+      "learning_rate": 8.735010927199837e-07,
+      "loss": 0.1233,
+      "step": 3992
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6245121951560472,
+      "learning_rate": 8.734324260223608e-07,
+      "loss": 0.01,
+      "step": 3993
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8564528634864154,
+      "learning_rate": 8.733637433933598e-07,
+      "loss": 0.3092,
+      "step": 3994
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6624523979672308,
+      "learning_rate": 8.732950448359109e-07,
+      "loss": 0.2709,
+      "step": 3995
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.876141690078714,
+      "learning_rate": 8.732263303529451e-07,
+      "loss": 0.2849,
+      "step": 3996
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0187704625173444,
+      "learning_rate": 8.731575999473936e-07,
+      "loss": 0.2479,
+      "step": 3997
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9978460960576566,
+      "learning_rate": 8.73088853622189e-07,
+      "loss": 0.146,
+      "step": 3998
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8034763387545482,
+      "learning_rate": 8.730200913802637e-07,
+      "loss": 0.1926,
+      "step": 3999
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4719686255204056,
+      "learning_rate": 8.729513132245514e-07,
+      "loss": 0.231,
+      "step": 4000
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3168863902789227,
+      "learning_rate": 8.728825191579865e-07,
+      "loss": 0.0533,
+      "step": 4001
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5907737292904186,
+      "learning_rate": 8.728137091835038e-07,
+      "loss": 0.1546,
+      "step": 4002
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.43436093732337505,
+      "learning_rate": 8.727448833040386e-07,
+      "loss": 0.1933,
+      "step": 4003
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.182335143143738,
+      "learning_rate": 8.726760415225274e-07,
+      "loss": 0.0765,
+      "step": 4004
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8972510251624722,
+      "learning_rate": 8.726071838419072e-07,
+      "loss": 0.1038,
+      "step": 4005
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6094007452363491,
+      "learning_rate": 8.725383102651153e-07,
+      "loss": 0.1785,
+      "step": 4006
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.29551837758871335,
+      "learning_rate": 8.724694207950905e-07,
+      "loss": 0.1089,
+      "step": 4007
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1253683787788942,
+      "learning_rate": 8.724005154347713e-07,
+      "loss": 0.2786,
+      "step": 4008
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.423186108471077,
+      "learning_rate": 8.723315941870974e-07,
+      "loss": 0.1108,
+      "step": 4009
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.48852532601750126,
+      "learning_rate": 8.722626570550093e-07,
+      "loss": 0.2561,
+      "step": 4010
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7463282720614175,
+      "learning_rate": 8.72193704041448e-07,
+      "loss": 0.3525,
+      "step": 4011
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5450143580604226,
+      "learning_rate": 8.72124735149355e-07,
+      "loss": 0.2108,
+      "step": 4012
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9253499962593161,
+      "learning_rate": 8.720557503816728e-07,
+      "loss": 0.3795,
+      "step": 4013
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.46562963399667523,
+      "learning_rate": 8.719867497413443e-07,
+      "loss": 0.007,
+      "step": 4014
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5332682655484945,
+      "learning_rate": 8.719177332313131e-07,
+      "loss": 0.186,
+      "step": 4015
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5535338783673729,
+      "learning_rate": 8.71848700854524e-07,
+      "loss": 0.1613,
+      "step": 4016
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6403299260618557,
+      "learning_rate": 8.717796526139217e-07,
+      "loss": 0.2925,
+      "step": 4017
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6254917532022537,
+      "learning_rate": 8.71710588512452e-07,
+      "loss": 0.2384,
+      "step": 4018
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5750364342194938,
+      "learning_rate": 8.716415085530615e-07,
+      "loss": 0.0331,
+      "step": 4019
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4569063266582358,
+      "learning_rate": 8.71572412738697e-07,
+      "loss": 0.3298,
+      "step": 4020
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.891163346975924,
+      "learning_rate": 8.715033010723065e-07,
+      "loss": 0.0961,
+      "step": 4021
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3413829687791063,
+      "learning_rate": 8.714341735568384e-07,
+      "loss": 0.0372,
+      "step": 4022
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.49756565405090214,
+      "learning_rate": 8.713650301952418e-07,
+      "loss": 0.0951,
+      "step": 4023
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9182125512484611,
+      "learning_rate": 8.712958709904664e-07,
+      "loss": 0.2331,
+      "step": 4024
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9725612485583259,
+      "learning_rate": 8.712266959454628e-07,
+      "loss": 0.2,
+      "step": 4025
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4010945358686009,
+      "learning_rate": 8.711575050631821e-07,
+      "loss": 0.031,
+      "step": 4026
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5290640232322119,
+      "learning_rate": 8.71088298346576e-07,
+      "loss": 0.2232,
+      "step": 4027
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.22726065289959244,
+      "learning_rate": 8.710190757985971e-07,
+      "loss": 0.0997,
+      "step": 4028
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5089060046635212,
+      "learning_rate": 8.709498374221987e-07,
+      "loss": 0.0809,
+      "step": 4029
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8505876302845143,
+      "learning_rate": 8.708805832203345e-07,
+      "loss": 0.1974,
+      "step": 4030
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5800888585351849,
+      "learning_rate": 8.708113131959591e-07,
+      "loss": 0.3174,
+      "step": 4031
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2127749732338962,
+      "learning_rate": 8.707420273520276e-07,
+      "loss": 0.2511,
+      "step": 4032
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9619904283254103,
+      "learning_rate": 8.706727256914959e-07,
+      "loss": 0.2079,
+      "step": 4033
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.788683859357756,
+      "learning_rate": 8.706034082173205e-07,
+      "loss": 0.2252,
+      "step": 4034
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8666167015743347,
+      "learning_rate": 8.70534074932459e-07,
+      "loss": 0.2964,
+      "step": 4035
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8339818264528626,
+      "learning_rate": 8.704647258398688e-07,
+      "loss": 0.1802,
+      "step": 4036
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7500161946388975,
+      "learning_rate": 8.703953609425087e-07,
+      "loss": 0.234,
+      "step": 4037
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0373680778071752,
+      "learning_rate": 8.703259802433378e-07,
+      "loss": 0.0323,
+      "step": 4038
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7686844705049926,
+      "learning_rate": 8.702565837453163e-07,
+      "loss": 0.1601,
+      "step": 4039
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.34216488586559624,
+      "learning_rate": 8.701871714514046e-07,
+      "loss": 0.0436,
+      "step": 4040
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7622924070033409,
+      "learning_rate": 8.701177433645639e-07,
+      "loss": 0.3274,
+      "step": 4041
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.876673436189718,
+      "learning_rate": 8.700482994877563e-07,
+      "loss": 0.1475,
+      "step": 4042
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3006986467282531,
+      "learning_rate": 8.699788398239442e-07,
+      "loss": 0.0065,
+      "step": 4043
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.35755714554411305,
+      "learning_rate": 8.699093643760913e-07,
+      "loss": 0.1083,
+      "step": 4044
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9828179439260103,
+      "learning_rate": 8.698398731471612e-07,
+      "loss": 0.1531,
+      "step": 4045
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.39576422761612184,
+      "learning_rate": 8.697703661401185e-07,
+      "loss": 0.2006,
+      "step": 4046
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8437275424897863,
+      "learning_rate": 8.697008433579289e-07,
+      "loss": 0.2849,
+      "step": 4047
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7544173146891293,
+      "learning_rate": 8.69631304803558e-07,
+      "loss": 0.3707,
+      "step": 4048
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7168024231112956,
+      "learning_rate": 8.695617504799726e-07,
+      "loss": 0.1958,
+      "step": 4049
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.19259379309971653,
+      "learning_rate": 8.694921803901401e-07,
+      "loss": 0.0996,
+      "step": 4050
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5194265600276591,
+      "learning_rate": 8.694225945370282e-07,
+      "loss": 0.0215,
+      "step": 4051
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6012828951597724,
+      "learning_rate": 8.693529929236058e-07,
+      "loss": 0.2555,
+      "step": 4052
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8690152717732422,
+      "learning_rate": 8.692833755528424e-07,
+      "loss": 0.0856,
+      "step": 4053
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3277992090830388,
+      "learning_rate": 8.692137424277079e-07,
+      "loss": 0.3708,
+      "step": 4054
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5039056616220217,
+      "learning_rate": 8.691440935511726e-07,
+      "loss": 0.3282,
+      "step": 4055
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.42331557662746544,
+      "learning_rate": 8.690744289262086e-07,
+      "loss": 0.3823,
+      "step": 4056
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9142575813248571,
+      "learning_rate": 8.690047485557873e-07,
+      "loss": 0.2637,
+      "step": 4057
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7196566229620287,
+      "learning_rate": 8.689350524428817e-07,
+      "loss": 0.24,
+      "step": 4058
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2910743427622435,
+      "learning_rate": 8.688653405904651e-07,
+      "loss": 0.2829,
+      "step": 4059
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9113426193291279,
+      "learning_rate": 8.687956130015115e-07,
+      "loss": 0.2902,
+      "step": 4060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5598130609352595,
+      "learning_rate": 8.687258696789957e-07,
+      "loss": 0.1831,
+      "step": 4061
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5193925121894146,
+      "learning_rate": 8.686561106258932e-07,
+      "loss": 0.0891,
+      "step": 4062
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6753090305556863,
+      "learning_rate": 8.685863358451797e-07,
+      "loss": 0.2076,
+      "step": 4063
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3933141989346814,
+      "learning_rate": 8.685165453398323e-07,
+      "loss": 0.1361,
+      "step": 4064
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9869841152758954,
+      "learning_rate": 8.684467391128283e-07,
+      "loss": 0.0818,
+      "step": 4065
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5617361231289953,
+      "learning_rate": 8.683769171671456e-07,
+      "loss": 0.0773,
+      "step": 4066
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9474663227051654,
+      "learning_rate": 8.683070795057632e-07,
+      "loss": 0.1526,
+      "step": 4067
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.08787256262239636,
+      "learning_rate": 8.682372261316603e-07,
+      "loss": 0.0033,
+      "step": 4068
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5999741487371676,
+      "learning_rate": 8.681673570478172e-07,
+      "loss": 0.0251,
+      "step": 4069
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5186519952801143,
+      "learning_rate": 8.680974722572144e-07,
+      "loss": 0.1501,
+      "step": 4070
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4551390871962548,
+      "learning_rate": 8.680275717628336e-07,
+      "loss": 0.185,
+      "step": 4071
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3686354176626132,
+      "learning_rate": 8.679576555676566e-07,
+      "loss": 0.1716,
+      "step": 4072
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7757396827823145,
+      "learning_rate": 8.678877236746664e-07,
+      "loss": 0.1621,
+      "step": 4073
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8828172389719617,
+      "learning_rate": 8.678177760868466e-07,
+      "loss": 0.2308,
+      "step": 4074
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6983117036471638,
+      "learning_rate": 8.677478128071808e-07,
+      "loss": 0.2866,
+      "step": 4075
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.38771195245275397,
+      "learning_rate": 8.676778338386541e-07,
+      "loss": 0.1309,
+      "step": 4076
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6801236831527018,
+      "learning_rate": 8.676078391842518e-07,
+      "loss": 0.2795,
+      "step": 4077
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5694146176378214,
+      "learning_rate": 8.675378288469603e-07,
+      "loss": 0.175,
+      "step": 4078
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6219395904385054,
+      "learning_rate": 8.674678028297659e-07,
+      "loss": 0.247,
+      "step": 4079
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.34342221249927024,
+      "learning_rate": 8.673977611356566e-07,
+      "loss": 0.1954,
+      "step": 4080
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7600292127363575,
+      "learning_rate": 8.673277037676201e-07,
+      "loss": 0.3134,
+      "step": 4081
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3481644163119903,
+      "learning_rate": 8.672576307286455e-07,
+      "loss": 0.0081,
+      "step": 4082
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6503751638113178,
+      "learning_rate": 8.671875420217218e-07,
+      "loss": 0.0743,
+      "step": 4083
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1873273700527,
+      "learning_rate": 8.671174376498396e-07,
+      "loss": 0.3154,
+      "step": 4084
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8396698016655777,
+      "learning_rate": 8.670473176159896e-07,
+      "loss": 0.0073,
+      "step": 4085
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6112612695587594,
+      "learning_rate": 8.66977181923163e-07,
+      "loss": 0.3248,
+      "step": 4086
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7350065130972876,
+      "learning_rate": 8.669070305743522e-07,
+      "loss": 0.1309,
+      "step": 4087
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0898361003584103,
+      "learning_rate": 8.668368635725498e-07,
+      "loss": 0.2625,
+      "step": 4088
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5157372002681769,
+      "learning_rate": 8.667666809207494e-07,
+      "loss": 0.1469,
+      "step": 4089
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7806264136291773,
+      "learning_rate": 8.666964826219452e-07,
+      "loss": 0.221,
+      "step": 4090
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.913072198088876,
+      "learning_rate": 8.666262686791317e-07,
+      "loss": 0.4681,
+      "step": 4091
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.764650493860681,
+      "learning_rate": 8.665560390953048e-07,
+      "loss": 0.4531,
+      "step": 4092
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8470516436514497,
+      "learning_rate": 8.664857938734603e-07,
+      "loss": 0.4609,
+      "step": 4093
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3627619097956996,
+      "learning_rate": 8.664155330165951e-07,
+      "loss": 0.315,
+      "step": 4094
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9197571070088587,
+      "learning_rate": 8.663452565277066e-07,
+      "loss": 0.4106,
+      "step": 4095
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4429333737292287,
+      "learning_rate": 8.66274964409793e-07,
+      "loss": 0.1465,
+      "step": 4096
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3301824337049717,
+      "learning_rate": 8.662046566658534e-07,
+      "loss": 0.155,
+      "step": 4097
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8237105848216212,
+      "learning_rate": 8.661343332988868e-07,
+      "loss": 0.3154,
+      "step": 4098
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8642760805583275,
+      "learning_rate": 8.660639943118935e-07,
+      "loss": 0.2097,
+      "step": 4099
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9988221169075769,
+      "learning_rate": 8.659936397078742e-07,
+      "loss": 0.1076,
+      "step": 4100
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.08165401706748,
+      "learning_rate": 8.659232694898306e-07,
+      "loss": 0.1478,
+      "step": 4101
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9392110429916637,
+      "learning_rate": 8.658528836607648e-07,
+      "loss": 0.2727,
+      "step": 4102
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6255102925376163,
+      "learning_rate": 8.657824822236794e-07,
+      "loss": 0.1134,
+      "step": 4103
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.531671919199018,
+      "learning_rate": 8.657120651815781e-07,
+      "loss": 0.097,
+      "step": 4104
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0349825853821242,
+      "learning_rate": 8.656416325374649e-07,
+      "loss": 0.248,
+      "step": 4105
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.41708388544839,
+      "learning_rate": 8.655711842943446e-07,
+      "loss": 0.2295,
+      "step": 4106
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.31534042476054,
+      "learning_rate": 8.655007204552227e-07,
+      "loss": 0.3639,
+      "step": 4107
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.40837283380562023,
+      "learning_rate": 8.654302410231054e-07,
+      "loss": 0.2332,
+      "step": 4108
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.30675613563980453,
+      "learning_rate": 8.653597460009993e-07,
+      "loss": 0.0117,
+      "step": 4109
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5425657633918686,
+      "learning_rate": 8.652892353919119e-07,
+      "loss": 0.0441,
+      "step": 4110
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4927982659656652,
+      "learning_rate": 8.652187091988516e-07,
+      "loss": 0.1758,
+      "step": 4111
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8854424625965419,
+      "learning_rate": 8.651481674248267e-07,
+      "loss": 0.4199,
+      "step": 4112
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7812687945958526,
+      "learning_rate": 8.650776100728471e-07,
+      "loss": 0.1715,
+      "step": 4113
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0582769701423547,
+      "learning_rate": 8.650070371459228e-07,
+      "loss": 0.0757,
+      "step": 4114
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3886860276188469,
+      "learning_rate": 8.649364486470646e-07,
+      "loss": 0.1217,
+      "step": 4115
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7357733802014849,
+      "learning_rate": 8.648658445792838e-07,
+      "loss": 0.3296,
+      "step": 4116
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7205992295519654,
+      "learning_rate": 8.647952249455925e-07,
+      "loss": 0.1298,
+      "step": 4117
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2698102001347191,
+      "learning_rate": 8.647245897490036e-07,
+      "loss": 0.0348,
+      "step": 4118
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0515583970453326,
+      "learning_rate": 8.646539389925307e-07,
+      "loss": 0.1677,
+      "step": 4119
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5393363410624852,
+      "learning_rate": 8.645832726791876e-07,
+      "loss": 0.1455,
+      "step": 4120
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7251985075075046,
+      "learning_rate": 8.645125908119892e-07,
+      "loss": 0.1008,
+      "step": 4121
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8636367474351939,
+      "learning_rate": 8.644418933939508e-07,
+      "loss": 0.3264,
+      "step": 4122
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6678994209994875,
+      "learning_rate": 8.643711804280888e-07,
+      "loss": 0.2842,
+      "step": 4123
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4542562535018774,
+      "learning_rate": 8.643004519174197e-07,
+      "loss": 0.078,
+      "step": 4124
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.569213117890287,
+      "learning_rate": 8.642297078649608e-07,
+      "loss": 0.2563,
+      "step": 4125
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.886986849481526,
+      "learning_rate": 8.641589482737306e-07,
+      "loss": 0.1211,
+      "step": 4126
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3409962692945223,
+      "learning_rate": 8.640881731467475e-07,
+      "loss": 0.1253,
+      "step": 4127
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0903431349800219,
+      "learning_rate": 8.640173824870311e-07,
+      "loss": 0.0725,
+      "step": 4128
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9233903438867485,
+      "learning_rate": 8.639465762976013e-07,
+      "loss": 0.2513,
+      "step": 4129
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.731236143951161,
+      "learning_rate": 8.638757545814789e-07,
+      "loss": 0.0069,
+      "step": 4130
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9535380261485513,
+      "learning_rate": 8.638049173416855e-07,
+      "loss": 0.2661,
+      "step": 4131
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.341231042585357,
+      "learning_rate": 8.637340645812429e-07,
+      "loss": 0.1597,
+      "step": 4132
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8995777054348525,
+      "learning_rate": 8.636631963031739e-07,
+      "loss": 0.2024,
+      "step": 4133
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16401430139744694,
+      "learning_rate": 8.635923125105018e-07,
+      "loss": 0.0235,
+      "step": 4134
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.843344059109508,
+      "learning_rate": 8.63521413206251e-07,
+      "loss": 0.0331,
+      "step": 4135
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6263545356874659,
+      "learning_rate": 8.634504983934456e-07,
+      "loss": 0.3176,
+      "step": 4136
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4808943140162613,
+      "learning_rate": 8.633795680751116e-07,
+      "loss": 0.1106,
+      "step": 4137
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6112905735795604,
+      "learning_rate": 8.633086222542746e-07,
+      "loss": 0.1585,
+      "step": 4138
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5330909187707842,
+      "learning_rate": 8.632376609339615e-07,
+      "loss": 0.0261,
+      "step": 4139
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5105147410937386,
+      "learning_rate": 8.631666841171995e-07,
+      "loss": 0.225,
+      "step": 4140
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4487846225878657,
+      "learning_rate": 8.630956918070167e-07,
+      "loss": 0.022,
+      "step": 4141
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8339169275027136,
+      "learning_rate": 8.630246840064419e-07,
+      "loss": 0.2695,
+      "step": 4142
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0898084150738283,
+      "learning_rate": 8.629536607185041e-07,
+      "loss": 0.3548,
+      "step": 4143
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2029661526671163,
+      "learning_rate": 8.628826219462337e-07,
+      "loss": 0.0799,
+      "step": 4144
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9993333429873376,
+      "learning_rate": 8.62811567692661e-07,
+      "loss": 0.3005,
+      "step": 4145
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4544022280862907,
+      "learning_rate": 8.627404979608177e-07,
+      "loss": 0.1412,
+      "step": 4146
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.992980608984845,
+      "learning_rate": 8.626694127537354e-07,
+      "loss": 0.1137,
+      "step": 4147
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.205693774665616,
+      "learning_rate": 8.625983120744469e-07,
+      "loss": 0.1241,
+      "step": 4148
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1628846331352964,
+      "learning_rate": 8.625271959259856e-07,
+      "loss": 0.2407,
+      "step": 4149
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.680904559539922,
+      "learning_rate": 8.624560643113852e-07,
+      "loss": 0.3343,
+      "step": 4150
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0543011777212534,
+      "learning_rate": 8.623849172336805e-07,
+      "loss": 0.1855,
+      "step": 4151
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8079673985740236,
+      "learning_rate": 8.623137546959068e-07,
+      "loss": 0.3205,
+      "step": 4152
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8083249109115577,
+      "learning_rate": 8.622425767011e-07,
+      "loss": 0.2182,
+      "step": 4153
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5073675655689701,
+      "learning_rate": 8.621713832522966e-07,
+      "loss": 0.2256,
+      "step": 4154
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6951868169192322,
+      "learning_rate": 8.62100174352534e-07,
+      "loss": 0.2103,
+      "step": 4155
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.46143325777616173,
+      "learning_rate": 8.620289500048501e-07,
+      "loss": 0.0863,
+      "step": 4156
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8654396278114236,
+      "learning_rate": 8.619577102122833e-07,
+      "loss": 0.21,
+      "step": 4157
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.844883718788873,
+      "learning_rate": 8.61886454977873e-07,
+      "loss": 0.1231,
+      "step": 4158
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9131056581684664,
+      "learning_rate": 8.618151843046591e-07,
+      "loss": 0.2919,
+      "step": 4159
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7942533845691214,
+      "learning_rate": 8.617438981956819e-07,
+      "loss": 0.1762,
+      "step": 4160
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5094814576754809,
+      "learning_rate": 8.616725966539831e-07,
+      "loss": 0.1071,
+      "step": 4161
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3741130726984353,
+      "learning_rate": 8.616012796826041e-07,
+      "loss": 0.1949,
+      "step": 4162
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2950232822520885,
+      "learning_rate": 8.615299472845876e-07,
+      "loss": 0.1617,
+      "step": 4163
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1629432705817841,
+      "learning_rate": 8.614585994629769e-07,
+      "loss": 0.072,
+      "step": 4164
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.43310283494739926,
+      "learning_rate": 8.613872362208157e-07,
+      "loss": 0.2264,
+      "step": 4165
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.53920155113599,
+      "learning_rate": 8.613158575611484e-07,
+      "loss": 0.1356,
+      "step": 4166
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.588776252123757,
+      "learning_rate": 8.612444634870204e-07,
+      "loss": 0.1094,
+      "step": 4167
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.255630474149303,
+      "learning_rate": 8.611730540014772e-07,
+      "loss": 0.125,
+      "step": 4168
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1919129974843073,
+      "learning_rate": 8.611016291075656e-07,
+      "loss": 0.1151,
+      "step": 4169
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.23080007066420352,
+      "learning_rate": 8.610301888083327e-07,
+      "loss": 0.129,
+      "step": 4170
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6946707730211346,
+      "learning_rate": 8.60958733106826e-07,
+      "loss": 0.591,
+      "step": 4171
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.45516192914906256,
+      "learning_rate": 8.608872620060943e-07,
+      "loss": 0.1313,
+      "step": 4172
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8978727104176235,
+      "learning_rate": 8.608157755091864e-07,
+      "loss": 0.4418,
+      "step": 4173
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8782263943138546,
+      "learning_rate": 8.607442736191521e-07,
+      "loss": 0.2521,
+      "step": 4174
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8028398729137047,
+      "learning_rate": 8.606727563390421e-07,
+      "loss": 0.188,
+      "step": 4175
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.21838660358954995,
+      "learning_rate": 8.606012236719073e-07,
+      "loss": 0.1483,
+      "step": 4176
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1902942487618677,
+      "learning_rate": 8.605296756207993e-07,
+      "loss": 0.2007,
+      "step": 4177
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.28907669710924344,
+      "learning_rate": 8.604581121887706e-07,
+      "loss": 0.1095,
+      "step": 4178
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.318027359896157,
+      "learning_rate": 8.60386533378874e-07,
+      "loss": 0.3918,
+      "step": 4179
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.751457980254541,
+      "learning_rate": 8.603149391941638e-07,
+      "loss": 0.2503,
+      "step": 4180
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5886811753906173,
+      "learning_rate": 8.602433296376937e-07,
+      "loss": 0.348,
+      "step": 4181
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5060503883283605,
+      "learning_rate": 8.601717047125191e-07,
+      "loss": 0.1095,
+      "step": 4182
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2121079761848257,
+      "learning_rate": 8.601000644216955e-07,
+      "loss": 0.223,
+      "step": 4183
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4177256732661798,
+      "learning_rate": 8.600284087682792e-07,
+      "loss": 0.4123,
+      "step": 4184
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.930620527742801,
+      "learning_rate": 8.599567377553274e-07,
+      "loss": 0.1856,
+      "step": 4185
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5694074812344485,
+      "learning_rate": 8.598850513858975e-07,
+      "loss": 0.1965,
+      "step": 4186
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.343892461363804,
+      "learning_rate": 8.598133496630477e-07,
+      "loss": 0.1279,
+      "step": 4187
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5437361057509675,
+      "learning_rate": 8.597416325898373e-07,
+      "loss": 0.0913,
+      "step": 4188
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.159113119356025,
+      "learning_rate": 8.596699001693255e-07,
+      "loss": 0.6262,
+      "step": 4189
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4849247105383991,
+      "learning_rate": 8.595981524045729e-07,
+      "loss": 0.2941,
+      "step": 4190
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.164368321820298,
+      "learning_rate": 8.595263892986402e-07,
+      "loss": 0.1936,
+      "step": 4191
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2799470303055762,
+      "learning_rate": 8.594546108545888e-07,
+      "loss": 0.4101,
+      "step": 4192
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0023004339422021,
+      "learning_rate": 8.593828170754813e-07,
+      "loss": 0.1801,
+      "step": 4193
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4085905230365117,
+      "learning_rate": 8.593110079643803e-07,
+      "loss": 0.2139,
+      "step": 4194
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5349107399844006,
+      "learning_rate": 8.592391835243495e-07,
+      "loss": 0.4644,
+      "step": 4195
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6752446175755001,
+      "learning_rate": 8.591673437584528e-07,
+      "loss": 0.0281,
+      "step": 4196
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.18305378645340323,
+      "learning_rate": 8.590954886697553e-07,
+      "loss": 0.0904,
+      "step": 4197
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.26359003237301,
+      "learning_rate": 8.590236182613224e-07,
+      "loss": 0.0882,
+      "step": 4198
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.38274271577641134,
+      "learning_rate": 8.589517325362201e-07,
+      "loss": 0.0959,
+      "step": 4199
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.41296319207016935,
+      "learning_rate": 8.588798314975154e-07,
+      "loss": 0.1613,
+      "step": 4200
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2904919767391199,
+      "learning_rate": 8.588079151482756e-07,
+      "loss": 0.1069,
+      "step": 4201
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5138270595551349,
+      "learning_rate": 8.58735983491569e-07,
+      "loss": 0.2439,
+      "step": 4202
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0077304062080574,
+      "learning_rate": 8.58664036530464e-07,
+      "loss": 0.2434,
+      "step": 4203
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.81302314365293,
+      "learning_rate": 8.585920742680304e-07,
+      "loss": 0.4034,
+      "step": 4204
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4243340845075125,
+      "learning_rate": 8.585200967073378e-07,
+      "loss": 0.2181,
+      "step": 4205
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.412047738650341,
+      "learning_rate": 8.584481038514572e-07,
+      "loss": 0.1865,
+      "step": 4206
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7237220104875748,
+      "learning_rate": 8.583760957034601e-07,
+      "loss": 0.3277,
+      "step": 4207
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7640779684273645,
+      "learning_rate": 8.583040722664183e-07,
+      "loss": 0.0173,
+      "step": 4208
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6214385840338672,
+      "learning_rate": 8.582320335434045e-07,
+      "loss": 0.2223,
+      "step": 4209
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9811865311502546,
+      "learning_rate": 8.581599795374918e-07,
+      "loss": 0.1144,
+      "step": 4210
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2670950809825193,
+      "learning_rate": 8.580879102517547e-07,
+      "loss": 0.1149,
+      "step": 4211
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1956863785449852,
+      "learning_rate": 8.580158256892672e-07,
+      "loss": 0.3023,
+      "step": 4212
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2071355650323525,
+      "learning_rate": 8.57943725853105e-07,
+      "loss": 0.4585,
+      "step": 4213
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6563889522905176,
+      "learning_rate": 8.578716107463439e-07,
+      "loss": 0.2213,
+      "step": 4214
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.215855606348245,
+      "learning_rate": 8.577994803720605e-07,
+      "loss": 0.1861,
+      "step": 4215
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8469151423347601,
+      "learning_rate": 8.57727334733332e-07,
+      "loss": 0.2754,
+      "step": 4216
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9923423963328847,
+      "learning_rate": 8.576551738332362e-07,
+      "loss": 0.2037,
+      "step": 4217
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.971817001544454,
+      "learning_rate": 8.575829976748519e-07,
+      "loss": 0.1913,
+      "step": 4218
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8981390336960575,
+      "learning_rate": 8.575108062612579e-07,
+      "loss": 0.1501,
+      "step": 4219
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2470736773943565,
+      "learning_rate": 8.574385995955342e-07,
+      "loss": 0.2306,
+      "step": 4220
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8319293764527479,
+      "learning_rate": 8.573663776807614e-07,
+      "loss": 0.318,
+      "step": 4221
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5835820336388515,
+      "learning_rate": 8.572941405200205e-07,
+      "loss": 0.2691,
+      "step": 4222
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0016540452258798,
+      "learning_rate": 8.572218881163932e-07,
+      "loss": 0.4274,
+      "step": 4223
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.360160874124541,
+      "learning_rate": 8.571496204729623e-07,
+      "loss": 0.0397,
+      "step": 4224
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.42308800478398084,
+      "learning_rate": 8.570773375928104e-07,
+      "loss": 0.2743,
+      "step": 4225
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8029133247581695,
+      "learning_rate": 8.570050394790216e-07,
+      "loss": 0.4695,
+      "step": 4226
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9014662793121879,
+      "learning_rate": 8.569327261346801e-07,
+      "loss": 0.2776,
+      "step": 4227
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.420264671288587,
+      "learning_rate": 8.568603975628709e-07,
+      "loss": 0.2015,
+      "step": 4228
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1059528845144877,
+      "learning_rate": 8.567880537666799e-07,
+      "loss": 0.2098,
+      "step": 4229
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.256259511897296,
+      "learning_rate": 8.567156947491932e-07,
+      "loss": 0.2977,
+      "step": 4230
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.23574349099195155,
+      "learning_rate": 8.56643320513498e-07,
+      "loss": 0.0856,
+      "step": 4231
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.341729503021465,
+      "learning_rate": 8.565709310626816e-07,
+      "loss": 0.1234,
+      "step": 4232
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.454600428281282,
+      "learning_rate": 8.564985263998325e-07,
+      "loss": 0.2918,
+      "step": 4233
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7518635243267678,
+      "learning_rate": 8.564261065280398e-07,
+      "loss": 0.0767,
+      "step": 4234
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8384725035293839,
+      "learning_rate": 8.563536714503928e-07,
+      "loss": 0.1559,
+      "step": 4235
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.800838341046286,
+      "learning_rate": 8.562812211699816e-07,
+      "loss": 0.1743,
+      "step": 4236
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8013847848128576,
+      "learning_rate": 8.562087556898976e-07,
+      "loss": 0.2045,
+      "step": 4237
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.31272994473471644,
+      "learning_rate": 8.56136275013232e-07,
+      "loss": 0.2826,
+      "step": 4238
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.49798282015021056,
+      "learning_rate": 8.560637791430768e-07,
+      "loss": 0.1582,
+      "step": 4239
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.621320727025723,
+      "learning_rate": 8.559912680825252e-07,
+      "loss": 0.3903,
+      "step": 4240
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.28427224685192864,
+      "learning_rate": 8.559187418346702e-07,
+      "loss": 0.166,
+      "step": 4241
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9318865664038639,
+      "learning_rate": 8.558462004026063e-07,
+      "loss": 0.2234,
+      "step": 4242
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4894180794341328,
+      "learning_rate": 8.557736437894283e-07,
+      "loss": 0.1665,
+      "step": 4243
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5977000699130905,
+      "learning_rate": 8.557010719982314e-07,
+      "loss": 0.3489,
+      "step": 4244
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8276768852244196,
+      "learning_rate": 8.556284850321116e-07,
+      "loss": 0.4377,
+      "step": 4245
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5917687601179266,
+      "learning_rate": 8.555558828941658e-07,
+      "loss": 0.1854,
+      "step": 4246
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0923027668181713,
+      "learning_rate": 8.554832655874913e-07,
+      "loss": 0.1182,
+      "step": 4247
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9908405100430573,
+      "learning_rate": 8.554106331151861e-07,
+      "loss": 0.1269,
+      "step": 4248
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9048833651868197,
+      "learning_rate": 8.553379854803488e-07,
+      "loss": 0.196,
+      "step": 4249
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.05747604125339,
+      "learning_rate": 8.552653226860787e-07,
+      "loss": 0.1749,
+      "step": 4250
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6215436704290188,
+      "learning_rate": 8.551926447354758e-07,
+      "loss": 0.1114,
+      "step": 4251
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5683414048674594,
+      "learning_rate": 8.551199516316407e-07,
+      "loss": 0.3162,
+      "step": 4252
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5778537672359699,
+      "learning_rate": 8.550472433776744e-07,
+      "loss": 0.1325,
+      "step": 4253
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3504835916634373,
+      "learning_rate": 8.549745199766791e-07,
+      "loss": 0.1099,
+      "step": 4254
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.18648009972585985,
+      "learning_rate": 8.549017814317572e-07,
+      "loss": 0.0076,
+      "step": 4255
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7118260180527443,
+      "learning_rate": 8.548290277460118e-07,
+      "loss": 0.4571,
+      "step": 4256
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7519858159781565,
+      "learning_rate": 8.547562589225468e-07,
+      "loss": 0.2112,
+      "step": 4257
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2938195918426418,
+      "learning_rate": 8.546834749644666e-07,
+      "loss": 0.2733,
+      "step": 4258
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3827592266894607,
+      "learning_rate": 8.546106758748763e-07,
+      "loss": 0.0178,
+      "step": 4259
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8115191964053644,
+      "learning_rate": 8.545378616568819e-07,
+      "loss": 0.1191,
+      "step": 4260
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.756630858578341,
+      "learning_rate": 8.544650323135895e-07,
+      "loss": 0.007,
+      "step": 4261
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8229417280241946,
+      "learning_rate": 8.543921878481063e-07,
+      "loss": 0.2933,
+      "step": 4262
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1633109845165468,
+      "learning_rate": 8.543193282635399e-07,
+      "loss": 0.209,
+      "step": 4263
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7959031999077241,
+      "learning_rate": 8.542464535629987e-07,
+      "loss": 0.0123,
+      "step": 4264
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.32068562583454563,
+      "learning_rate": 8.541735637495917e-07,
+      "loss": 0.161,
+      "step": 4265
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1121281956030695,
+      "learning_rate": 8.541006588264286e-07,
+      "loss": 0.1259,
+      "step": 4266
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7871506502797747,
+      "learning_rate": 8.540277387966193e-07,
+      "loss": 0.1888,
+      "step": 4267
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3010533321259848,
+      "learning_rate": 8.539548036632751e-07,
+      "loss": 0.1007,
+      "step": 4268
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6552887264914901,
+      "learning_rate": 8.538818534295075e-07,
+      "loss": 0.0432,
+      "step": 4269
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4525615232710638,
+      "learning_rate": 8.538088880984285e-07,
+      "loss": 0.1468,
+      "step": 4270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7399268411262727,
+      "learning_rate": 8.537359076731512e-07,
+      "loss": 0.1885,
+      "step": 4271
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3202130344918199,
+      "learning_rate": 8.53662912156789e-07,
+      "loss": 0.0391,
+      "step": 4272
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5685864471737712,
+      "learning_rate": 8.535899015524559e-07,
+      "loss": 0.1587,
+      "step": 4273
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6777539456984496,
+      "learning_rate": 8.535168758632669e-07,
+      "loss": 0.1231,
+      "step": 4274
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2755821239711442,
+      "learning_rate": 8.534438350923373e-07,
+      "loss": 0.1061,
+      "step": 4275
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5404488903288633,
+      "learning_rate": 8.533707792427831e-07,
+      "loss": 0.2662,
+      "step": 4276
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6068717340366714,
+      "learning_rate": 8.532977083177213e-07,
+      "loss": 0.2832,
+      "step": 4277
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6639554778520406,
+      "learning_rate": 8.532246223202688e-07,
+      "loss": 0.1044,
+      "step": 4278
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1836937541968538,
+      "learning_rate": 8.53151521253544e-07,
+      "loss": 0.1085,
+      "step": 4279
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4925808917133856,
+      "learning_rate": 8.530784051206653e-07,
+      "loss": 0.1829,
+      "step": 4280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7013712454806474,
+      "learning_rate": 8.530052739247521e-07,
+      "loss": 0.1747,
+      "step": 4281
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.36553437198870564,
+      "learning_rate": 8.529321276689244e-07,
+      "loss": 0.218,
+      "step": 4282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1997752663695453,
+      "learning_rate": 8.528589663563024e-07,
+      "loss": 0.224,
+      "step": 4283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7877892821714088,
+      "learning_rate": 8.527857899900077e-07,
+      "loss": 0.0509,
+      "step": 4284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5189093467424057,
+      "learning_rate": 8.527125985731621e-07,
+      "loss": 0.1515,
+      "step": 4285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.794490678121487,
+      "learning_rate": 8.526393921088877e-07,
+      "loss": 0.0195,
+      "step": 4286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9463137408634047,
+      "learning_rate": 8.525661706003082e-07,
+      "loss": 0.1803,
+      "step": 4287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.733742604782018,
+      "learning_rate": 8.524929340505471e-07,
+      "loss": 0.3151,
+      "step": 4288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.888426562772552,
+      "learning_rate": 8.524196824627288e-07,
+      "loss": 0.3374,
+      "step": 4289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5635952352373081,
+      "learning_rate": 8.523464158399783e-07,
+      "loss": 0.0777,
+      "step": 4290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9068369267471855,
+      "learning_rate": 8.522731341854215e-07,
+      "loss": 0.1167,
+      "step": 4291
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16553678920009054,
+      "learning_rate": 8.521998375021846e-07,
+      "loss": 0.0062,
+      "step": 4292
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.751025825975747,
+      "learning_rate": 8.521265257933947e-07,
+      "loss": 0.2014,
+      "step": 4293
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5086890057313618,
+      "learning_rate": 8.520531990621792e-07,
+      "loss": 0.0529,
+      "step": 4294
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5195701383131921,
+      "learning_rate": 8.519798573116666e-07,
+      "loss": 0.1368,
+      "step": 4295
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4886367835918936,
+      "learning_rate": 8.519065005449857e-07,
+      "loss": 0.3366,
+      "step": 4296
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.345860065984868,
+      "learning_rate": 8.51833128765266e-07,
+      "loss": 0.2093,
+      "step": 4297
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6163713218739876,
+      "learning_rate": 8.51759741975638e-07,
+      "loss": 0.1095,
+      "step": 4298
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2497052208533497,
+      "learning_rate": 8.51686340179232e-07,
+      "loss": 0.1417,
+      "step": 4299
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0159735011566458,
+      "learning_rate": 8.516129233791799e-07,
+      "loss": 0.1864,
+      "step": 4300
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.48030318363378866,
+      "learning_rate": 8.515394915786136e-07,
+      "loss": 0.0638,
+      "step": 4301
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7073627873066558,
+      "learning_rate": 8.51466044780666e-07,
+      "loss": 0.5101,
+      "step": 4302
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.42053252529749324,
+      "learning_rate": 8.513925829884703e-07,
+      "loss": 0.244,
+      "step": 4303
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7435353546687125,
+      "learning_rate": 8.513191062051607e-07,
+      "loss": 0.1459,
+      "step": 4304
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.47268336912859865,
+      "learning_rate": 8.512456144338716e-07,
+      "loss": 0.3455,
+      "step": 4305
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7420844477051433,
+      "learning_rate": 8.511721076777387e-07,
+      "loss": 0.169,
+      "step": 4306
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9693912136338695,
+      "learning_rate": 8.510985859398976e-07,
+      "loss": 0.4203,
+      "step": 4307
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2386222437024794,
+      "learning_rate": 8.510250492234852e-07,
+      "loss": 0.0616,
+      "step": 4308
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9317269982300498,
+      "learning_rate": 8.509514975316384e-07,
+      "loss": 0.2988,
+      "step": 4309
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3509798007559648,
+      "learning_rate": 8.508779308674952e-07,
+      "loss": 0.1942,
+      "step": 4310
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8425443795757862,
+      "learning_rate": 8.508043492341943e-07,
+      "loss": 0.1486,
+      "step": 4311
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3790680364923795,
+      "learning_rate": 8.507307526348744e-07,
+      "loss": 0.0867,
+      "step": 4312
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6909365406472099,
+      "learning_rate": 8.506571410726758e-07,
+      "loss": 0.1258,
+      "step": 4313
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.112657815616951,
+      "learning_rate": 8.505835145507385e-07,
+      "loss": 0.1601,
+      "step": 4314
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8192066693388006,
+      "learning_rate": 8.505098730722038e-07,
+      "loss": 0.0203,
+      "step": 4315
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.023798586627928,
+      "learning_rate": 8.504362166402131e-07,
+      "loss": 0.41,
+      "step": 4316
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4184967784167302,
+      "learning_rate": 8.503625452579091e-07,
+      "loss": 0.1804,
+      "step": 4317
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2020954289720547,
+      "learning_rate": 8.502888589284346e-07,
+      "loss": 0.1918,
+      "step": 4318
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9166679015975862,
+      "learning_rate": 8.502151576549332e-07,
+      "loss": 0.0869,
+      "step": 4319
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7368593006208842,
+      "learning_rate": 8.50141441440549e-07,
+      "loss": 0.0869,
+      "step": 4320
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9989025002647027,
+      "learning_rate": 8.500677102884273e-07,
+      "loss": 0.0395,
+      "step": 4321
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8430282057065523,
+      "learning_rate": 8.499939642017131e-07,
+      "loss": 0.3408,
+      "step": 4322
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4278914053106401,
+      "learning_rate": 8.49920203183553e-07,
+      "loss": 0.2442,
+      "step": 4323
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8225644744674708,
+      "learning_rate": 8.498464272370936e-07,
+      "loss": 0.3567,
+      "step": 4324
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8640436344809874,
+      "learning_rate": 8.497726363654823e-07,
+      "loss": 0.1426,
+      "step": 4325
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3736799735416585,
+      "learning_rate": 8.496988305718671e-07,
+      "loss": 0.2412,
+      "step": 4326
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7029669234188077,
+      "learning_rate": 8.496250098593968e-07,
+      "loss": 0.2283,
+      "step": 4327
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4224616227847302,
+      "learning_rate": 8.495511742312209e-07,
+      "loss": 0.439,
+      "step": 4328
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.585775683575167,
+      "learning_rate": 8.494773236904891e-07,
+      "loss": 0.2986,
+      "step": 4329
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6835802848282123,
+      "learning_rate": 8.494034582403522e-07,
+      "loss": 0.2607,
+      "step": 4330
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2957920462701642,
+      "learning_rate": 8.493295778839614e-07,
+      "loss": 0.0239,
+      "step": 4331
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5438956127282539,
+      "learning_rate": 8.492556826244686e-07,
+      "loss": 0.0947,
+      "step": 4332
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9298790936968806,
+      "learning_rate": 8.491817724650261e-07,
+      "loss": 0.2266,
+      "step": 4333
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6943324528089971,
+      "learning_rate": 8.491078474087873e-07,
+      "loss": 0.4088,
+      "step": 4334
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.32700302844867757,
+      "learning_rate": 8.49033907458906e-07,
+      "loss": 0.4245,
+      "step": 4335
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.598715524527703,
+      "learning_rate": 8.489599526185366e-07,
+      "loss": 0.2194,
+      "step": 4336
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7196597826352433,
+      "learning_rate": 8.488859828908341e-07,
+      "loss": 0.291,
+      "step": 4337
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.118949806060416,
+      "learning_rate": 8.488119982789541e-07,
+      "loss": 0.1923,
+      "step": 4338
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5843986075098508,
+      "learning_rate": 8.487379987860533e-07,
+      "loss": 0.0867,
+      "step": 4339
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9454531375694244,
+      "learning_rate": 8.486639844152881e-07,
+      "loss": 0.3082,
+      "step": 4340
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9987469523016238,
+      "learning_rate": 8.485899551698166e-07,
+      "loss": 0.3858,
+      "step": 4341
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40715029117676954,
+      "learning_rate": 8.485159110527969e-07,
+      "loss": 0.1751,
+      "step": 4342
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.499792871528072,
+      "learning_rate": 8.484418520673878e-07,
+      "loss": 0.0946,
+      "step": 4343
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.262308483993313,
+      "learning_rate": 8.483677782167489e-07,
+      "loss": 0.1008,
+      "step": 4344
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4171213140551655,
+      "learning_rate": 8.482936895040402e-07,
+      "loss": 0.1927,
+      "step": 4345
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5700755835615721,
+      "learning_rate": 8.482195859324225e-07,
+      "loss": 0.3416,
+      "step": 4346
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.525540193469076,
+      "learning_rate": 8.481454675050572e-07,
+      "loss": 0.0943,
+      "step": 4347
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7161781408228356,
+      "learning_rate": 8.480713342251065e-07,
+      "loss": 0.1144,
+      "step": 4348
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.48178956519723415,
+      "learning_rate": 8.47997186095733e-07,
+      "loss": 0.0822,
+      "step": 4349
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4131590219410887,
+      "learning_rate": 8.479230231201e-07,
+      "loss": 0.1703,
+      "step": 4350
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4610693527349393,
+      "learning_rate": 8.478488453013713e-07,
+      "loss": 0.0302,
+      "step": 4351
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.572453851246035,
+      "learning_rate": 8.477746526427117e-07,
+      "loss": 0.2856,
+      "step": 4352
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0447256574844934,
+      "learning_rate": 8.477004451472862e-07,
+      "loss": 0.1349,
+      "step": 4353
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1982465611298891,
+      "learning_rate": 8.476262228182608e-07,
+      "loss": 0.3739,
+      "step": 4354
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.524392966559799,
+      "learning_rate": 8.475519856588019e-07,
+      "loss": 0.1272,
+      "step": 4355
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40241971349987826,
+      "learning_rate": 8.474777336720766e-07,
+      "loss": 0.1078,
+      "step": 4356
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6715665943043577,
+      "learning_rate": 8.474034668612528e-07,
+      "loss": 0.3504,
+      "step": 4357
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0836109662832707,
+      "learning_rate": 8.473291852294986e-07,
+      "loss": 0.0272,
+      "step": 4358
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2938766956526506,
+      "learning_rate": 8.472548887799833e-07,
+      "loss": 0.0779,
+      "step": 4359
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.816943845780774,
+      "learning_rate": 8.471805775158762e-07,
+      "loss": 0.2372,
+      "step": 4360
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.75841659991107,
+      "learning_rate": 8.471062514403478e-07,
+      "loss": 0.1443,
+      "step": 4361
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.47206538795515435,
+      "learning_rate": 8.470319105565689e-07,
+      "loss": 0.2239,
+      "step": 4362
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.411996083151186,
+      "learning_rate": 8.469575548677111e-07,
+      "loss": 0.2346,
+      "step": 4363
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9391920864355068,
+      "learning_rate": 8.468831843769466e-07,
+      "loss": 0.0221,
+      "step": 4364
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.49146848990345166,
+      "learning_rate": 8.468087990874479e-07,
+      "loss": 0.175,
+      "step": 4365
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5375892507560971,
+      "learning_rate": 8.46734399002389e-07,
+      "loss": 0.0121,
+      "step": 4366
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0964068343453515,
+      "learning_rate": 8.466599841249435e-07,
+      "loss": 0.3299,
+      "step": 4367
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5268867637463387,
+      "learning_rate": 8.465855544582861e-07,
+      "loss": 0.0925,
+      "step": 4368
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5820936578181061,
+      "learning_rate": 8.465111100055922e-07,
+      "loss": 0.2211,
+      "step": 4369
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.45649807209567056,
+      "learning_rate": 8.46436650770038e-07,
+      "loss": 0.1642,
+      "step": 4370
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3095959954200143,
+      "learning_rate": 8.463621767547997e-07,
+      "loss": 0.0989,
+      "step": 4371
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.032850155819076,
+      "learning_rate": 8.462876879630547e-07,
+      "loss": 0.0369,
+      "step": 4372
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7607701314968326,
+      "learning_rate": 8.462131843979808e-07,
+      "loss": 0.1293,
+      "step": 4373
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.47250294306162127,
+      "learning_rate": 8.461386660627563e-07,
+      "loss": 0.0511,
+      "step": 4374
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8788969238131757,
+      "learning_rate": 8.460641329605607e-07,
+      "loss": 0.2171,
+      "step": 4375
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.72297044819535,
+      "learning_rate": 8.459895850945735e-07,
+      "loss": 0.2224,
+      "step": 4376
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4394715100207545,
+      "learning_rate": 8.45915022467975e-07,
+      "loss": 0.1797,
+      "step": 4377
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3353120173309568,
+      "learning_rate": 8.458404450839462e-07,
+      "loss": 0.1442,
+      "step": 4378
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6724250722787823,
+      "learning_rate": 8.457658529456689e-07,
+      "loss": 0.2112,
+      "step": 4379
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8829546336921816,
+      "learning_rate": 8.456912460563253e-07,
+      "loss": 0.1709,
+      "step": 4380
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6053766450016911,
+      "learning_rate": 8.456166244190981e-07,
+      "loss": 0.2125,
+      "step": 4381
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1657277703372586,
+      "learning_rate": 8.455419880371709e-07,
+      "loss": 0.2216,
+      "step": 4382
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5844690456974323,
+      "learning_rate": 8.45467336913728e-07,
+      "loss": 0.152,
+      "step": 4383
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6908481820040442,
+      "learning_rate": 8.453926710519539e-07,
+      "loss": 0.1361,
+      "step": 4384
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.45749695868996737,
+      "learning_rate": 8.453179904550343e-07,
+      "loss": 0.0473,
+      "step": 4385
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3220879391832774,
+      "learning_rate": 8.452432951261548e-07,
+      "loss": 0.2768,
+      "step": 4386
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7971791292876084,
+      "learning_rate": 8.451685850685025e-07,
+      "loss": 0.0975,
+      "step": 4387
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8164272301833383,
+      "learning_rate": 8.450938602852644e-07,
+      "loss": 0.1351,
+      "step": 4388
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.015208488226374,
+      "learning_rate": 8.450191207796285e-07,
+      "loss": 0.2406,
+      "step": 4389
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0417216634065258,
+      "learning_rate": 8.449443665547833e-07,
+      "loss": 0.2907,
+      "step": 4390
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.38297480315899257,
+      "learning_rate": 8.44869597613918e-07,
+      "loss": 0.2583,
+      "step": 4391
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8137261665206976,
+      "learning_rate": 8.447948139602225e-07,
+      "loss": 0.1434,
+      "step": 4392
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8724312544785039,
+      "learning_rate": 8.44720015596887e-07,
+      "loss": 0.118,
+      "step": 4393
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.307410897826849,
+      "learning_rate": 8.446452025271027e-07,
+      "loss": 0.1632,
+      "step": 4394
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17111584161678214,
+      "learning_rate": 8.445703747540613e-07,
+      "loss": 0.0935,
+      "step": 4395
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.286088804021365,
+      "learning_rate": 8.444955322809548e-07,
+      "loss": 0.212,
+      "step": 4396
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9218959587526094,
+      "learning_rate": 8.444206751109766e-07,
+      "loss": 0.2434,
+      "step": 4397
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.608305385898315,
+      "learning_rate": 8.443458032473201e-07,
+      "loss": 0.2542,
+      "step": 4398
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7082549088828657,
+      "learning_rate": 8.442709166931792e-07,
+      "loss": 0.0726,
+      "step": 4399
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6564970763572002,
+      "learning_rate": 8.44196015451749e-07,
+      "loss": 0.0889,
+      "step": 4400
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5232565841103223,
+      "learning_rate": 8.441210995262249e-07,
+      "loss": 0.1873,
+      "step": 4401
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6616604078127134,
+      "learning_rate": 8.440461689198029e-07,
+      "loss": 0.2932,
+      "step": 4402
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7353454228622197,
+      "learning_rate": 8.439712236356797e-07,
+      "loss": 0.224,
+      "step": 4403
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0530861035964485,
+      "learning_rate": 8.438962636770526e-07,
+      "loss": 0.2295,
+      "step": 4404
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2168156362266176,
+      "learning_rate": 8.438212890471197e-07,
+      "loss": 0.3593,
+      "step": 4405
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.48474040718209976,
+      "learning_rate": 8.437462997490793e-07,
+      "loss": 0.006,
+      "step": 4406
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7252387637225124,
+      "learning_rate": 8.436712957861308e-07,
+      "loss": 0.0512,
+      "step": 4407
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5362488869203919,
+      "learning_rate": 8.435962771614739e-07,
+      "loss": 0.1622,
+      "step": 4408
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7438715173511999,
+      "learning_rate": 8.435212438783091e-07,
+      "loss": 0.269,
+      "step": 4409
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.120037485777825,
+      "learning_rate": 8.434461959398376e-07,
+      "loss": 0.1691,
+      "step": 4410
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.47257631506788905,
+      "learning_rate": 8.433711333492608e-07,
+      "loss": 0.2809,
+      "step": 4411
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.824371668720856,
+      "learning_rate": 8.432960561097814e-07,
+      "loss": 0.3591,
+      "step": 4412
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6413136399100594,
+      "learning_rate": 8.43220964224602e-07,
+      "loss": 0.1596,
+      "step": 4413
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.30765493077013994,
+      "learning_rate": 8.431458576969263e-07,
+      "loss": 0.1539,
+      "step": 4414
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.817059073970843,
+      "learning_rate": 8.430707365299585e-07,
+      "loss": 0.3303,
+      "step": 4415
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.219533996366765,
+      "learning_rate": 8.429956007269035e-07,
+      "loss": 0.0526,
+      "step": 4416
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9711890001749441,
+      "learning_rate": 8.429204502909666e-07,
+      "loss": 0.2074,
+      "step": 4417
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5516315541274354,
+      "learning_rate": 8.428452852253538e-07,
+      "loss": 0.1186,
+      "step": 4418
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.382794531359862,
+      "learning_rate": 8.427701055332721e-07,
+      "loss": 0.1941,
+      "step": 4419
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3404852144539467,
+      "learning_rate": 8.426949112179287e-07,
+      "loss": 0.1036,
+      "step": 4420
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6916350369653512,
+      "learning_rate": 8.426197022825313e-07,
+      "loss": 0.1553,
+      "step": 4421
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8270012871780776,
+      "learning_rate": 8.425444787302887e-07,
+      "loss": 0.1212,
+      "step": 4422
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.246385309263991,
+      "learning_rate": 8.4246924056441e-07,
+      "loss": 0.2484,
+      "step": 4423
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.673906575969224,
+      "learning_rate": 8.423939877881053e-07,
+      "loss": 0.1861,
+      "step": 4424
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7879070177283408,
+      "learning_rate": 8.423187204045845e-07,
+      "loss": 0.339,
+      "step": 4425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.348886836199494,
+      "learning_rate": 8.422434384170591e-07,
+      "loss": 0.1119,
+      "step": 4426
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6745275690026145,
+      "learning_rate": 8.421681418287405e-07,
+      "loss": 0.3574,
+      "step": 4427
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.43383469870148433,
+      "learning_rate": 8.420928306428412e-07,
+      "loss": 0.123,
+      "step": 4428
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7763078183179507,
+      "learning_rate": 8.420175048625742e-07,
+      "loss": 0.1554,
+      "step": 4429
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7748784863518428,
+      "learning_rate": 8.419421644911526e-07,
+      "loss": 0.0832,
+      "step": 4430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6505815947911184,
+      "learning_rate": 8.418668095317911e-07,
+      "loss": 0.248,
+      "step": 4431
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5520398041901158,
+      "learning_rate": 8.417914399877042e-07,
+      "loss": 0.1497,
+      "step": 4432
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0835156592478004,
+      "learning_rate": 8.417160558621073e-07,
+      "loss": 0.3796,
+      "step": 4433
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5971967168656509,
+      "learning_rate": 8.416406571582167e-07,
+      "loss": 0.242,
+      "step": 4434
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.916643175209122,
+      "learning_rate": 8.415652438792486e-07,
+      "loss": 0.3095,
+      "step": 4435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.775772313594287,
+      "learning_rate": 8.414898160284207e-07,
+      "loss": 0.1477,
+      "step": 4436
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9152256704284947,
+      "learning_rate": 8.414143736089509e-07,
+      "loss": 0.0795,
+      "step": 4437
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0285415379650333,
+      "learning_rate": 8.413389166240574e-07,
+      "loss": 0.3897,
+      "step": 4438
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6570578741704823,
+      "learning_rate": 8.412634450769598e-07,
+      "loss": 0.181,
+      "step": 4439
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2676443027375748,
+      "learning_rate": 8.411879589708775e-07,
+      "loss": 0.1724,
+      "step": 4440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7798554241070158,
+      "learning_rate": 8.411124583090307e-07,
+      "loss": 0.0496,
+      "step": 4441
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5629140076740263,
+      "learning_rate": 8.410369430946412e-07,
+      "loss": 0.2308,
+      "step": 4442
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.780273770448044,
+      "learning_rate": 8.409614133309297e-07,
+      "loss": 0.2241,
+      "step": 4443
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5670293097966727,
+      "learning_rate": 8.408858690211191e-07,
+      "loss": 0.3358,
+      "step": 4444
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5042621290216677,
+      "learning_rate": 8.408103101684321e-07,
+      "loss": 0.142,
+      "step": 4445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2387795036670779,
+      "learning_rate": 8.40734736776092e-07,
+      "loss": 0.0405,
+      "step": 4446
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7124827909652498,
+      "learning_rate": 8.406591488473232e-07,
+      "loss": 0.2548,
+      "step": 4447
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5194044697111061,
+      "learning_rate": 8.405835463853502e-07,
+      "loss": 0.1652,
+      "step": 4448
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5701851869266625,
+      "learning_rate": 8.405079293933985e-07,
+      "loss": 0.0565,
+      "step": 4449
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0910496102773226,
+      "learning_rate": 8.404322978746941e-07,
+      "loss": 0.4179,
+      "step": 4450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6895464900298918,
+      "learning_rate": 8.403566518324634e-07,
+      "loss": 0.1947,
+      "step": 4451
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6471583129533248,
+      "learning_rate": 8.402809912699338e-07,
+      "loss": 0.3407,
+      "step": 4452
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0732438394333603,
+      "learning_rate": 8.402053161903331e-07,
+      "loss": 0.2065,
+      "step": 4453
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.47199993676472796,
+      "learning_rate": 8.401296265968896e-07,
+      "loss": 0.0839,
+      "step": 4454
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6270475259464904,
+      "learning_rate": 8.400539224928325e-07,
+      "loss": 0.2564,
+      "step": 4455
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.084180740135046,
+      "learning_rate": 8.399782038813916e-07,
+      "loss": 0.1497,
+      "step": 4456
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5701749049464911,
+      "learning_rate": 8.399024707657969e-07,
+      "loss": 0.1572,
+      "step": 4457
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6142287732929258,
+      "learning_rate": 8.398267231492797e-07,
+      "loss": 0.0791,
+      "step": 4458
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.42223800371297737,
+      "learning_rate": 8.397509610350712e-07,
+      "loss": 0.1058,
+      "step": 4459
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.18619950043951156,
+      "learning_rate": 8.396751844264038e-07,
+      "loss": 0.002,
+      "step": 4460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0950956996937822,
+      "learning_rate": 8.395993933265101e-07,
+      "loss": 0.295,
+      "step": 4461
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1258189986203382,
+      "learning_rate": 8.395235877386236e-07,
+      "loss": 0.1735,
+      "step": 4462
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6911988427258985,
+      "learning_rate": 8.394477676659784e-07,
+      "loss": 0.2315,
+      "step": 4463
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8369721166034685,
+      "learning_rate": 8.39371933111809e-07,
+      "loss": 0.4423,
+      "step": 4464
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2261142755252696,
+      "learning_rate": 8.392960840793508e-07,
+      "loss": 0.346,
+      "step": 4465
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6335384642094622,
+      "learning_rate": 8.392202205718395e-07,
+      "loss": 0.1489,
+      "step": 4466
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5759397326677361,
+      "learning_rate": 8.391443425925118e-07,
+      "loss": 0.1198,
+      "step": 4467
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8138314939314969,
+      "learning_rate": 8.390684501446046e-07,
+      "loss": 0.3187,
+      "step": 4468
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9922259606701328,
+      "learning_rate": 8.389925432313556e-07,
+      "loss": 0.2434,
+      "step": 4469
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9439010131970127,
+      "learning_rate": 8.389166218560033e-07,
+      "loss": 0.281,
+      "step": 4470
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.928976400579361,
+      "learning_rate": 8.388406860217867e-07,
+      "loss": 0.1049,
+      "step": 4471
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8344175767333075,
+      "learning_rate": 8.387647357319452e-07,
+      "loss": 0.1803,
+      "step": 4472
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3053728221696224,
+      "learning_rate": 8.38688770989719e-07,
+      "loss": 0.2961,
+      "step": 4473
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6094112650877064,
+      "learning_rate": 8.386127917983492e-07,
+      "loss": 0.2262,
+      "step": 4474
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.39069129879624453,
+      "learning_rate": 8.385367981610769e-07,
+      "loss": 0.0516,
+      "step": 4475
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.33031838581037737,
+      "learning_rate": 8.384607900811441e-07,
+      "loss": 0.1315,
+      "step": 4476
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.073025608080206,
+      "learning_rate": 8.383847675617938e-07,
+      "loss": 0.1077,
+      "step": 4477
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.47848041576443184,
+      "learning_rate": 8.383087306062689e-07,
+      "loss": 0.1378,
+      "step": 4478
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.092192500525181,
+      "learning_rate": 8.382326792178136e-07,
+      "loss": 0.1555,
+      "step": 4479
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4073851225514589,
+      "learning_rate": 8.381566133996722e-07,
+      "loss": 0.1103,
+      "step": 4480
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5885160317150906,
+      "learning_rate": 8.3808053315509e-07,
+      "loss": 0.1013,
+      "step": 4481
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4601262119205156,
+      "learning_rate": 8.380044384873126e-07,
+      "loss": 0.0265,
+      "step": 4482
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4609037119067469,
+      "learning_rate": 8.379283293995862e-07,
+      "loss": 0.1188,
+      "step": 4483
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7100195224733987,
+      "learning_rate": 8.378522058951582e-07,
+      "loss": 0.1741,
+      "step": 4484
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9272890709236896,
+      "learning_rate": 8.377760679772759e-07,
+      "loss": 0.2117,
+      "step": 4485
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.396616398343421,
+      "learning_rate": 8.376999156491873e-07,
+      "loss": 0.2864,
+      "step": 4486
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9904552251003063,
+      "learning_rate": 8.376237489141416e-07,
+      "loss": 0.3021,
+      "step": 4487
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.668455169482055,
+      "learning_rate": 8.37547567775388e-07,
+      "loss": 0.4558,
+      "step": 4488
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3726506649987562,
+      "learning_rate": 8.374713722361766e-07,
+      "loss": 0.3314,
+      "step": 4489
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.50682511573133,
+      "learning_rate": 8.373951622997581e-07,
+      "loss": 0.2216,
+      "step": 4490
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6726497042068822,
+      "learning_rate": 8.373189379693837e-07,
+      "loss": 0.2344,
+      "step": 4491
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.609016184030008,
+      "learning_rate": 8.372426992483054e-07,
+      "loss": 0.3915,
+      "step": 4492
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1723781811917382,
+      "learning_rate": 8.371664461397756e-07,
+      "loss": 0.3616,
+      "step": 4493
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5127086883005437,
+      "learning_rate": 8.370901786470474e-07,
+      "loss": 0.1553,
+      "step": 4494
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.67868702046723,
+      "learning_rate": 8.370138967733744e-07,
+      "loss": 0.1686,
+      "step": 4495
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0072948668837383,
+      "learning_rate": 8.369376005220113e-07,
+      "loss": 0.3462,
+      "step": 4496
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9595075385103153,
+      "learning_rate": 8.368612898962126e-07,
+      "loss": 0.1108,
+      "step": 4497
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.602483397325656,
+      "learning_rate": 8.367849648992342e-07,
+      "loss": 0.0887,
+      "step": 4498
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.374804428447395,
+      "learning_rate": 8.367086255343322e-07,
+      "loss": 0.4079,
+      "step": 4499
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.874141071120875,
+      "learning_rate": 8.366322718047634e-07,
+      "loss": 0.299,
+      "step": 4500
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2232745684819886,
+      "learning_rate": 8.36555903713785e-07,
+      "loss": 0.3907,
+      "step": 4501
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9824008189289976,
+      "learning_rate": 8.364795212646553e-07,
+      "loss": 0.2109,
+      "step": 4502
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2373988796750734,
+      "learning_rate": 8.364031244606329e-07,
+      "loss": 0.0894,
+      "step": 4503
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2971308108811155,
+      "learning_rate": 8.363267133049768e-07,
+      "loss": 0.145,
+      "step": 4504
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6923696073736749,
+      "learning_rate": 8.362502878009472e-07,
+      "loss": 0.1401,
+      "step": 4505
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8011960752370959,
+      "learning_rate": 8.361738479518041e-07,
+      "loss": 0.2743,
+      "step": 4506
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9141753106075413,
+      "learning_rate": 8.360973937608091e-07,
+      "loss": 0.2063,
+      "step": 4507
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9111172786342697,
+      "learning_rate": 8.360209252312236e-07,
+      "loss": 0.3673,
+      "step": 4508
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.21164414406498477,
+      "learning_rate": 8.359444423663099e-07,
+      "loss": 0.1103,
+      "step": 4509
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2468528092728297,
+      "learning_rate": 8.358679451693309e-07,
+      "loss": 0.1975,
+      "step": 4510
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3175666792175712,
+      "learning_rate": 8.357914336435503e-07,
+      "loss": 0.1048,
+      "step": 4511
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41181223836699016,
+      "learning_rate": 8.357149077922321e-07,
+      "loss": 0.003,
+      "step": 4512
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0454166567368108,
+      "learning_rate": 8.356383676186409e-07,
+      "loss": 0.3953,
+      "step": 4513
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6133492766439936,
+      "learning_rate": 8.355618131260424e-07,
+      "loss": 0.1993,
+      "step": 4514
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5328039269781779,
+      "learning_rate": 8.354852443177023e-07,
+      "loss": 0.0975,
+      "step": 4515
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5075796705113474,
+      "learning_rate": 8.354086611968871e-07,
+      "loss": 0.1081,
+      "step": 4516
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6732698907190383,
+      "learning_rate": 8.353320637668644e-07,
+      "loss": 0.1204,
+      "step": 4517
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8895655181978893,
+      "learning_rate": 8.352554520309017e-07,
+      "loss": 0.2238,
+      "step": 4518
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.36560431686336975,
+      "learning_rate": 8.351788259922676e-07,
+      "loss": 0.1056,
+      "step": 4519
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.565130468908776,
+      "learning_rate": 8.351021856542308e-07,
+      "loss": 0.2859,
+      "step": 4520
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7907019642042231,
+      "learning_rate": 8.350255310200611e-07,
+      "loss": 0.1872,
+      "step": 4521
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9747517445554856,
+      "learning_rate": 8.349488620930289e-07,
+      "loss": 0.1934,
+      "step": 4522
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9964218880977478,
+      "learning_rate": 8.348721788764048e-07,
+      "loss": 0.4043,
+      "step": 4523
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.487042549188682,
+      "learning_rate": 8.347954813734605e-07,
+      "loss": 0.1708,
+      "step": 4524
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8260511257866386,
+      "learning_rate": 8.347187695874678e-07,
+      "loss": 0.0331,
+      "step": 4525
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2344444427781944,
+      "learning_rate": 8.346420435216996e-07,
+      "loss": 0.1984,
+      "step": 4526
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.441167080806813,
+      "learning_rate": 8.34565303179429e-07,
+      "loss": 0.239,
+      "step": 4527
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.164332167445458,
+      "learning_rate": 8.344885485639302e-07,
+      "loss": 0.1012,
+      "step": 4528
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.774718838597696,
+      "learning_rate": 8.344117796784773e-07,
+      "loss": 0.1094,
+      "step": 4529
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5371450454458172,
+      "learning_rate": 8.343349965263457e-07,
+      "loss": 0.2501,
+      "step": 4530
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5861537448207714,
+      "learning_rate": 8.342581991108112e-07,
+      "loss": 0.1648,
+      "step": 4531
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7117029575482539,
+      "learning_rate": 8.341813874351499e-07,
+      "loss": 0.1814,
+      "step": 4532
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.49056028703580873,
+      "learning_rate": 8.341045615026388e-07,
+      "loss": 0.1694,
+      "step": 4533
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.48242432652565337,
+      "learning_rate": 8.340277213165554e-07,
+      "loss": 0.2004,
+      "step": 4534
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.922951045137883,
+      "learning_rate": 8.33950866880178e-07,
+      "loss": 0.1574,
+      "step": 4535
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3845890214990481,
+      "learning_rate": 8.338739981967853e-07,
+      "loss": 0.2496,
+      "step": 4536
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1293397170787918,
+      "learning_rate": 8.337971152696565e-07,
+      "loss": 0.2678,
+      "step": 4537
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9948576701873844,
+      "learning_rate": 8.33720218102072e-07,
+      "loss": 0.1849,
+      "step": 4538
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5148305889359297,
+      "learning_rate": 8.336433066973121e-07,
+      "loss": 0.1714,
+      "step": 4539
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1373115981841952,
+      "learning_rate": 8.33566381058658e-07,
+      "loss": 0.1738,
+      "step": 4540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6299221472633558,
+      "learning_rate": 8.334894411893913e-07,
+      "loss": 0.4132,
+      "step": 4541
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7694773051375428,
+      "learning_rate": 8.33412487092795e-07,
+      "loss": 0.189,
+      "step": 4542
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9193780317310255,
+      "learning_rate": 8.333355187721515e-07,
+      "loss": 0.5219,
+      "step": 4543
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3733278175440414,
+      "learning_rate": 8.332585362307447e-07,
+      "loss": 0.1146,
+      "step": 4544
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.39541841523059135,
+      "learning_rate": 8.331815394718589e-07,
+      "loss": 0.1744,
+      "step": 4545
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5466051194717116,
+      "learning_rate": 8.331045284987789e-07,
+      "loss": 0.1114,
+      "step": 4546
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5900345574222747,
+      "learning_rate": 8.3302750331479e-07,
+      "loss": 0.114,
+      "step": 4547
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9652657532302261,
+      "learning_rate": 8.329504639231783e-07,
+      "loss": 0.5105,
+      "step": 4548
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9475445901641059,
+      "learning_rate": 8.328734103272306e-07,
+      "loss": 0.3432,
+      "step": 4549
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6756970679755269,
+      "learning_rate": 8.32796342530234e-07,
+      "loss": 0.3576,
+      "step": 4550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7712110044567839,
+      "learning_rate": 8.327192605354765e-07,
+      "loss": 0.278,
+      "step": 4551
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9991316278556664,
+      "learning_rate": 8.326421643462465e-07,
+      "loss": 0.3683,
+      "step": 4552
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4824304657385116,
+      "learning_rate": 8.325650539658329e-07,
+      "loss": 0.2923,
+      "step": 4553
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6166678229650034,
+      "learning_rate": 8.324879293975257e-07,
+      "loss": 0.2142,
+      "step": 4554
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9273594079410605,
+      "learning_rate": 8.32410790644615e-07,
+      "loss": 0.1816,
+      "step": 4555
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2826245959745397,
+      "learning_rate": 8.323336377103917e-07,
+      "loss": 0.3449,
+      "step": 4556
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6230622406872659,
+      "learning_rate": 8.322564705981474e-07,
+      "loss": 0.6153,
+      "step": 4557
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3643263834950898,
+      "learning_rate": 8.321792893111741e-07,
+      "loss": 0.1063,
+      "step": 4558
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0041656352019737,
+      "learning_rate": 8.321020938527645e-07,
+      "loss": 0.3541,
+      "step": 4559
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.875425035071312,
+      "learning_rate": 8.320248842262121e-07,
+      "loss": 0.2437,
+      "step": 4560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.158182413474993,
+      "learning_rate": 8.319476604348106e-07,
+      "loss": 0.0853,
+      "step": 4561
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1630444341751074,
+      "learning_rate": 8.318704224818547e-07,
+      "loss": 0.2336,
+      "step": 4562
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.741685929124548,
+      "learning_rate": 8.317931703706393e-07,
+      "loss": 0.2526,
+      "step": 4563
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1489538592135694,
+      "learning_rate": 8.317159041044604e-07,
+      "loss": 0.0017,
+      "step": 4564
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5622834596398911,
+      "learning_rate": 8.316386236866142e-07,
+      "loss": 0.2137,
+      "step": 4565
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4250045625899355,
+      "learning_rate": 8.315613291203976e-07,
+      "loss": 0.1515,
+      "step": 4566
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9054182827221595,
+      "learning_rate": 8.314840204091082e-07,
+      "loss": 0.3344,
+      "step": 4567
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0175592120730144,
+      "learning_rate": 8.314066975560441e-07,
+      "loss": 0.3002,
+      "step": 4568
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7709846955208957,
+      "learning_rate": 8.31329360564504e-07,
+      "loss": 0.2873,
+      "step": 4569
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1544419692367198,
+      "learning_rate": 8.312520094377872e-07,
+      "loss": 0.0196,
+      "step": 4570
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.860307808512491,
+      "learning_rate": 8.311746441791941e-07,
+      "loss": 0.2551,
+      "step": 4571
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.36251735774144683,
+      "learning_rate": 8.310972647920247e-07,
+      "loss": 0.4366,
+      "step": 4572
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.193437641023994,
+      "learning_rate": 8.310198712795806e-07,
+      "loss": 0.1652,
+      "step": 4573
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4583817126495332,
+      "learning_rate": 8.309424636451632e-07,
+      "loss": 0.0906,
+      "step": 4574
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.57855699259455,
+      "learning_rate": 8.308650418920751e-07,
+      "loss": 0.3078,
+      "step": 4575
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6839847152348562,
+      "learning_rate": 8.307876060236191e-07,
+      "loss": 0.1316,
+      "step": 4576
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7829264233824589,
+      "learning_rate": 8.307101560430989e-07,
+      "loss": 0.2309,
+      "step": 4577
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2290653221772467,
+      "learning_rate": 8.306326919538186e-07,
+      "loss": 0.1449,
+      "step": 4578
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7162220578243333,
+      "learning_rate": 8.305552137590831e-07,
+      "loss": 0.206,
+      "step": 4579
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9666480360558779,
+      "learning_rate": 8.304777214621976e-07,
+      "loss": 0.1349,
+      "step": 4580
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1651099365529307,
+      "learning_rate": 8.304002150664682e-07,
+      "loss": 0.2618,
+      "step": 4581
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.15035581837467768,
+      "learning_rate": 8.303226945752014e-07,
+      "loss": 0.0857,
+      "step": 4582
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6239645041727994,
+      "learning_rate": 8.302451599917044e-07,
+      "loss": 0.3263,
+      "step": 4583
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3917517143758265,
+      "learning_rate": 8.301676113192852e-07,
+      "loss": 0.28,
+      "step": 4584
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3866381229855671,
+      "learning_rate": 8.300900485612518e-07,
+      "loss": 0.0685,
+      "step": 4585
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4654222159644312,
+      "learning_rate": 8.300124717209134e-07,
+      "loss": 0.2056,
+      "step": 4586
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6494826203489248,
+      "learning_rate": 8.299348808015795e-07,
+      "loss": 0.0846,
+      "step": 4587
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8289824743122928,
+      "learning_rate": 8.298572758065602e-07,
+      "loss": 0.1841,
+      "step": 4588
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.35527394307717897,
+      "learning_rate": 8.297796567391668e-07,
+      "loss": 0.1913,
+      "step": 4589
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6139715393711419,
+      "learning_rate": 8.297020236027101e-07,
+      "loss": 0.2243,
+      "step": 4590
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.037744256135214,
+      "learning_rate": 8.296243764005022e-07,
+      "loss": 0.0647,
+      "step": 4591
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.40674151977736805,
+      "learning_rate": 8.295467151358559e-07,
+      "loss": 0.098,
+      "step": 4592
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3260904276436847,
+      "learning_rate": 8.294690398120842e-07,
+      "loss": 0.1931,
+      "step": 4593
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3688959676407841,
+      "learning_rate": 8.29391350432501e-07,
+      "loss": 0.1907,
+      "step": 4594
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7784545250374711,
+      "learning_rate": 8.293136470004206e-07,
+      "loss": 0.2911,
+      "step": 4595
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7557252652780297,
+      "learning_rate": 8.29235929519158e-07,
+      "loss": 0.1921,
+      "step": 4596
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.703423784255771,
+      "learning_rate": 8.291581979920288e-07,
+      "loss": 0.3141,
+      "step": 4597
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5752349269239349,
+      "learning_rate": 8.290804524223491e-07,
+      "loss": 0.0607,
+      "step": 4598
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9029668416968109,
+      "learning_rate": 8.29002692813436e-07,
+      "loss": 0.1323,
+      "step": 4599
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2662017446158336,
+      "learning_rate": 8.289249191686063e-07,
+      "loss": 0.3032,
+      "step": 4600
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7251310077502929,
+      "learning_rate": 8.288471314911786e-07,
+      "loss": 0.1822,
+      "step": 4601
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.11923768026772609,
+      "learning_rate": 8.287693297844711e-07,
+      "loss": 0.0059,
+      "step": 4602
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6914129724847398,
+      "learning_rate": 8.286915140518032e-07,
+      "loss": 0.081,
+      "step": 4603
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5500244242234917,
+      "learning_rate": 8.286136842964944e-07,
+      "loss": 0.3442,
+      "step": 4604
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5948898502347938,
+      "learning_rate": 8.285358405218654e-07,
+      "loss": 0.204,
+      "step": 4605
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5806936398897938,
+      "learning_rate": 8.284579827312368e-07,
+      "loss": 0.1383,
+      "step": 4606
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2924043730778653,
+      "learning_rate": 8.283801109279305e-07,
+      "loss": 0.007,
+      "step": 4607
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.888021047655419,
+      "learning_rate": 8.283022251152685e-07,
+      "loss": 0.2459,
+      "step": 4608
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4635528743666432,
+      "learning_rate": 8.282243252965737e-07,
+      "loss": 0.2149,
+      "step": 4609
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7244105226734352,
+      "learning_rate": 8.281464114751692e-07,
+      "loss": 0.2014,
+      "step": 4610
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4907907933620943,
+      "learning_rate": 8.280684836543793e-07,
+      "loss": 0.1493,
+      "step": 4611
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6492772531144928,
+      "learning_rate": 8.279905418375283e-07,
+      "loss": 0.1192,
+      "step": 4612
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1096932116011753,
+      "learning_rate": 8.279125860279415e-07,
+      "loss": 0.4534,
+      "step": 4613
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5776216364701081,
+      "learning_rate": 8.278346162289445e-07,
+      "loss": 0.1065,
+      "step": 4614
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7320969443695299,
+      "learning_rate": 8.277566324438639e-07,
+      "loss": 0.3739,
+      "step": 4615
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5258340744849652,
+      "learning_rate": 8.276786346760262e-07,
+      "loss": 0.2819,
+      "step": 4616
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.405493453938572,
+      "learning_rate": 8.276006229287594e-07,
+      "loss": 0.0787,
+      "step": 4617
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9396627536486883,
+      "learning_rate": 8.275225972053917e-07,
+      "loss": 0.0345,
+      "step": 4618
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8877227374545172,
+      "learning_rate": 8.274445575092512e-07,
+      "loss": 0.0973,
+      "step": 4619
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.189231121656869,
+      "learning_rate": 8.273665038436679e-07,
+      "loss": 0.111,
+      "step": 4620
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.423798873505761,
+      "learning_rate": 8.272884362119712e-07,
+      "loss": 0.2148,
+      "step": 4621
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.554313924010514,
+      "learning_rate": 8.272103546174921e-07,
+      "loss": 0.0598,
+      "step": 4622
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.46607967107758846,
+      "learning_rate": 8.271322590635615e-07,
+      "loss": 0.1261,
+      "step": 4623
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0127076897066063,
+      "learning_rate": 8.27054149553511e-07,
+      "loss": 0.2251,
+      "step": 4624
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9804034283357304,
+      "learning_rate": 8.269760260906731e-07,
+      "loss": 0.1116,
+      "step": 4625
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7108680094980235,
+      "learning_rate": 8.268978886783805e-07,
+      "loss": 0.311,
+      "step": 4626
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.47837300243419517,
+      "learning_rate": 8.268197373199669e-07,
+      "loss": 0.0157,
+      "step": 4627
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8288377237677884,
+      "learning_rate": 8.267415720187662e-07,
+      "loss": 0.1729,
+      "step": 4628
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7902554788474534,
+      "learning_rate": 8.266633927781135e-07,
+      "loss": 0.3295,
+      "step": 4629
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.550425026591602,
+      "learning_rate": 8.265851996013436e-07,
+      "loss": 0.0477,
+      "step": 4630
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7396059757718584,
+      "learning_rate": 8.265069924917924e-07,
+      "loss": 0.0871,
+      "step": 4631
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6696288100632679,
+      "learning_rate": 8.264287714527969e-07,
+      "loss": 0.0493,
+      "step": 4632
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5719594532710218,
+      "learning_rate": 8.263505364876937e-07,
+      "loss": 0.1919,
+      "step": 4633
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7864193915426383,
+      "learning_rate": 8.262722875998204e-07,
+      "loss": 0.1095,
+      "step": 4634
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9594146892344215,
+      "learning_rate": 8.261940247925154e-07,
+      "loss": 0.1684,
+      "step": 4635
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6935617213124876,
+      "learning_rate": 8.261157480691178e-07,
+      "loss": 0.3149,
+      "step": 4636
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.38494366850966,
+      "learning_rate": 8.260374574329668e-07,
+      "loss": 0.1426,
+      "step": 4637
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7821993996292126,
+      "learning_rate": 8.259591528874022e-07,
+      "loss": 0.1792,
+      "step": 4638
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5951570960976268,
+      "learning_rate": 8.25880834435765e-07,
+      "loss": 0.1402,
+      "step": 4639
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3405780359458321,
+      "learning_rate": 8.258025020813963e-07,
+      "loss": 0.1172,
+      "step": 4640
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6995719513746204,
+      "learning_rate": 8.25724155827638e-07,
+      "loss": 0.0942,
+      "step": 4641
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2475965460830065,
+      "learning_rate": 8.256457956778324e-07,
+      "loss": 0.1576,
+      "step": 4642
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6485662326946411,
+      "learning_rate": 8.255674216353224e-07,
+      "loss": 0.0878,
+      "step": 4643
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9023323330171991,
+      "learning_rate": 8.254890337034519e-07,
+      "loss": 0.3953,
+      "step": 4644
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1302258505314244,
+      "learning_rate": 8.254106318855648e-07,
+      "loss": 0.0984,
+      "step": 4645
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3699190511740946,
+      "learning_rate": 8.25332216185006e-07,
+      "loss": 0.0037,
+      "step": 4646
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.336241396869838,
+      "learning_rate": 8.252537866051208e-07,
+      "loss": 0.3013,
+      "step": 4647
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0047472333264649,
+      "learning_rate": 8.251753431492553e-07,
+      "loss": 0.0824,
+      "step": 4648
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7740210959644076,
+      "learning_rate": 8.250968858207559e-07,
+      "loss": 0.1275,
+      "step": 4649
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6199947318242983,
+      "learning_rate": 8.2501841462297e-07,
+      "loss": 0.1217,
+      "step": 4650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8882994311764234,
+      "learning_rate": 8.24939929559245e-07,
+      "loss": 0.1321,
+      "step": 4651
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.889286149368149,
+      "learning_rate": 8.248614306329295e-07,
+      "loss": 0.3605,
+      "step": 4652
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.649976919422674,
+      "learning_rate": 8.247829178473722e-07,
+      "loss": 0.145,
+      "step": 4653
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9265333723529212,
+      "learning_rate": 8.247043912059228e-07,
+      "loss": 0.2944,
+      "step": 4654
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2498562724746331,
+      "learning_rate": 8.246258507119313e-07,
+      "loss": 0.0102,
+      "step": 4655
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.876708751259279,
+      "learning_rate": 8.245472963687484e-07,
+      "loss": 0.0302,
+      "step": 4656
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8298423753477335,
+      "learning_rate": 8.244687281797254e-07,
+      "loss": 0.4475,
+      "step": 4657
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6697123832860922,
+      "learning_rate": 8.243901461482143e-07,
+      "loss": 0.096,
+      "step": 4658
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.49985156868772807,
+      "learning_rate": 8.243115502775675e-07,
+      "loss": 0.1561,
+      "step": 4659
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8748840466630781,
+      "learning_rate": 8.242329405711379e-07,
+      "loss": 0.3036,
+      "step": 4660
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4727218219758068,
+      "learning_rate": 8.241543170322793e-07,
+      "loss": 0.1853,
+      "step": 4661
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6569204405804752,
+      "learning_rate": 8.240756796643459e-07,
+      "loss": 0.1419,
+      "step": 4662
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7974699904017554,
+      "learning_rate": 8.239970284706925e-07,
+      "loss": 0.1973,
+      "step": 4663
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2116589416155774,
+      "learning_rate": 8.239183634546746e-07,
+      "loss": 0.0917,
+      "step": 4664
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9486569978203159,
+      "learning_rate": 8.238396846196481e-07,
+      "loss": 0.2162,
+      "step": 4665
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.585708348381636,
+      "learning_rate": 8.237609919689696e-07,
+      "loss": 0.2014,
+      "step": 4666
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7825492050299566,
+      "learning_rate": 8.236822855059965e-07,
+      "loss": 0.1459,
+      "step": 4667
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8328841281850551,
+      "learning_rate": 8.236035652340864e-07,
+      "loss": 0.4653,
+      "step": 4668
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7004520417155415,
+      "learning_rate": 8.235248311565977e-07,
+      "loss": 0.0492,
+      "step": 4669
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5439465978243794,
+      "learning_rate": 8.234460832768893e-07,
+      "loss": 0.266,
+      "step": 4670
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.40523210034565077,
+      "learning_rate": 8.233673215983205e-07,
+      "loss": 0.2039,
+      "step": 4671
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8095650622236058,
+      "learning_rate": 8.23288546124252e-07,
+      "loss": 0.0136,
+      "step": 4672
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9191328261543144,
+      "learning_rate": 8.232097568580443e-07,
+      "loss": 0.0245,
+      "step": 4673
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0014710037084338,
+      "learning_rate": 8.231309538030585e-07,
+      "loss": 0.1874,
+      "step": 4674
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8227465043781447,
+      "learning_rate": 8.230521369626567e-07,
+      "loss": 0.1479,
+      "step": 4675
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8263908331466868,
+      "learning_rate": 8.229733063402012e-07,
+      "loss": 0.0944,
+      "step": 4676
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2417722896625074,
+      "learning_rate": 8.228944619390554e-07,
+      "loss": 0.1235,
+      "step": 4677
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5775048761976556,
+      "learning_rate": 8.228156037625826e-07,
+      "loss": 0.2096,
+      "step": 4678
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14880514998075245,
+      "learning_rate": 8.227367318141471e-07,
+      "loss": 0.0576,
+      "step": 4679
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8514689445420394,
+      "learning_rate": 8.226578460971141e-07,
+      "loss": 0.2681,
+      "step": 4680
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.353638450122576,
+      "learning_rate": 8.225789466148487e-07,
+      "loss": 0.0365,
+      "step": 4681
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3520576242029907,
+      "learning_rate": 8.225000333707169e-07,
+      "loss": 0.2149,
+      "step": 4682
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2241922239463783,
+      "learning_rate": 8.224211063680852e-07,
+      "loss": 0.0974,
+      "step": 4683
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0618219535173103,
+      "learning_rate": 8.223421656103212e-07,
+      "loss": 0.1163,
+      "step": 4684
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0534455178092248,
+      "learning_rate": 8.222632111007924e-07,
+      "loss": 0.1829,
+      "step": 4685
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5360778566064571,
+      "learning_rate": 8.221842428428672e-07,
+      "loss": 0.3163,
+      "step": 4686
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.407809610791214,
+      "learning_rate": 8.221052608399144e-07,
+      "loss": 0.0037,
+      "step": 4687
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5739234045881964,
+      "learning_rate": 8.220262650953037e-07,
+      "loss": 0.1646,
+      "step": 4688
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4669324054169145,
+      "learning_rate": 8.219472556124052e-07,
+      "loss": 0.4065,
+      "step": 4689
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5825268662027375,
+      "learning_rate": 8.218682323945895e-07,
+      "loss": 0.2036,
+      "step": 4690
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.4307318733168914,
+      "learning_rate": 8.217891954452281e-07,
+      "loss": 0.3523,
+      "step": 4691
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0414104009912146,
+      "learning_rate": 8.217101447676928e-07,
+      "loss": 0.4357,
+      "step": 4692
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.35544067754905573,
+      "learning_rate": 8.216310803653559e-07,
+      "loss": 0.2828,
+      "step": 4693
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.760208699033786,
+      "learning_rate": 8.215520022415905e-07,
+      "loss": 0.2042,
+      "step": 4694
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5296508099772066,
+      "learning_rate": 8.214729103997704e-07,
+      "loss": 0.2238,
+      "step": 4695
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5131090269687375,
+      "learning_rate": 8.213938048432696e-07,
+      "loss": 0.152,
+      "step": 4696
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.266830854764602,
+      "learning_rate": 8.213146855754632e-07,
+      "loss": 0.1553,
+      "step": 4697
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7225666498136933,
+      "learning_rate": 8.212355525997261e-07,
+      "loss": 0.2243,
+      "step": 4698
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1794197600511782,
+      "learning_rate": 8.211564059194347e-07,
+      "loss": 0.0153,
+      "step": 4699
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8186364889420014,
+      "learning_rate": 8.210772455379656e-07,
+      "loss": 0.1994,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8741105299284755,
+      "learning_rate": 8.209980714586955e-07,
+      "loss": 0.3851,
+      "step": 4701
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9295748485479789,
+      "learning_rate": 8.209188836850024e-07,
+      "loss": 0.145,
+      "step": 4702
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8603070684898152,
+      "learning_rate": 8.208396822202647e-07,
+      "loss": 0.1078,
+      "step": 4703
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1274263718944066,
+      "learning_rate": 8.207604670678612e-07,
+      "loss": 0.0527,
+      "step": 4704
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4563964387411023,
+      "learning_rate": 8.206812382311712e-07,
+      "loss": 0.2161,
+      "step": 4705
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7493729818410559,
+      "learning_rate": 8.20601995713575e-07,
+      "loss": 0.4807,
+      "step": 4706
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8077790117145862,
+      "learning_rate": 8.205227395184533e-07,
+      "loss": 0.0051,
+      "step": 4707
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3994268722811354,
+      "learning_rate": 8.204434696491871e-07,
+      "loss": 0.0114,
+      "step": 4708
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9107950272394667,
+      "learning_rate": 8.203641861091582e-07,
+      "loss": 0.0118,
+      "step": 4709
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7385816562045615,
+      "learning_rate": 8.202848889017493e-07,
+      "loss": 0.2511,
+      "step": 4710
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.14026937237341053,
+      "learning_rate": 8.202055780303431e-07,
+      "loss": 0.0637,
+      "step": 4711
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0653851814224091,
+      "learning_rate": 8.201262534983232e-07,
+      "loss": 0.156,
+      "step": 4712
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8707820745859838,
+      "learning_rate": 8.200469153090739e-07,
+      "loss": 0.202,
+      "step": 4713
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.559567352341435,
+      "learning_rate": 8.199675634659798e-07,
+      "loss": 0.2786,
+      "step": 4714
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8831184985543606,
+      "learning_rate": 8.198881979724262e-07,
+      "loss": 0.023,
+      "step": 4715
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7523378082412079,
+      "learning_rate": 8.19808818831799e-07,
+      "loss": 0.2363,
+      "step": 4716
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2402894662560693,
+      "learning_rate": 8.197294260474848e-07,
+      "loss": 0.2434,
+      "step": 4717
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9662177624727837,
+      "learning_rate": 8.196500196228704e-07,
+      "loss": 0.2969,
+      "step": 4718
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8447923730615003,
+      "learning_rate": 8.195705995613436e-07,
+      "loss": 0.0784,
+      "step": 4719
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1200294678811002,
+      "learning_rate": 8.194911658662927e-07,
+      "loss": 0.356,
+      "step": 4720
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4122942864613039,
+      "learning_rate": 8.194117185411062e-07,
+      "loss": 0.1679,
+      "step": 4721
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5700926088637928,
+      "learning_rate": 8.193322575891739e-07,
+      "loss": 0.3432,
+      "step": 4722
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5156157924268693,
+      "learning_rate": 8.192527830138856e-07,
+      "loss": 0.0841,
+      "step": 4723
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4370874859474639,
+      "learning_rate": 8.191732948186316e-07,
+      "loss": 0.239,
+      "step": 4724
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.093057648590071,
+      "learning_rate": 8.190937930068033e-07,
+      "loss": 0.2754,
+      "step": 4725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2555123969253632,
+      "learning_rate": 8.190142775817923e-07,
+      "loss": 0.3539,
+      "step": 4726
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.41472778104884717,
+      "learning_rate": 8.189347485469911e-07,
+      "loss": 0.1164,
+      "step": 4727
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8913508137970223,
+      "learning_rate": 8.188552059057923e-07,
+      "loss": 0.071,
+      "step": 4728
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.278132125183237,
+      "learning_rate": 8.187756496615895e-07,
+      "loss": 0.17,
+      "step": 4729
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8443359027948496,
+      "learning_rate": 8.186960798177765e-07,
+      "loss": 0.2811,
+      "step": 4730
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5137199003709494,
+      "learning_rate": 8.18616496377748e-07,
+      "loss": 0.0148,
+      "step": 4731
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.848923482668164,
+      "learning_rate": 8.185368993448993e-07,
+      "loss": 0.0536,
+      "step": 4732
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6488707373978129,
+      "learning_rate": 8.184572887226263e-07,
+      "loss": 0.0111,
+      "step": 4733
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27362005094595204,
+      "learning_rate": 8.183776645143252e-07,
+      "loss": 0.2206,
+      "step": 4734
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2135317880173803,
+      "learning_rate": 8.182980267233927e-07,
+      "loss": 0.016,
+      "step": 4735
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5562540739822275,
+      "learning_rate": 8.182183753532268e-07,
+      "loss": 0.3809,
+      "step": 4736
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4851006141667179,
+      "learning_rate": 8.18138710407225e-07,
+      "loss": 0.1858,
+      "step": 4737
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8018425586726883,
+      "learning_rate": 8.180590318887866e-07,
+      "loss": 0.1847,
+      "step": 4738
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3074670578761211,
+      "learning_rate": 8.179793398013102e-07,
+      "loss": 0.0349,
+      "step": 4739
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.998361311194355,
+      "learning_rate": 8.178996341481961e-07,
+      "loss": 0.1196,
+      "step": 4740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.668081158103692,
+      "learning_rate": 8.178199149328446e-07,
+      "loss": 0.2594,
+      "step": 4741
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5350107573420417,
+      "learning_rate": 8.177401821586567e-07,
+      "loss": 0.2036,
+      "step": 4742
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3575336661143542,
+      "learning_rate": 8.176604358290338e-07,
+      "loss": 0.1067,
+      "step": 4743
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.34226669859146513,
+      "learning_rate": 8.17580675947378e-07,
+      "loss": 0.1925,
+      "step": 4744
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5353285593550388,
+      "learning_rate": 8.175009025170922e-07,
+      "loss": 0.2514,
+      "step": 4745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4845438326924945,
+      "learning_rate": 8.174211155415798e-07,
+      "loss": 0.2897,
+      "step": 4746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6661578403739732,
+      "learning_rate": 8.173413150242444e-07,
+      "loss": 0.2717,
+      "step": 4747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.46315836179789466,
+      "learning_rate": 8.172615009684905e-07,
+      "loss": 0.2408,
+      "step": 4748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8786415967748011,
+      "learning_rate": 8.171816733777232e-07,
+      "loss": 0.1703,
+      "step": 4749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3677112592552782,
+      "learning_rate": 8.171018322553482e-07,
+      "loss": 0.1777,
+      "step": 4750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.531046619906957,
+      "learning_rate": 8.170219776047715e-07,
+      "loss": 0.3927,
+      "step": 4751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6553025686291939,
+      "learning_rate": 8.169421094294e-07,
+      "loss": 0.28,
+      "step": 4752
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7449111866321586,
+      "learning_rate": 8.168622277326409e-07,
+      "loss": 0.308,
+      "step": 4753
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.49963409292098165,
+      "learning_rate": 8.167823325179023e-07,
+      "loss": 0.1949,
+      "step": 4754
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4718151113525169,
+      "learning_rate": 8.167024237885927e-07,
+      "loss": 0.2263,
+      "step": 4755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.45236383778448064,
+      "learning_rate": 8.166225015481208e-07,
+      "loss": 0.267,
+      "step": 4756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6146614210267652,
+      "learning_rate": 8.165425657998966e-07,
+      "loss": 0.1569,
+      "step": 4757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6358796438298233,
+      "learning_rate": 8.164626165473302e-07,
+      "loss": 0.243,
+      "step": 4758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.720794840933295,
+      "learning_rate": 8.163826537938323e-07,
+      "loss": 0.3877,
+      "step": 4759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6339632423895353,
+      "learning_rate": 8.163026775428146e-07,
+      "loss": 0.1155,
+      "step": 4760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7540032251864254,
+      "learning_rate": 8.162226877976886e-07,
+      "loss": 0.1105,
+      "step": 4761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7409483817849162,
+      "learning_rate": 8.161426845618671e-07,
+      "loss": 0.1984,
+      "step": 4762
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8535988151051185,
+      "learning_rate": 8.160626678387632e-07,
+      "loss": 0.2088,
+      "step": 4763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.0317250470698331,
+      "learning_rate": 8.159826376317906e-07,
+      "loss": 0.2576,
+      "step": 4764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7235580007157351,
+      "learning_rate": 8.159025939443634e-07,
+      "loss": 0.2651,
+      "step": 4765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.42120452784390594,
+      "learning_rate": 8.158225367798966e-07,
+      "loss": 0.2216,
+      "step": 4766
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2914330147067489,
+      "learning_rate": 8.157424661418054e-07,
+      "loss": 0.0033,
+      "step": 4767
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3571703923846106,
+      "learning_rate": 8.156623820335058e-07,
+      "loss": 0.3403,
+      "step": 4768
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9458826621621397,
+      "learning_rate": 8.155822844584145e-07,
+      "loss": 0.2017,
+      "step": 4769
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8301991987674295,
+      "learning_rate": 8.155021734199486e-07,
+      "loss": 0.0587,
+      "step": 4770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.13018554076094138,
+      "learning_rate": 8.154220489215256e-07,
+      "loss": 0.0026,
+      "step": 4771
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4448594031418288,
+      "learning_rate": 8.153419109665641e-07,
+      "loss": 0.1517,
+      "step": 4772
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5194526639357139,
+      "learning_rate": 8.152617595584825e-07,
+      "loss": 0.1278,
+      "step": 4773
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1607749301962398,
+      "learning_rate": 8.151815947007007e-07,
+      "loss": 0.4137,
+      "step": 4774
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8357605391741285,
+      "learning_rate": 8.151014163966384e-07,
+      "loss": 0.0445,
+      "step": 4775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5136069215013145,
+      "learning_rate": 8.150212246497164e-07,
+      "loss": 0.099,
+      "step": 4776
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9389764063044246,
+      "learning_rate": 8.149410194633554e-07,
+      "loss": 0.0236,
+      "step": 4777
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.23778314579438078,
+      "learning_rate": 8.148608008409775e-07,
+      "loss": 0.0924,
+      "step": 4778
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7102426735803246,
+      "learning_rate": 8.14780568786005e-07,
+      "loss": 0.0872,
+      "step": 4779
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6122280657389112,
+      "learning_rate": 8.147003233018605e-07,
+      "loss": 0.1284,
+      "step": 4780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5522734828408397,
+      "learning_rate": 8.146200643919676e-07,
+      "loss": 0.162,
+      "step": 4781
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3747575208807832,
+      "learning_rate": 8.145397920597505e-07,
+      "loss": 0.0953,
+      "step": 4782
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7088931277898864,
+      "learning_rate": 8.144595063086335e-07,
+      "loss": 0.0691,
+      "step": 4783
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6193736346032409,
+      "learning_rate": 8.143792071420417e-07,
+      "loss": 0.0099,
+      "step": 4784
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0455965018983127,
+      "learning_rate": 8.142988945634009e-07,
+      "loss": 0.1422,
+      "step": 4785
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6153189687685371,
+      "learning_rate": 8.142185685761375e-07,
+      "loss": 0.1013,
+      "step": 4786
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5786634886782436,
+      "learning_rate": 8.141382291836783e-07,
+      "loss": 0.1876,
+      "step": 4787
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.33442578786895083,
+      "learning_rate": 8.140578763894508e-07,
+      "loss": 0.0659,
+      "step": 4788
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7147668461843721,
+      "learning_rate": 8.139775101968829e-07,
+      "loss": 0.164,
+      "step": 4789
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.151225999367275,
+      "learning_rate": 8.138971306094033e-07,
+      "loss": 0.3153,
+      "step": 4790
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0706416998260024,
+      "learning_rate": 8.13816737630441e-07,
+      "loss": 0.0871,
+      "step": 4791
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.584298758890585,
+      "learning_rate": 8.137363312634258e-07,
+      "loss": 0.1895,
+      "step": 4792
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6246688718089787,
+      "learning_rate": 8.136559115117881e-07,
+      "loss": 0.2332,
+      "step": 4793
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2684598818504136,
+      "learning_rate": 8.135754783789587e-07,
+      "loss": 0.0355,
+      "step": 4794
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5161168130043927,
+      "learning_rate": 8.134950318683691e-07,
+      "loss": 0.1363,
+      "step": 4795
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.645203944916089,
+      "learning_rate": 8.134145719834511e-07,
+      "loss": 0.1471,
+      "step": 4796
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7895951644051908,
+      "learning_rate": 8.133340987276375e-07,
+      "loss": 0.2596,
+      "step": 4797
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6105455471126823,
+      "learning_rate": 8.132536121043613e-07,
+      "loss": 0.1016,
+      "step": 4798
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7463146149380439,
+      "learning_rate": 8.131731121170563e-07,
+      "loss": 0.4144,
+      "step": 4799
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.703424109933384,
+      "learning_rate": 8.130925987691568e-07,
+      "loss": 0.215,
+      "step": 4800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1396604059930846,
+      "learning_rate": 8.130120720640976e-07,
+      "loss": 0.3308,
+      "step": 4801
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1795720050065013,
+      "learning_rate": 8.129315320053143e-07,
+      "loss": 0.1339,
+      "step": 4802
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8063374908340881,
+      "learning_rate": 8.128509785962427e-07,
+      "loss": 0.2179,
+      "step": 4803
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.38986296951702754,
+      "learning_rate": 8.127704118403194e-07,
+      "loss": 0.2449,
+      "step": 4804
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3440784561137122,
+      "learning_rate": 8.126898317409816e-07,
+      "loss": 0.1668,
+      "step": 4805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.27625082109998883,
+      "learning_rate": 8.12609238301667e-07,
+      "loss": 0.1278,
+      "step": 4806
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5367519695412856,
+      "learning_rate": 8.125286315258139e-07,
+      "loss": 0.1201,
+      "step": 4807
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3586161860975536,
+      "learning_rate": 8.124480114168611e-07,
+      "loss": 0.1583,
+      "step": 4808
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5083697358079993,
+      "learning_rate": 8.12367377978248e-07,
+      "loss": 0.1727,
+      "step": 4809
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.37507640697416395,
+      "learning_rate": 8.122867312134147e-07,
+      "loss": 0.2108,
+      "step": 4810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9980083401077641,
+      "learning_rate": 8.122060711258017e-07,
+      "loss": 0.2243,
+      "step": 4811
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8726441713134545,
+      "learning_rate": 8.121253977188499e-07,
+      "loss": 0.266,
+      "step": 4812
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8992933782380833,
+      "learning_rate": 8.120447109960014e-07,
+      "loss": 0.3619,
+      "step": 4813
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26986011613380834,
+      "learning_rate": 8.11964010960698e-07,
+      "loss": 0.0722,
+      "step": 4814
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8749899382429204,
+      "learning_rate": 8.11883297616383e-07,
+      "loss": 0.2008,
+      "step": 4815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.41960032165463595,
+      "learning_rate": 8.118025709664994e-07,
+      "loss": 0.1805,
+      "step": 4816
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5619040980184855,
+      "learning_rate": 8.117218310144913e-07,
+      "loss": 0.3494,
+      "step": 4817
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3700513371571906,
+      "learning_rate": 8.116410777638034e-07,
+      "loss": 0.06,
+      "step": 4818
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.669604798969257,
+      "learning_rate": 8.115603112178806e-07,
+      "loss": 0.2491,
+      "step": 4819
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.265929445107414,
+      "learning_rate": 8.114795313801686e-07,
+      "loss": 0.1283,
+      "step": 4820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.125882682580364,
+      "learning_rate": 8.113987382541136e-07,
+      "loss": 0.0843,
+      "step": 4821
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5303452032060689,
+      "learning_rate": 8.113179318431624e-07,
+      "loss": 0.3097,
+      "step": 4822
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6740244575310982,
+      "learning_rate": 8.112371121507625e-07,
+      "loss": 0.1511,
+      "step": 4823
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8220889534531641,
+      "learning_rate": 8.111562791803617e-07,
+      "loss": 0.132,
+      "step": 4824
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.358675217109917,
+      "learning_rate": 8.110754329354086e-07,
+      "loss": 0.3406,
+      "step": 4825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.891636297348018,
+      "learning_rate": 8.10994573419352e-07,
+      "loss": 0.2714,
+      "step": 4826
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2897380692260539,
+      "learning_rate": 8.109137006356419e-07,
+      "loss": 0.1168,
+      "step": 4827
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.487431741184058,
+      "learning_rate": 8.108328145877282e-07,
+      "loss": 0.3414,
+      "step": 4828
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.661817392387723,
+      "learning_rate": 8.107519152790619e-07,
+      "loss": 0.3947,
+      "step": 4829
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0208401780032037,
+      "learning_rate": 8.10671002713094e-07,
+      "loss": 0.2575,
+      "step": 4830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2116693190295287,
+      "learning_rate": 8.105900768932767e-07,
+      "loss": 0.3512,
+      "step": 4831
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4987501829098024,
+      "learning_rate": 8.105091378230624e-07,
+      "loss": 0.0436,
+      "step": 4832
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8342103669152526,
+      "learning_rate": 8.10428185505904e-07,
+      "loss": 0.5072,
+      "step": 4833
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8757327035704235,
+      "learning_rate": 8.103472199452553e-07,
+      "loss": 0.4112,
+      "step": 4834
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5415438431193039,
+      "learning_rate": 8.102662411445702e-07,
+      "loss": 0.2089,
+      "step": 4835
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4508114582543965,
+      "learning_rate": 8.101852491073036e-07,
+      "loss": 0.262,
+      "step": 4836
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5815245402928737,
+      "learning_rate": 8.101042438369108e-07,
+      "loss": 0.1776,
+      "step": 4837
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8965872238094595,
+      "learning_rate": 8.100232253368474e-07,
+      "loss": 0.33,
+      "step": 4838
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.35410971342035186,
+      "learning_rate": 8.099421936105702e-07,
+      "loss": 0.0168,
+      "step": 4839
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2150666514128075,
+      "learning_rate": 8.098611486615357e-07,
+      "loss": 0.0392,
+      "step": 4840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.033711989512499,
+      "learning_rate": 8.097800904932018e-07,
+      "loss": 0.2871,
+      "step": 4841
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0445909649421055,
+      "learning_rate": 8.096990191090265e-07,
+      "loss": 0.211,
+      "step": 4842
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.588543664066317,
+      "learning_rate": 8.096179345124685e-07,
+      "loss": 0.1082,
+      "step": 4843
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.914117310489967,
+      "learning_rate": 8.09536836706987e-07,
+      "loss": 0.0876,
+      "step": 4844
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.36171155517429354,
+      "learning_rate": 8.094557256960419e-07,
+      "loss": 0.1091,
+      "step": 4845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9676226298196176,
+      "learning_rate": 8.093746014830933e-07,
+      "loss": 0.0858,
+      "step": 4846
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6714200761593854,
+      "learning_rate": 8.092934640716023e-07,
+      "loss": 0.2552,
+      "step": 4847
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7545950416077731,
+      "learning_rate": 8.092123134650304e-07,
+      "loss": 0.3534,
+      "step": 4848
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4545882543203872,
+      "learning_rate": 8.091311496668396e-07,
+      "loss": 0.0389,
+      "step": 4849
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4086507364432175,
+      "learning_rate": 8.090499726804924e-07,
+      "loss": 0.3278,
+      "step": 4850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6553200148490134,
+      "learning_rate": 8.089687825094524e-07,
+      "loss": 0.1967,
+      "step": 4851
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9493328856123927,
+      "learning_rate": 8.088875791571829e-07,
+      "loss": 0.0726,
+      "step": 4852
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7316791064460415,
+      "learning_rate": 8.088063626271482e-07,
+      "loss": 0.3481,
+      "step": 4853
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8532338216162878,
+      "learning_rate": 8.087251329228135e-07,
+      "loss": 0.4165,
+      "step": 4854
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.182696331300221,
+      "learning_rate": 8.086438900476437e-07,
+      "loss": 0.0125,
+      "step": 4855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2246967929530859,
+      "learning_rate": 8.085626340051054e-07,
+      "loss": 0.0899,
+      "step": 4856
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5421316152513138,
+      "learning_rate": 8.084813647986648e-07,
+      "loss": 0.1918,
+      "step": 4857
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5652342853883778,
+      "learning_rate": 8.08400082431789e-07,
+      "loss": 0.1306,
+      "step": 4858
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.585916229816176,
+      "learning_rate": 8.083187869079458e-07,
+      "loss": 0.1555,
+      "step": 4859
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4553496465574186,
+      "learning_rate": 8.082374782306032e-07,
+      "loss": 0.1156,
+      "step": 4860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2827301344247135,
+      "learning_rate": 8.081561564032302e-07,
+      "loss": 0.2535,
+      "step": 4861
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4708930856967913,
+      "learning_rate": 8.080748214292961e-07,
+      "loss": 0.1788,
+      "step": 4862
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.895992974866067,
+      "learning_rate": 8.079934733122707e-07,
+      "loss": 0.2384,
+      "step": 4863
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9405193118592018,
+      "learning_rate": 8.079121120556247e-07,
+      "loss": 0.0869,
+      "step": 4864
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7232272466958303,
+      "learning_rate": 8.07830737662829e-07,
+      "loss": 0.1209,
+      "step": 4865
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.066643556053722,
+      "learning_rate": 8.077493501373554e-07,
+      "loss": 0.1455,
+      "step": 4866
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.47089877034493316,
+      "learning_rate": 8.076679494826757e-07,
+      "loss": 0.141,
+      "step": 4867
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.121471486827043,
+      "learning_rate": 8.075865357022628e-07,
+      "loss": 0.046,
+      "step": 4868
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.158549015956193,
+      "learning_rate": 8.075051087995899e-07,
+      "loss": 0.1862,
+      "step": 4869
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9730069506178508,
+      "learning_rate": 8.074236687781309e-07,
+      "loss": 0.4972,
+      "step": 4870
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0793488888012408,
+      "learning_rate": 8.073422156413603e-07,
+      "loss": 0.373,
+      "step": 4871
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1993090734989529,
+      "learning_rate": 8.072607493927528e-07,
+      "loss": 0.0676,
+      "step": 4872
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5660302550557087,
+      "learning_rate": 8.071792700357842e-07,
+      "loss": 0.2421,
+      "step": 4873
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7597048493635756,
+      "learning_rate": 8.070977775739304e-07,
+      "loss": 0.2999,
+      "step": 4874
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.38808110568411536,
+      "learning_rate": 8.070162720106679e-07,
+      "loss": 0.2452,
+      "step": 4875
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8209025724927607,
+      "learning_rate": 8.069347533494744e-07,
+      "loss": 0.1285,
+      "step": 4876
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8209839213021729,
+      "learning_rate": 8.068532215938269e-07,
+      "loss": 0.2401,
+      "step": 4877
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9038983135784068,
+      "learning_rate": 8.067716767472044e-07,
+      "loss": 0.1235,
+      "step": 4878
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7128631175040728,
+      "learning_rate": 8.066901188130854e-07,
+      "loss": 0.3511,
+      "step": 4879
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7170598170129175,
+      "learning_rate": 8.066085477949494e-07,
+      "loss": 0.0227,
+      "step": 4880
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.62173391419443,
+      "learning_rate": 8.065269636962763e-07,
+      "loss": 0.4253,
+      "step": 4881
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.947325453129527,
+      "learning_rate": 8.06445366520547e-07,
+      "loss": 0.1273,
+      "step": 4882
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0286287147679303,
+      "learning_rate": 8.063637562712421e-07,
+      "loss": 0.1125,
+      "step": 4883
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.187943991890367,
+      "learning_rate": 8.062821329518435e-07,
+      "loss": 0.1838,
+      "step": 4884
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2909373566917518,
+      "learning_rate": 8.062004965658336e-07,
+      "loss": 0.0796,
+      "step": 4885
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.11145047618391998,
+      "learning_rate": 8.061188471166947e-07,
+      "loss": 0.0973,
+      "step": 4886
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5975929567822871,
+      "learning_rate": 8.060371846079106e-07,
+      "loss": 0.2333,
+      "step": 4887
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8888907966553493,
+      "learning_rate": 8.059555090429649e-07,
+      "loss": 0.3266,
+      "step": 4888
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6178344911289139,
+      "learning_rate": 8.058738204253421e-07,
+      "loss": 0.4046,
+      "step": 4889
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3801275160196431,
+      "learning_rate": 8.057921187585273e-07,
+      "loss": 0.2519,
+      "step": 4890
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8939016166809852,
+      "learning_rate": 8.057104040460061e-07,
+      "loss": 0.2632,
+      "step": 4891
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6356544303041343,
+      "learning_rate": 8.056286762912643e-07,
+      "loss": 0.1726,
+      "step": 4892
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3755835281239818,
+      "learning_rate": 8.055469354977889e-07,
+      "loss": 0.1538,
+      "step": 4893
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.45209701749726516,
+      "learning_rate": 8.054651816690669e-07,
+      "loss": 0.0788,
+      "step": 4894
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.584728209799264,
+      "learning_rate": 8.053834148085864e-07,
+      "loss": 0.1629,
+      "step": 4895
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6760059551912412,
+      "learning_rate": 8.053016349198354e-07,
+      "loss": 0.1113,
+      "step": 4896
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.36028504095486713,
+      "learning_rate": 8.052198420063029e-07,
+      "loss": 0.0753,
+      "step": 4897
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3252513191684513,
+      "learning_rate": 8.051380360714783e-07,
+      "loss": 0.1073,
+      "step": 4898
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9650826262026042,
+      "learning_rate": 8.050562171188519e-07,
+      "loss": 0.3438,
+      "step": 4899
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6943397455835096,
+      "learning_rate": 8.049743851519139e-07,
+      "loss": 0.0816,
+      "step": 4900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6936148521736064,
+      "learning_rate": 8.048925401741555e-07,
+      "loss": 0.2057,
+      "step": 4901
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5132755435461887,
+      "learning_rate": 8.048106821890686e-07,
+      "loss": 0.1608,
+      "step": 4902
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6639518584758515,
+      "learning_rate": 8.04728811200145e-07,
+      "loss": 0.37,
+      "step": 4903
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7177421838414897,
+      "learning_rate": 8.046469272108779e-07,
+      "loss": 0.2359,
+      "step": 4904
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5432046176468095,
+      "learning_rate": 8.045650302247604e-07,
+      "loss": 0.2579,
+      "step": 4905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2688296023595715,
+      "learning_rate": 8.044831202452864e-07,
+      "loss": 0.111,
+      "step": 4906
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1354848906039784,
+      "learning_rate": 8.044011972759507e-07,
+      "loss": 0.0047,
+      "step": 4907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3150780133923562,
+      "learning_rate": 8.043192613202479e-07,
+      "loss": 0.0636,
+      "step": 4908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.43616701414274484,
+      "learning_rate": 8.042373123816735e-07,
+      "loss": 0.1428,
+      "step": 4909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4719856867245222,
+      "learning_rate": 8.041553504637237e-07,
+      "loss": 0.1362,
+      "step": 4910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.52111897260746,
+      "learning_rate": 8.040733755698955e-07,
+      "loss": 0.1843,
+      "step": 4911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8024011424709545,
+      "learning_rate": 8.039913877036855e-07,
+      "loss": 0.3262,
+      "step": 4912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9411942662342757,
+      "learning_rate": 8.03909386868592e-07,
+      "loss": 0.1619,
+      "step": 4913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.379404535635887,
+      "learning_rate": 8.038273730681131e-07,
+      "loss": 0.1127,
+      "step": 4914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3482281884152491,
+      "learning_rate": 8.037453463057476e-07,
+      "loss": 0.2595,
+      "step": 4915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5181406813260363,
+      "learning_rate": 8.036633065849952e-07,
+      "loss": 0.1973,
+      "step": 4916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9929727899045164,
+      "learning_rate": 8.035812539093556e-07,
+      "loss": 0.2721,
+      "step": 4917
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.41468132224453363,
+      "learning_rate": 8.034991882823295e-07,
+      "loss": 0.1691,
+      "step": 4918
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7564116362954902,
+      "learning_rate": 8.034171097074178e-07,
+      "loss": 0.2941,
+      "step": 4919
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0419087637830673,
+      "learning_rate": 8.033350181881223e-07,
+      "loss": 0.3469,
+      "step": 4920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9600973468805437,
+      "learning_rate": 8.032529137279452e-07,
+      "loss": 0.3099,
+      "step": 4921
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8859863115949071,
+      "learning_rate": 8.03170796330389e-07,
+      "loss": 0.2808,
+      "step": 4922
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3729753295058565,
+      "learning_rate": 8.030886659989575e-07,
+      "loss": 0.1751,
+      "step": 4923
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7550129699030261,
+      "learning_rate": 8.03006522737154e-07,
+      "loss": 0.294,
+      "step": 4924
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1835190904054413,
+      "learning_rate": 8.029243665484832e-07,
+      "loss": 0.2282,
+      "step": 4925
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4322986351748098,
+      "learning_rate": 8.028421974364499e-07,
+      "loss": 0.3293,
+      "step": 4926
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6129394757652573,
+      "learning_rate": 8.027600154045597e-07,
+      "loss": 0.2775,
+      "step": 4927
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6999076153066456,
+      "learning_rate": 8.026778204563186e-07,
+      "loss": 0.3713,
+      "step": 4928
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.528830069210588,
+      "learning_rate": 8.025956125952333e-07,
+      "loss": 0.1196,
+      "step": 4929
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.061599656854315,
+      "learning_rate": 8.025133918248108e-07,
+      "loss": 0.2902,
+      "step": 4930
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5694961102214035,
+      "learning_rate": 8.024311581485588e-07,
+      "loss": 0.0822,
+      "step": 4931
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5561115290492764,
+      "learning_rate": 8.023489115699857e-07,
+      "loss": 0.1893,
+      "step": 4932
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6327645403669329,
+      "learning_rate": 8.022666520926003e-07,
+      "loss": 0.3066,
+      "step": 4933
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.589508861362452,
+      "learning_rate": 8.021843797199119e-07,
+      "loss": 0.0339,
+      "step": 4934
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8778632804172244,
+      "learning_rate": 8.021020944554304e-07,
+      "loss": 0.0416,
+      "step": 4935
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7568867874491414,
+      "learning_rate": 8.020197963026662e-07,
+      "loss": 0.333,
+      "step": 4936
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5578812210514541,
+      "learning_rate": 8.019374852651302e-07,
+      "loss": 0.1473,
+      "step": 4937
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.30723336690661857,
+      "learning_rate": 8.018551613463344e-07,
+      "loss": 0.2139,
+      "step": 4938
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2492561977176489,
+      "learning_rate": 8.017728245497903e-07,
+      "loss": 0.1933,
+      "step": 4939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6182145102174467,
+      "learning_rate": 8.016904748790112e-07,
+      "loss": 0.1276,
+      "step": 4940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.31288799637361114,
+      "learning_rate": 8.016081123375097e-07,
+      "loss": 0.0768,
+      "step": 4941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.578768112587766,
+      "learning_rate": 8.015257369287999e-07,
+      "loss": 0.2145,
+      "step": 4942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6243010586183911,
+      "learning_rate": 8.014433486563961e-07,
+      "loss": 0.2143,
+      "step": 4943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5772522433540507,
+      "learning_rate": 8.01360947523813e-07,
+      "loss": 0.1336,
+      "step": 4944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.154009492996312,
+      "learning_rate": 8.01278533534566e-07,
+      "loss": 0.1029,
+      "step": 4945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5637675702015749,
+      "learning_rate": 8.011961066921712e-07,
+      "loss": 0.3088,
+      "step": 4946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8447865031556825,
+      "learning_rate": 8.01113667000145e-07,
+      "loss": 0.2675,
+      "step": 4947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8427574294223372,
+      "learning_rate": 8.010312144620045e-07,
+      "loss": 0.0268,
+      "step": 4948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6887712855980462,
+      "learning_rate": 8.009487490812671e-07,
+      "loss": 0.1334,
+      "step": 4949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.150689543029537,
+      "learning_rate": 8.008662708614513e-07,
+      "loss": 0.11,
+      "step": 4950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.19335587316152167,
+      "learning_rate": 8.007837798060754e-07,
+      "loss": 0.07,
+      "step": 4951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.580050594643492,
+      "learning_rate": 8.007012759186589e-07,
+      "loss": 0.2524,
+      "step": 4952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3589227727435151,
+      "learning_rate": 8.006187592027213e-07,
+      "loss": 0.1275,
+      "step": 4953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0979953933170263,
+      "learning_rate": 8.005362296617833e-07,
+      "loss": 0.4041,
+      "step": 4954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8801438577089142,
+      "learning_rate": 8.004536872993655e-07,
+      "loss": 0.0907,
+      "step": 4955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6296143796044339,
+      "learning_rate": 8.003711321189895e-07,
+      "loss": 0.0866,
+      "step": 4956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7793038701861162,
+      "learning_rate": 8.00288564124177e-07,
+      "loss": 0.2095,
+      "step": 4957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.37715338537872434,
+      "learning_rate": 8.002059833184509e-07,
+      "loss": 0.0245,
+      "step": 4958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7644760400658014,
+      "learning_rate": 8.001233897053339e-07,
+      "loss": 0.3113,
+      "step": 4959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6204781145389356,
+      "learning_rate": 8.000407832883498e-07,
+      "loss": 0.2237,
+      "step": 4960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6320265579175591,
+      "learning_rate": 7.999581640710229e-07,
+      "loss": 0.1138,
+      "step": 4961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8650039911746136,
+      "learning_rate": 7.998755320568777e-07,
+      "loss": 0.0433,
+      "step": 4962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4204365329912512,
+      "learning_rate": 7.997928872494393e-07,
+      "loss": 0.0049,
+      "step": 4963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5922103444989494,
+      "learning_rate": 7.997102296522338e-07,
+      "loss": 0.3073,
+      "step": 4964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4635430753988371,
+      "learning_rate": 7.996275592687873e-07,
+      "loss": 0.1315,
+      "step": 4965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4221298468018966,
+      "learning_rate": 7.995448761026269e-07,
+      "loss": 0.0955,
+      "step": 4966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7395847734958292,
+      "learning_rate": 7.994621801572799e-07,
+      "loss": 0.0599,
+      "step": 4967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.738676426321658,
+      "learning_rate": 7.993794714362743e-07,
+      "loss": 0.2013,
+      "step": 4968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.726919285608865,
+      "learning_rate": 7.992967499431386e-07,
+      "loss": 0.3061,
+      "step": 4969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6451475246976672,
+      "learning_rate": 7.992140156814018e-07,
+      "loss": 0.1609,
+      "step": 4970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7994856861678853,
+      "learning_rate": 7.991312686545937e-07,
+      "loss": 0.2259,
+      "step": 4971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5418123855404976,
+      "learning_rate": 7.990485088662444e-07,
+      "loss": 0.0922,
+      "step": 4972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2629885242231957,
+      "learning_rate": 7.989657363198844e-07,
+      "loss": 0.3647,
+      "step": 4973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7995087081555955,
+      "learning_rate": 7.988829510190451e-07,
+      "loss": 0.3445,
+      "step": 4974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6500852431645989,
+      "learning_rate": 7.988001529672586e-07,
+      "loss": 0.0953,
+      "step": 4975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4354991781289608,
+      "learning_rate": 7.987173421680566e-07,
+      "loss": 0.3204,
+      "step": 4976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.45323271759854744,
+      "learning_rate": 7.986345186249724e-07,
+      "loss": 0.1694,
+      "step": 4977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6257471185124001,
+      "learning_rate": 7.985516823415393e-07,
+      "loss": 0.2966,
+      "step": 4978
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9748700051738195,
+      "learning_rate": 7.984688333212911e-07,
+      "loss": 0.3093,
+      "step": 4979
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.766471716127174,
+      "learning_rate": 7.983859715677626e-07,
+      "loss": 0.0133,
+      "step": 4980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.780145636884804,
+      "learning_rate": 7.983030970844886e-07,
+      "loss": 0.1167,
+      "step": 4981
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.514590505070607,
+      "learning_rate": 7.98220209875005e-07,
+      "loss": 0.2056,
+      "step": 4982
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7500044643537416,
+      "learning_rate": 7.981373099428477e-07,
+      "loss": 0.1141,
+      "step": 4983
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9443011383275393,
+      "learning_rate": 7.980543972915534e-07,
+      "loss": 0.0075,
+      "step": 4984
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3823654877624576,
+      "learning_rate": 7.979714719246594e-07,
+      "loss": 0.1604,
+      "step": 4985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5027893630783101,
+      "learning_rate": 7.978885338457033e-07,
+      "loss": 0.1724,
+      "step": 4986
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.490008379196501,
+      "learning_rate": 7.978055830582235e-07,
+      "loss": 0.1139,
+      "step": 4987
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.825093526811515,
+      "learning_rate": 7.97722619565759e-07,
+      "loss": 0.1198,
+      "step": 4988
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.312071666363919,
+      "learning_rate": 7.976396433718491e-07,
+      "loss": 0.1773,
+      "step": 4989
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.620296038300229,
+      "learning_rate": 7.975566544800336e-07,
+      "loss": 0.3288,
+      "step": 4990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5364438222770149,
+      "learning_rate": 7.97473652893853e-07,
+      "loss": 0.229,
+      "step": 4991
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6438409456228207,
+      "learning_rate": 7.973906386168484e-07,
+      "loss": 0.3674,
+      "step": 4992
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1231179404140257,
+      "learning_rate": 7.973076116525613e-07,
+      "loss": 0.2368,
+      "step": 4993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.486323983764233,
+      "learning_rate": 7.972245720045339e-07,
+      "loss": 0.0239,
+      "step": 4994
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5149627780660425,
+      "learning_rate": 7.971415196763087e-07,
+      "loss": 0.1882,
+      "step": 4995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.48207292341649965,
+      "learning_rate": 7.97058454671429e-07,
+      "loss": 0.2135,
+      "step": 4996
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.451393549295626,
+      "learning_rate": 7.969753769934385e-07,
+      "loss": 0.1189,
+      "step": 4997
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8268227433301185,
+      "learning_rate": 7.968922866458812e-07,
+      "loss": 0.0217,
+      "step": 4998
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8272394906628622,
+      "learning_rate": 7.968091836323024e-07,
+      "loss": 0.1134,
+      "step": 4999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.47506817233651355,
+      "learning_rate": 7.967260679562469e-07,
+      "loss": 0.2445,
+      "step": 5000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5423948336077351,
+      "learning_rate": 7.966429396212609e-07,
+      "loss": 0.3635,
+      "step": 5001
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.06826896756584,
+      "learning_rate": 7.96559798630891e-07,
+      "loss": 0.2838,
+      "step": 5002
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.416550947442645,
+      "learning_rate": 7.964766449886837e-07,
+      "loss": 0.1613,
+      "step": 5003
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4322128171498775,
+      "learning_rate": 7.963934786981869e-07,
+      "loss": 0.0897,
+      "step": 5004
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3111004790221444,
+      "learning_rate": 7.963102997629483e-07,
+      "loss": 0.0062,
+      "step": 5005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4275767116908408,
+      "learning_rate": 7.962271081865168e-07,
+      "loss": 0.088,
+      "step": 5006
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6975253562041295,
+      "learning_rate": 7.961439039724413e-07,
+      "loss": 0.1739,
+      "step": 5007
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4395987336849303,
+      "learning_rate": 7.960606871242715e-07,
+      "loss": 0.0743,
+      "step": 5008
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1549968254191867,
+      "learning_rate": 7.959774576455575e-07,
+      "loss": 0.3855,
+      "step": 5009
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9447495405402501,
+      "learning_rate": 7.958942155398505e-07,
+      "loss": 0.2621,
+      "step": 5010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9002141884337671,
+      "learning_rate": 7.958109608107013e-07,
+      "loss": 0.2056,
+      "step": 5011
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.47126443233971976,
+      "learning_rate": 7.957276934616617e-07,
+      "loss": 0.1012,
+      "step": 5012
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8064655018044103,
+      "learning_rate": 7.956444134962845e-07,
+      "loss": 0.209,
+      "step": 5013
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5132628000304005,
+      "learning_rate": 7.955611209181221e-07,
+      "loss": 0.3676,
+      "step": 5014
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7680638476418366,
+      "learning_rate": 7.954778157307282e-07,
+      "loss": 0.2247,
+      "step": 5015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3801852438888622,
+      "learning_rate": 7.953944979376566e-07,
+      "loss": 0.1818,
+      "step": 5016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0267432775653922,
+      "learning_rate": 7.953111675424621e-07,
+      "loss": 0.1018,
+      "step": 5017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3157116719404107,
+      "learning_rate": 7.952278245486994e-07,
+      "loss": 0.0912,
+      "step": 5018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.677428114293082,
+      "learning_rate": 7.951444689599243e-07,
+      "loss": 0.062,
+      "step": 5019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.538623523501602,
+      "learning_rate": 7.950611007796926e-07,
+      "loss": 0.1322,
+      "step": 5020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.734947706556767,
+      "learning_rate": 7.949777200115614e-07,
+      "loss": 0.3189,
+      "step": 5021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5005699296473773,
+      "learning_rate": 7.948943266590877e-07,
+      "loss": 0.1052,
+      "step": 5022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.956372170598805,
+      "learning_rate": 7.948109207258291e-07,
+      "loss": 0.038,
+      "step": 5023
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.29887402833314614,
+      "learning_rate": 7.947275022153442e-07,
+      "loss": 0.1527,
+      "step": 5024
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5773722082686125,
+      "learning_rate": 7.946440711311913e-07,
+      "loss": 0.0651,
+      "step": 5025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7473159928478327,
+      "learning_rate": 7.9456062747693e-07,
+      "loss": 0.3616,
+      "step": 5026
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1540229629180792,
+      "learning_rate": 7.944771712561205e-07,
+      "loss": 0.1379,
+      "step": 5027
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.402763511775212,
+      "learning_rate": 7.943937024723227e-07,
+      "loss": 0.1207,
+      "step": 5028
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1898707403402686,
+      "learning_rate": 7.943102211290978e-07,
+      "loss": 0.157,
+      "step": 5029
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4718896143639624,
+      "learning_rate": 7.942267272300073e-07,
+      "loss": 0.3093,
+      "step": 5030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6827378006526623,
+      "learning_rate": 7.941432207786129e-07,
+      "loss": 0.3476,
+      "step": 5031
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.838382983533253,
+      "learning_rate": 7.940597017784777e-07,
+      "loss": 0.5095,
+      "step": 5032
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.25771237914452,
+      "learning_rate": 7.939761702331643e-07,
+      "loss": 0.0783,
+      "step": 5033
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6697000274673348,
+      "learning_rate": 7.938926261462365e-07,
+      "loss": 0.1997,
+      "step": 5034
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.45422559182740535,
+      "learning_rate": 7.938090695212586e-07,
+      "loss": 0.1168,
+      "step": 5035
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.749265570823233,
+      "learning_rate": 7.937255003617951e-07,
+      "loss": 0.177,
+      "step": 5036
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2582617138738326,
+      "learning_rate": 7.936419186714112e-07,
+      "loss": 0.2065,
+      "step": 5037
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1733341664774195,
+      "learning_rate": 7.935583244536729e-07,
+      "loss": 0.1761,
+      "step": 5038
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.290772020618876,
+      "learning_rate": 7.934747177121462e-07,
+      "loss": 0.1028,
+      "step": 5039
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6410520876912739,
+      "learning_rate": 7.933910984503982e-07,
+      "loss": 0.2144,
+      "step": 5040
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6934250674178204,
+      "learning_rate": 7.933074666719961e-07,
+      "loss": 0.0993,
+      "step": 5041
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.047608512089884,
+      "learning_rate": 7.932238223805078e-07,
+      "loss": 0.2721,
+      "step": 5042
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.18622970671123631,
+      "learning_rate": 7.93140165579502e-07,
+      "loss": 0.0026,
+      "step": 5043
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4924091168199843,
+      "learning_rate": 7.930564962725474e-07,
+      "loss": 0.0448,
+      "step": 5044
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0139061701689729,
+      "learning_rate": 7.929728144632134e-07,
+      "loss": 0.2306,
+      "step": 5045
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.42514525358423894,
+      "learning_rate": 7.928891201550702e-07,
+      "loss": 0.1029,
+      "step": 5046
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6606443876500947,
+      "learning_rate": 7.928054133516884e-07,
+      "loss": 0.0526,
+      "step": 5047
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5533387802104182,
+      "learning_rate": 7.92721694056639e-07,
+      "loss": 0.2507,
+      "step": 5048
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.392260737879013,
+      "learning_rate": 7.926379622734938e-07,
+      "loss": 0.0962,
+      "step": 5049
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6639063362367161,
+      "learning_rate": 7.925542180058248e-07,
+      "loss": 0.059,
+      "step": 5050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5527057815532743,
+      "learning_rate": 7.924704612572048e-07,
+      "loss": 0.1333,
+      "step": 5051
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8813921275691163,
+      "learning_rate": 7.923866920312067e-07,
+      "loss": 0.245,
+      "step": 5052
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6499958517425046,
+      "learning_rate": 7.923029103314049e-07,
+      "loss": 0.2011,
+      "step": 5053
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6815565286916628,
+      "learning_rate": 7.922191161613731e-07,
+      "loss": 0.0448,
+      "step": 5054
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9708873600600691,
+      "learning_rate": 7.921353095246866e-07,
+      "loss": 0.2493,
+      "step": 5055
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3973793893255023,
+      "learning_rate": 7.920514904249204e-07,
+      "loss": 0.095,
+      "step": 5056
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4297452153589135,
+      "learning_rate": 7.919676588656505e-07,
+      "loss": 0.1796,
+      "step": 5057
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8032506156239824,
+      "learning_rate": 7.918838148504535e-07,
+      "loss": 0.1595,
+      "step": 5058
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.622115335804525,
+      "learning_rate": 7.91799958382906e-07,
+      "loss": 0.1206,
+      "step": 5059
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7769521696973315,
+      "learning_rate": 7.917160894665858e-07,
+      "loss": 0.1069,
+      "step": 5060
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2420123982376446,
+      "learning_rate": 7.916322081050709e-07,
+      "loss": 0.1345,
+      "step": 5061
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3595717039836064,
+      "learning_rate": 7.915483143019395e-07,
+      "loss": 0.1815,
+      "step": 5062
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5631361585386891,
+      "learning_rate": 7.91464408060771e-07,
+      "loss": 0.2543,
+      "step": 5063
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5781748938014992,
+      "learning_rate": 7.91380489385145e-07,
+      "loss": 0.321,
+      "step": 5064
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4852625379275442,
+      "learning_rate": 7.912965582786415e-07,
+      "loss": 0.2118,
+      "step": 5065
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5333212149403556,
+      "learning_rate": 7.912126147448413e-07,
+      "loss": 0.1468,
+      "step": 5066
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.18167122602406438,
+      "learning_rate": 7.911286587873256e-07,
+      "loss": 0.1121,
+      "step": 5067
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3149254689417239,
+      "learning_rate": 7.910446904096759e-07,
+      "loss": 0.1077,
+      "step": 5068
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.907826203685545,
+      "learning_rate": 7.909607096154749e-07,
+      "loss": 0.318,
+      "step": 5069
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2544560723678784,
+      "learning_rate": 7.908767164083049e-07,
+      "loss": 0.3798,
+      "step": 5070
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1243881852506736,
+      "learning_rate": 7.907927107917495e-07,
+      "loss": 0.1088,
+      "step": 5071
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4039041042046467,
+      "learning_rate": 7.907086927693925e-07,
+      "loss": 0.1204,
+      "step": 5072
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.337441881657516,
+      "learning_rate": 7.906246623448183e-07,
+      "loss": 0.2662,
+      "step": 5073
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3869484615283691,
+      "learning_rate": 7.905406195216117e-07,
+      "loss": 0.194,
+      "step": 5074
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3772397922865434,
+      "learning_rate": 7.904565643033583e-07,
+      "loss": 0.1906,
+      "step": 5075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.075791385822625,
+      "learning_rate": 7.90372496693644e-07,
+      "loss": 0.2967,
+      "step": 5076
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1951627905017583,
+      "learning_rate": 7.902884166960551e-07,
+      "loss": 0.0529,
+      "step": 5077
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4299172986861356,
+      "learning_rate": 7.902043243141789e-07,
+      "loss": 0.1034,
+      "step": 5078
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27911503588386616,
+      "learning_rate": 7.901202195516028e-07,
+      "loss": 0.0044,
+      "step": 5079
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3465537627965502,
+      "learning_rate": 7.90036102411915e-07,
+      "loss": 0.0845,
+      "step": 5080
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8558313789097676,
+      "learning_rate": 7.89951972898704e-07,
+      "loss": 0.169,
+      "step": 5081
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4629701720174717,
+      "learning_rate": 7.898678310155589e-07,
+      "loss": 0.1314,
+      "step": 5082
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5101844974842896,
+      "learning_rate": 7.897836767660695e-07,
+      "loss": 0.1962,
+      "step": 5083
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5551973788374136,
+      "learning_rate": 7.896995101538259e-07,
+      "loss": 0.1197,
+      "step": 5084
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5410604563063769,
+      "learning_rate": 7.896153311824188e-07,
+      "loss": 0.245,
+      "step": 5085
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.950281909471845,
+      "learning_rate": 7.895311398554394e-07,
+      "loss": 0.2554,
+      "step": 5086
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8591973063541584,
+      "learning_rate": 7.894469361764798e-07,
+      "loss": 0.1654,
+      "step": 5087
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.522447476544494,
+      "learning_rate": 7.893627201491318e-07,
+      "loss": 0.2683,
+      "step": 5088
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5860219876403301,
+      "learning_rate": 7.892784917769884e-07,
+      "loss": 0.1917,
+      "step": 5089
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3691056995316721,
+      "learning_rate": 7.891942510636431e-07,
+      "loss": 0.1302,
+      "step": 5090
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5598536136114467,
+      "learning_rate": 7.891099980126899e-07,
+      "loss": 0.4858,
+      "step": 5091
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6224550751677805,
+      "learning_rate": 7.890257326277227e-07,
+      "loss": 0.1352,
+      "step": 5092
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5249354254426418,
+      "learning_rate": 7.889414549123369e-07,
+      "loss": 0.0301,
+      "step": 5093
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5549734263590133,
+      "learning_rate": 7.888571648701277e-07,
+      "loss": 0.4364,
+      "step": 5094
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.252075031879599,
+      "learning_rate": 7.887728625046912e-07,
+      "loss": 0.2175,
+      "step": 5095
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.577590550688623,
+      "learning_rate": 7.88688547819624e-07,
+      "loss": 0.2646,
+      "step": 5096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9450411854298872,
+      "learning_rate": 7.886042208185229e-07,
+      "loss": 0.2347,
+      "step": 5097
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5053901836705134,
+      "learning_rate": 7.885198815049857e-07,
+      "loss": 0.2114,
+      "step": 5098
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3266340655180504,
+      "learning_rate": 7.884355298826102e-07,
+      "loss": 0.1107,
+      "step": 5099
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9160849313524885,
+      "learning_rate": 7.883511659549952e-07,
+      "loss": 0.1369,
+      "step": 5100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8819874018164324,
+      "learning_rate": 7.882667897257398e-07,
+      "loss": 0.3088,
+      "step": 5101
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6339437987921024,
+      "learning_rate": 7.881824011984437e-07,
+      "loss": 0.2548,
+      "step": 5102
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.18967550245166595,
+      "learning_rate": 7.880980003767071e-07,
+      "loss": 0.0126,
+      "step": 5103
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7208254564609355,
+      "learning_rate": 7.880135872641305e-07,
+      "loss": 0.2752,
+      "step": 5104
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4640709049335785,
+      "learning_rate": 7.879291618643154e-07,
+      "loss": 0.1532,
+      "step": 5105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8370361202927852,
+      "learning_rate": 7.878447241808634e-07,
+      "loss": 0.0573,
+      "step": 5106
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7902582978217804,
+      "learning_rate": 7.877602742173768e-07,
+      "loss": 0.1405,
+      "step": 5107
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6673761349895175,
+      "learning_rate": 7.876758119774585e-07,
+      "loss": 0.0331,
+      "step": 5108
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9481514783975585,
+      "learning_rate": 7.875913374647116e-07,
+      "loss": 0.2157,
+      "step": 5109
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7219554855726629,
+      "learning_rate": 7.875068506827403e-07,
+      "loss": 0.1573,
+      "step": 5110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.723677411654189,
+      "learning_rate": 7.874223516351487e-07,
+      "loss": 0.1056,
+      "step": 5111
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.314401714303733,
+      "learning_rate": 7.873378403255419e-07,
+      "loss": 0.209,
+      "step": 5112
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9923260613096916,
+      "learning_rate": 7.872533167575251e-07,
+      "loss": 0.1989,
+      "step": 5113
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3756478802285395,
+      "learning_rate": 7.871687809347045e-07,
+      "loss": 0.2573,
+      "step": 5114
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.960812771431285,
+      "learning_rate": 7.870842328606862e-07,
+      "loss": 0.1823,
+      "step": 5115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7207169323889557,
+      "learning_rate": 7.869996725390775e-07,
+      "loss": 0.2595,
+      "step": 5116
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7463231305737968,
+      "learning_rate": 7.869150999734859e-07,
+      "loss": 0.0896,
+      "step": 5117
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.819749616163352,
+      "learning_rate": 7.868305151675192e-07,
+      "loss": 0.2348,
+      "step": 5118
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1100931362103976,
+      "learning_rate": 7.867459181247863e-07,
+      "loss": 0.2892,
+      "step": 5119
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.511772839504525,
+      "learning_rate": 7.866613088488961e-07,
+      "loss": 0.2312,
+      "step": 5120
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3402164407726527,
+      "learning_rate": 7.86576687343458e-07,
+      "loss": 0.0916,
+      "step": 5121
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9386732144001584,
+      "learning_rate": 7.864920536120825e-07,
+      "loss": 0.5051,
+      "step": 5122
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5137049325530657,
+      "learning_rate": 7.864074076583799e-07,
+      "loss": 0.2519,
+      "step": 5123
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1003313269417925,
+      "learning_rate": 7.863227494859619e-07,
+      "loss": 0.0859,
+      "step": 5124
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6060528816479874,
+      "learning_rate": 7.862380790984396e-07,
+      "loss": 0.2056,
+      "step": 5125
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6734469477166636,
+      "learning_rate": 7.861533964994254e-07,
+      "loss": 0.1005,
+      "step": 5126
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9326064729223768,
+      "learning_rate": 7.860687016925321e-07,
+      "loss": 0.1176,
+      "step": 5127
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5140312436999763,
+      "learning_rate": 7.859839946813729e-07,
+      "loss": 0.1963,
+      "step": 5128
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.046058161699342,
+      "learning_rate": 7.858992754695617e-07,
+      "loss": 0.1158,
+      "step": 5129
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5555687076704139,
+      "learning_rate": 7.858145440607124e-07,
+      "loss": 0.2217,
+      "step": 5130
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0269123056921516,
+      "learning_rate": 7.857298004584403e-07,
+      "loss": 0.1093,
+      "step": 5131
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.51625517947423,
+      "learning_rate": 7.856450446663605e-07,
+      "loss": 0.187,
+      "step": 5132
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.529660144764917,
+      "learning_rate": 7.855602766880889e-07,
+      "loss": 0.1998,
+      "step": 5133
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7936212130851568,
+      "learning_rate": 7.854754965272419e-07,
+      "loss": 0.1397,
+      "step": 5134
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.38627722105736007,
+      "learning_rate": 7.853907041874361e-07,
+      "loss": 0.2149,
+      "step": 5135
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7981630064468679,
+      "learning_rate": 7.853058996722895e-07,
+      "loss": 0.1107,
+      "step": 5136
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4979561136612249,
+      "learning_rate": 7.852210829854193e-07,
+      "loss": 0.2579,
+      "step": 5137
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6970440182009874,
+      "learning_rate": 7.851362541304446e-07,
+      "loss": 0.2837,
+      "step": 5138
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5434395421011616,
+      "learning_rate": 7.85051413110984e-07,
+      "loss": 0.0616,
+      "step": 5139
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9180870135310615,
+      "learning_rate": 7.849665599306572e-07,
+      "loss": 0.3237,
+      "step": 5140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7764686954870078,
+      "learning_rate": 7.848816945930841e-07,
+      "loss": 0.1006,
+      "step": 5141
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.104607388805604,
+      "learning_rate": 7.84796817101885e-07,
+      "loss": 0.2477,
+      "step": 5142
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.474006178564293,
+      "learning_rate": 7.847119274606814e-07,
+      "loss": 0.2312,
+      "step": 5143
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7354791717268591,
+      "learning_rate": 7.846270256730945e-07,
+      "loss": 0.4119,
+      "step": 5144
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8495632459382707,
+      "learning_rate": 7.845421117427466e-07,
+      "loss": 0.2803,
+      "step": 5145
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9405873283152801,
+      "learning_rate": 7.844571856732601e-07,
+      "loss": 0.275,
+      "step": 5146
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1669588160239925,
+      "learning_rate": 7.843722474682583e-07,
+      "loss": 0.258,
+      "step": 5147
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.41844449084585966,
+      "learning_rate": 7.842872971313648e-07,
+      "loss": 0.0111,
+      "step": 5148
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.18090898926874738,
+      "learning_rate": 7.842023346662036e-07,
+      "loss": 0.0212,
+      "step": 5149
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9171412220481466,
+      "learning_rate": 7.841173600763996e-07,
+      "loss": 0.2245,
+      "step": 5150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.40540167800773685,
+      "learning_rate": 7.840323733655778e-07,
+      "loss": 0.1728,
+      "step": 5151
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.37619284814137943,
+      "learning_rate": 7.839473745373642e-07,
+      "loss": 0.1055,
+      "step": 5152
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.264378141577581,
+      "learning_rate": 7.838623635953845e-07,
+      "loss": 0.0205,
+      "step": 5153
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.866356084279496,
+      "learning_rate": 7.83777340543266e-07,
+      "loss": 0.1617,
+      "step": 5154
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7757190835594977,
+      "learning_rate": 7.836923053846354e-07,
+      "loss": 0.2728,
+      "step": 5155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3825698622205671,
+      "learning_rate": 7.836072581231209e-07,
+      "loss": 0.1509,
+      "step": 5156
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8802356116921449,
+      "learning_rate": 7.835221987623506e-07,
+      "loss": 0.2452,
+      "step": 5157
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3535140613776956,
+      "learning_rate": 7.834371273059534e-07,
+      "loss": 0.2698,
+      "step": 5158
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7063955805418285,
+      "learning_rate": 7.833520437575585e-07,
+      "loss": 0.4772,
+      "step": 5159
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8677992657140369,
+      "learning_rate": 7.832669481207958e-07,
+      "loss": 0.1712,
+      "step": 5160
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0564033872324807,
+      "learning_rate": 7.831818403992958e-07,
+      "loss": 0.2168,
+      "step": 5161
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1737715122182201,
+      "learning_rate": 7.830967205966892e-07,
+      "loss": 0.0726,
+      "step": 5162
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4813717911308954,
+      "learning_rate": 7.830115887166072e-07,
+      "loss": 0.0935,
+      "step": 5163
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.21965268421480677,
+      "learning_rate": 7.82926444762682e-07,
+      "loss": 0.0828,
+      "step": 5164
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6091583475193905,
+      "learning_rate": 7.82841288738546e-07,
+      "loss": 0.1261,
+      "step": 5165
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.41979584647571394,
+      "learning_rate": 7.82756120647832e-07,
+      "loss": 0.012,
+      "step": 5166
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3935984589136889,
+      "learning_rate": 7.826709404941735e-07,
+      "loss": 0.0103,
+      "step": 5167
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.822421978783359,
+      "learning_rate": 7.825857482812044e-07,
+      "loss": 0.1331,
+      "step": 5168
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1806423913775659,
+      "learning_rate": 7.825005440125593e-07,
+      "loss": 0.2662,
+      "step": 5169
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.15141360722514163,
+      "learning_rate": 7.824153276918731e-07,
+      "loss": 0.0022,
+      "step": 5170
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.003885880421857,
+      "learning_rate": 7.82330099322781e-07,
+      "loss": 0.3928,
+      "step": 5171
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.43733114030772724,
+      "learning_rate": 7.822448589089197e-07,
+      "loss": 0.1057,
+      "step": 5172
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6463146289694512,
+      "learning_rate": 7.821596064539251e-07,
+      "loss": 0.1843,
+      "step": 5173
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7131295128071986,
+      "learning_rate": 7.820743419614345e-07,
+      "loss": 0.2946,
+      "step": 5174
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3734179210841508,
+      "learning_rate": 7.819890654350855e-07,
+      "loss": 0.0989,
+      "step": 5175
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0249790109579513,
+      "learning_rate": 7.819037768785159e-07,
+      "loss": 0.3991,
+      "step": 5176
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4857390485229206,
+      "learning_rate": 7.818184762953648e-07,
+      "loss": 0.0315,
+      "step": 5177
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.364107220300501,
+      "learning_rate": 7.817331636892708e-07,
+      "loss": 0.0135,
+      "step": 5178
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6163913761357585,
+      "learning_rate": 7.816478390638737e-07,
+      "loss": 0.0153,
+      "step": 5179
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7892463179352985,
+      "learning_rate": 7.815625024228136e-07,
+      "loss": 0.198,
+      "step": 5180
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4749420067595575,
+      "learning_rate": 7.814771537697311e-07,
+      "loss": 0.1432,
+      "step": 5181
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6439197328856424,
+      "learning_rate": 7.813917931082675e-07,
+      "loss": 0.1727,
+      "step": 5182
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9286148777239651,
+      "learning_rate": 7.813064204420643e-07,
+      "loss": 0.2073,
+      "step": 5183
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5687769427647333,
+      "learning_rate": 7.812210357747635e-07,
+      "loss": 0.1604,
+      "step": 5184
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1008413881744683,
+      "learning_rate": 7.811356391100081e-07,
+      "loss": 0.2216,
+      "step": 5185
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2699507225473614,
+      "learning_rate": 7.810502304514413e-07,
+      "loss": 0.2641,
+      "step": 5186
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1134241983281936,
+      "learning_rate": 7.809648098027066e-07,
+      "loss": 0.3247,
+      "step": 5187
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.625339701996283,
+      "learning_rate": 7.808793771674484e-07,
+      "loss": 0.2065,
+      "step": 5188
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3914872799377487,
+      "learning_rate": 7.807939325493112e-07,
+      "loss": 0.2097,
+      "step": 5189
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.376255327123015,
+      "learning_rate": 7.807084759519404e-07,
+      "loss": 0.1322,
+      "step": 5190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6763194881524864,
+      "learning_rate": 7.806230073789818e-07,
+      "loss": 0.3152,
+      "step": 5191
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.863660255909624,
+      "learning_rate": 7.805375268340814e-07,
+      "loss": 0.0923,
+      "step": 5192
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8275494096837781,
+      "learning_rate": 7.804520343208865e-07,
+      "loss": 0.2021,
+      "step": 5193
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.269752146604219,
+      "learning_rate": 7.803665298430437e-07,
+      "loss": 0.1851,
+      "step": 5194
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.699971489648117,
+      "learning_rate": 7.802810134042013e-07,
+      "loss": 0.1585,
+      "step": 5195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.795300966736381,
+      "learning_rate": 7.801954850080074e-07,
+      "loss": 0.105,
+      "step": 5196
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5764921562418399,
+      "learning_rate": 7.801099446581107e-07,
+      "loss": 0.2153,
+      "step": 5197
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5092916387115642,
+      "learning_rate": 7.800243923581609e-07,
+      "loss": 0.0748,
+      "step": 5198
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.45864637235860056,
+      "learning_rate": 7.799388281118076e-07,
+      "loss": 0.1253,
+      "step": 5199
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2302589845756009,
+      "learning_rate": 7.79853251922701e-07,
+      "loss": 0.1033,
+      "step": 5200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6796973251578036,
+      "learning_rate": 7.797676637944921e-07,
+      "loss": 0.2755,
+      "step": 5201
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9209543229656066,
+      "learning_rate": 7.796820637308323e-07,
+      "loss": 0.3127,
+      "step": 5202
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5949097161132205,
+      "learning_rate": 7.795964517353733e-07,
+      "loss": 0.3117,
+      "step": 5203
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6011857914950647,
+      "learning_rate": 7.795108278117678e-07,
+      "loss": 0.0846,
+      "step": 5204
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7723605848731894,
+      "learning_rate": 7.794251919636685e-07,
+      "loss": 0.1317,
+      "step": 5205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.42760420529026383,
+      "learning_rate": 7.793395441947287e-07,
+      "loss": 0.1835,
+      "step": 5206
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8185202619388978,
+      "learning_rate": 7.792538845086024e-07,
+      "loss": 0.1726,
+      "step": 5207
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3600276855074998,
+      "learning_rate": 7.79168212908944e-07,
+      "loss": 0.194,
+      "step": 5208
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1362800072731472,
+      "learning_rate": 7.790825293994086e-07,
+      "loss": 0.2714,
+      "step": 5209
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.039725504216318,
+      "learning_rate": 7.789968339836514e-07,
+      "loss": 0.2195,
+      "step": 5210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2588782562309624,
+      "learning_rate": 7.789111266653283e-07,
+      "loss": 0.3284,
+      "step": 5211
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9196694656825979,
+      "learning_rate": 7.78825407448096e-07,
+      "loss": 0.0946,
+      "step": 5212
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8374289363909836,
+      "learning_rate": 7.787396763356111e-07,
+      "loss": 0.3104,
+      "step": 5213
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7024580538454515,
+      "learning_rate": 7.786539333315315e-07,
+      "loss": 0.1579,
+      "step": 5214
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7842770416764756,
+      "learning_rate": 7.785681784395148e-07,
+      "loss": 0.2168,
+      "step": 5215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9262774147299662,
+      "learning_rate": 7.784824116632196e-07,
+      "loss": 0.4278,
+      "step": 5216
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8510804981741832,
+      "learning_rate": 7.78396633006305e-07,
+      "loss": 0.261,
+      "step": 5217
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5361973129473463,
+      "learning_rate": 7.783108424724303e-07,
+      "loss": 0.399,
+      "step": 5218
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.024259909503765,
+      "learning_rate": 7.782250400652556e-07,
+      "loss": 0.0766,
+      "step": 5219
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6120064942509154,
+      "learning_rate": 7.781392257884415e-07,
+      "loss": 0.1121,
+      "step": 5220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3654686798068923,
+      "learning_rate": 7.780533996456489e-07,
+      "loss": 0.0905,
+      "step": 5221
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6877879776820551,
+      "learning_rate": 7.779675616405391e-07,
+      "loss": 0.2864,
+      "step": 5222
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8962972809336581,
+      "learning_rate": 7.778817117767747e-07,
+      "loss": 0.2471,
+      "step": 5223
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1438333180741294,
+      "learning_rate": 7.777958500580175e-07,
+      "loss": 0.2553,
+      "step": 5224
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.01063537119916,
+      "learning_rate": 7.777099764879311e-07,
+      "loss": 0.1714,
+      "step": 5225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.547708033777952,
+      "learning_rate": 7.776240910701787e-07,
+      "loss": 0.2227,
+      "step": 5226
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9399754562836553,
+      "learning_rate": 7.775381938084245e-07,
+      "loss": 0.1651,
+      "step": 5227
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.573631678102885,
+      "learning_rate": 7.77452284706333e-07,
+      "loss": 0.2619,
+      "step": 5228
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8042545184195301,
+      "learning_rate": 7.773663637675694e-07,
+      "loss": 0.3167,
+      "step": 5229
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7307787292330183,
+      "learning_rate": 7.77280430995799e-07,
+      "loss": 0.357,
+      "step": 5230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.41078368085224987,
+      "learning_rate": 7.771944863946882e-07,
+      "loss": 0.1782,
+      "step": 5231
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5626768386257761,
+      "learning_rate": 7.771085299679033e-07,
+      "loss": 0.1987,
+      "step": 5232
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.30757453219627906,
+      "learning_rate": 7.770225617191115e-07,
+      "loss": 0.1097,
+      "step": 5233
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7249081777321619,
+      "learning_rate": 7.769365816519802e-07,
+      "loss": 0.0807,
+      "step": 5234
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9128744385671228,
+      "learning_rate": 7.768505897701777e-07,
+      "loss": 0.2948,
+      "step": 5235
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5775201829528736,
+      "learning_rate": 7.767645860773725e-07,
+      "loss": 0.1744,
+      "step": 5236
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7369713102491516,
+      "learning_rate": 7.766785705772338e-07,
+      "loss": 0.1896,
+      "step": 5237
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.35607296837526103,
+      "learning_rate": 7.765925432734309e-07,
+      "loss": 0.186,
+      "step": 5238
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4936113410080766,
+      "learning_rate": 7.765065041696341e-07,
+      "loss": 0.1866,
+      "step": 5239
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6865003619440185,
+      "learning_rate": 7.764204532695141e-07,
+      "loss": 0.17,
+      "step": 5240
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3432388715961265,
+      "learning_rate": 7.763343905767419e-07,
+      "loss": 0.1504,
+      "step": 5241
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6067028995860944,
+      "learning_rate": 7.762483160949888e-07,
+      "loss": 0.1617,
+      "step": 5242
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9284293087567614,
+      "learning_rate": 7.761622298279276e-07,
+      "loss": 0.1262,
+      "step": 5243
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6042827755347948,
+      "learning_rate": 7.760761317792303e-07,
+      "loss": 0.0869,
+      "step": 5244
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6440464766518622,
+      "learning_rate": 7.759900219525703e-07,
+      "loss": 0.3204,
+      "step": 5245
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6139440459427701,
+      "learning_rate": 7.759039003516211e-07,
+      "loss": 0.0814,
+      "step": 5246
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3709636354163115,
+      "learning_rate": 7.758177669800568e-07,
+      "loss": 0.1332,
+      "step": 5247
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5544713127409657,
+      "learning_rate": 7.757316218415523e-07,
+      "loss": 0.2422,
+      "step": 5248
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.157099304716162,
+      "learning_rate": 7.756454649397824e-07,
+      "loss": 0.052,
+      "step": 5249
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6744716019691676,
+      "learning_rate": 7.75559296278423e-07,
+      "loss": 0.1596,
+      "step": 5250
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6174406562037895,
+      "learning_rate": 7.754731158611498e-07,
+      "loss": 0.3179,
+      "step": 5251
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5190619052521371,
+      "learning_rate": 7.753869236916399e-07,
+      "loss": 0.3108,
+      "step": 5252
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9681623585174963,
+      "learning_rate": 7.753007197735703e-07,
+      "loss": 0.0746,
+      "step": 5253
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7248986660787387,
+      "learning_rate": 7.752145041106184e-07,
+      "loss": 0.2915,
+      "step": 5254
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8032978124242338,
+      "learning_rate": 7.751282767064626e-07,
+      "loss": 0.1974,
+      "step": 5255
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7604321560764785,
+      "learning_rate": 7.750420375647815e-07,
+      "loss": 0.3739,
+      "step": 5256
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8691280985179755,
+      "learning_rate": 7.74955786689254e-07,
+      "loss": 0.3987,
+      "step": 5257
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7048945497271688,
+      "learning_rate": 7.7486952408356e-07,
+      "loss": 0.1943,
+      "step": 5258
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.26554239065647,
+      "learning_rate": 7.747832497513795e-07,
+      "loss": 0.2652,
+      "step": 5259
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43491317356013665,
+      "learning_rate": 7.746969636963933e-07,
+      "loss": 0.2595,
+      "step": 5260
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6478578418172165,
+      "learning_rate": 7.746106659222823e-07,
+      "loss": 0.25,
+      "step": 5261
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1504782887633436,
+      "learning_rate": 7.745243564327283e-07,
+      "loss": 0.1144,
+      "step": 5262
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.101033691732244,
+      "learning_rate": 7.744380352314134e-07,
+      "loss": 0.0991,
+      "step": 5263
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8055943990726835,
+      "learning_rate": 7.743517023220203e-07,
+      "loss": 0.3228,
+      "step": 5264
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7353106485944705,
+      "learning_rate": 7.742653577082318e-07,
+      "loss": 0.1819,
+      "step": 5265
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0442489787901628,
+      "learning_rate": 7.741790013937321e-07,
+      "loss": 0.3745,
+      "step": 5266
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4147080737370854,
+      "learning_rate": 7.740926333822049e-07,
+      "loss": 0.298,
+      "step": 5267
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4203104667063722,
+      "learning_rate": 7.74006253677335e-07,
+      "loss": 0.2799,
+      "step": 5268
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5690448686724064,
+      "learning_rate": 7.739198622828073e-07,
+      "loss": 0.0067,
+      "step": 5269
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.37927647647532053,
+      "learning_rate": 7.738334592023079e-07,
+      "loss": 0.2907,
+      "step": 5270
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9868200960221822,
+      "learning_rate": 7.737470444395226e-07,
+      "loss": 0.3532,
+      "step": 5271
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.577393204628036,
+      "learning_rate": 7.73660617998138e-07,
+      "loss": 0.3059,
+      "step": 5272
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.33371205203811,
+      "learning_rate": 7.735741798818414e-07,
+      "loss": 0.1558,
+      "step": 5273
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.547719657318448,
+      "learning_rate": 7.734877300943202e-07,
+      "loss": 0.1946,
+      "step": 5274
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9383937743435579,
+      "learning_rate": 7.734012686392628e-07,
+      "loss": 0.1982,
+      "step": 5275
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9772115285261509,
+      "learning_rate": 7.733147955203576e-07,
+      "loss": 0.1183,
+      "step": 5276
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6408425941334519,
+      "learning_rate": 7.732283107412938e-07,
+      "loss": 0.2941,
+      "step": 5277
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2603429994239863,
+      "learning_rate": 7.731418143057611e-07,
+      "loss": 0.2918,
+      "step": 5278
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.190932591688707,
+      "learning_rate": 7.730553062174494e-07,
+      "loss": 0.4235,
+      "step": 5279
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9899504115425867,
+      "learning_rate": 7.729687864800494e-07,
+      "loss": 0.0565,
+      "step": 5280
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0326428444784788,
+      "learning_rate": 7.728822550972522e-07,
+      "loss": 0.016,
+      "step": 5281
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7889003525642555,
+      "learning_rate": 7.727957120727495e-07,
+      "loss": 0.0446,
+      "step": 5282
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4504325124157612,
+      "learning_rate": 7.727091574102334e-07,
+      "loss": 0.1024,
+      "step": 5283
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5993252725635362,
+      "learning_rate": 7.726225911133965e-07,
+      "loss": 0.2273,
+      "step": 5284
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6508119301918696,
+      "learning_rate": 7.725360131859317e-07,
+      "loss": 0.0937,
+      "step": 5285
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7281171500449641,
+      "learning_rate": 7.724494236315327e-07,
+      "loss": 0.3123,
+      "step": 5286
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7887649925558102,
+      "learning_rate": 7.723628224538937e-07,
+      "loss": 0.0142,
+      "step": 5287
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3721712379544822,
+      "learning_rate": 7.722762096567089e-07,
+      "loss": 0.1807,
+      "step": 5288
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.570330342971187,
+      "learning_rate": 7.721895852436739e-07,
+      "loss": 0.1508,
+      "step": 5289
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.017985400053795,
+      "learning_rate": 7.72102949218484e-07,
+      "loss": 0.1733,
+      "step": 5290
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6570959670651932,
+      "learning_rate": 7.720163015848352e-07,
+      "loss": 0.1342,
+      "step": 5291
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.31935624622912157,
+      "learning_rate": 7.719296423464243e-07,
+      "loss": 0.0046,
+      "step": 5292
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.02227589469212,
+      "learning_rate": 7.718429715069481e-07,
+      "loss": 0.2811,
+      "step": 5293
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0304873467200701,
+      "learning_rate": 7.717562890701043e-07,
+      "loss": 0.278,
+      "step": 5294
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0786198227661057,
+      "learning_rate": 7.716695950395908e-07,
+      "loss": 0.2424,
+      "step": 5295
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.161837307128368,
+      "learning_rate": 7.715828894191063e-07,
+      "loss": 0.1475,
+      "step": 5296
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0594379087178116,
+      "learning_rate": 7.714961722123498e-07,
+      "loss": 0.0915,
+      "step": 5297
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.613491070317713,
+      "learning_rate": 7.71409443423021e-07,
+      "loss": 0.1684,
+      "step": 5298
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.48638051637562585,
+      "learning_rate": 7.713227030548195e-07,
+      "loss": 0.1002,
+      "step": 5299
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9718802165761276,
+      "learning_rate": 7.712359511114461e-07,
+      "loss": 0.2009,
+      "step": 5300
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.755756688591537,
+      "learning_rate": 7.711491875966019e-07,
+      "loss": 0.2435,
+      "step": 5301
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4700821926865363,
+      "learning_rate": 7.710624125139882e-07,
+      "loss": 0.1341,
+      "step": 5302
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7815767874893984,
+      "learning_rate": 7.70975625867307e-07,
+      "loss": 0.2079,
+      "step": 5303
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.41626744975865426,
+      "learning_rate": 7.708888276602609e-07,
+      "loss": 0.2388,
+      "step": 5304
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.498746829276991,
+      "learning_rate": 7.70802017896553e-07,
+      "loss": 0.2425,
+      "step": 5305
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.38162158973715743,
+      "learning_rate": 7.707151965798866e-07,
+      "loss": 0.2501,
+      "step": 5306
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.016607126681126,
+      "learning_rate": 7.706283637139657e-07,
+      "loss": 0.2299,
+      "step": 5307
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7391466955217589,
+      "learning_rate": 7.705415193024947e-07,
+      "loss": 0.1713,
+      "step": 5308
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.020352974328961,
+      "learning_rate": 7.704546633491787e-07,
+      "loss": 0.0339,
+      "step": 5309
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6292426340417446,
+      "learning_rate": 7.703677958577231e-07,
+      "loss": 0.217,
+      "step": 5310
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.829214248599567,
+      "learning_rate": 7.702809168318337e-07,
+      "loss": 0.1083,
+      "step": 5311
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5662842095404212,
+      "learning_rate": 7.701940262752171e-07,
+      "loss": 0.2041,
+      "step": 5312
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.650322089205414,
+      "learning_rate": 7.701071241915802e-07,
+      "loss": 0.2903,
+      "step": 5313
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5903015047276998,
+      "learning_rate": 7.700202105846303e-07,
+      "loss": 0.3205,
+      "step": 5314
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0010473667544635,
+      "learning_rate": 7.699332854580756e-07,
+      "loss": 0.1309,
+      "step": 5315
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.267577849455619,
+      "learning_rate": 7.698463488156241e-07,
+      "loss": 0.1097,
+      "step": 5316
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8435166915977652,
+      "learning_rate": 7.69759400660985e-07,
+      "loss": 0.0634,
+      "step": 5317
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0060092200291906,
+      "learning_rate": 7.696724409978677e-07,
+      "loss": 0.1808,
+      "step": 5318
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6870243220663309,
+      "learning_rate": 7.695854698299819e-07,
+      "loss": 0.0972,
+      "step": 5319
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5387727882044846,
+      "learning_rate": 7.694984871610379e-07,
+      "loss": 0.349,
+      "step": 5320
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.325458812570029,
+      "learning_rate": 7.694114929947469e-07,
+      "loss": 0.1727,
+      "step": 5321
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5208017490544538,
+      "learning_rate": 7.693244873348197e-07,
+      "loss": 0.1576,
+      "step": 5322
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.49418080354995575,
+      "learning_rate": 7.692374701849687e-07,
+      "loss": 0.0074,
+      "step": 5323
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7438736074259235,
+      "learning_rate": 7.691504415489058e-07,
+      "loss": 0.2522,
+      "step": 5324
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.271699322155713,
+      "learning_rate": 7.690634014303441e-07,
+      "loss": 0.3671,
+      "step": 5325
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8152545212787211,
+      "learning_rate": 7.689763498329969e-07,
+      "loss": 0.2574,
+      "step": 5326
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9140083684797955,
+      "learning_rate": 7.688892867605778e-07,
+      "loss": 0.1236,
+      "step": 5327
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6380556696314743,
+      "learning_rate": 7.688022122168012e-07,
+      "loss": 0.319,
+      "step": 5328
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6314997933841358,
+      "learning_rate": 7.68715126205382e-07,
+      "loss": 0.3729,
+      "step": 5329
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8182762182779711,
+      "learning_rate": 7.686280287300352e-07,
+      "loss": 0.0136,
+      "step": 5330
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3156776685054188,
+      "learning_rate": 7.685409197944768e-07,
+      "loss": 0.0872,
+      "step": 5331
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.831457989118222,
+      "learning_rate": 7.684537994024228e-07,
+      "loss": 0.1813,
+      "step": 5332
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7341932180942895,
+      "learning_rate": 7.683666675575901e-07,
+      "loss": 0.1579,
+      "step": 5333
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6616775004102179,
+      "learning_rate": 7.682795242636958e-07,
+      "loss": 0.1285,
+      "step": 5334
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.092916168305884,
+      "learning_rate": 7.681923695244578e-07,
+      "loss": 0.0534,
+      "step": 5335
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1168417053413657,
+      "learning_rate": 7.681052033435942e-07,
+      "loss": 0.2868,
+      "step": 5336
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.866520975312506,
+      "learning_rate": 7.680180257248235e-07,
+      "loss": 0.2316,
+      "step": 5337
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3393368998848683,
+      "learning_rate": 7.679308366718652e-07,
+      "loss": 0.1601,
+      "step": 5338
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.427738334861401,
+      "learning_rate": 7.678436361884388e-07,
+      "loss": 0.2265,
+      "step": 5339
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.454088902427455,
+      "learning_rate": 7.677564242782644e-07,
+      "loss": 0.2034,
+      "step": 5340
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4295246616964228,
+      "learning_rate": 7.676692009450626e-07,
+      "loss": 0.0805,
+      "step": 5341
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.056853686692961,
+      "learning_rate": 7.675819661925547e-07,
+      "loss": 0.3044,
+      "step": 5342
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9531499999715173,
+      "learning_rate": 7.674947200244622e-07,
+      "loss": 0.1795,
+      "step": 5343
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.417559042119717,
+      "learning_rate": 7.67407462444507e-07,
+      "loss": 0.1089,
+      "step": 5344
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6925665703216968,
+      "learning_rate": 7.673201934564122e-07,
+      "loss": 0.1718,
+      "step": 5345
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2011792930434808,
+      "learning_rate": 7.672329130639005e-07,
+      "loss": 0.3044,
+      "step": 5346
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5730655337707138,
+      "learning_rate": 7.671456212706956e-07,
+      "loss": 0.1644,
+      "step": 5347
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3752020337562845,
+      "learning_rate": 7.670583180805213e-07,
+      "loss": 0.1373,
+      "step": 5348
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6841256839270571,
+      "learning_rate": 7.669710034971024e-07,
+      "loss": 0.2897,
+      "step": 5349
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4700650050761052,
+      "learning_rate": 7.668836775241638e-07,
+      "loss": 0.0782,
+      "step": 5350
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6904821541206747,
+      "learning_rate": 7.667963401654308e-07,
+      "loss": 0.3029,
+      "step": 5351
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.25120294168718676,
+      "learning_rate": 7.667089914246299e-07,
+      "loss": 0.1555,
+      "step": 5352
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.77480443285652,
+      "learning_rate": 7.666216313054871e-07,
+      "loss": 0.2512,
+      "step": 5353
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9440982341360855,
+      "learning_rate": 7.665342598117296e-07,
+      "loss": 0.4071,
+      "step": 5354
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5085383472991699,
+      "learning_rate": 7.664468769470847e-07,
+      "loss": 0.3125,
+      "step": 5355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.48361497281116317,
+      "learning_rate": 7.663594827152805e-07,
+      "loss": 0.1925,
+      "step": 5356
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3529085530710126,
+      "learning_rate": 7.662720771200452e-07,
+      "loss": 0.2793,
+      "step": 5357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5916127837381652,
+      "learning_rate": 7.66184660165108e-07,
+      "loss": 0.1994,
+      "step": 5358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4505598947898049,
+      "learning_rate": 7.660972318541981e-07,
+      "loss": 0.1177,
+      "step": 5359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1316654222220077,
+      "learning_rate": 7.660097921910451e-07,
+      "loss": 0.002,
+      "step": 5360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4002509006801685,
+      "learning_rate": 7.659223411793799e-07,
+      "loss": 0.1185,
+      "step": 5361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.40040102302262925,
+      "learning_rate": 7.658348788229329e-07,
+      "loss": 0.2263,
+      "step": 5362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.19910629247316,
+      "learning_rate": 7.657474051254356e-07,
+      "loss": 0.3614,
+      "step": 5363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2013994840857332,
+      "learning_rate": 7.656599200906197e-07,
+      "loss": 0.1998,
+      "step": 5364
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6985400370745665,
+      "learning_rate": 7.655724237222177e-07,
+      "loss": 0.1717,
+      "step": 5365
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8095266896357952,
+      "learning_rate": 7.654849160239623e-07,
+      "loss": 0.216,
+      "step": 5366
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.264126168233839,
+      "learning_rate": 7.653973969995865e-07,
+      "loss": 0.195,
+      "step": 5367
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1730880943862239,
+      "learning_rate": 7.653098666528244e-07,
+      "loss": 0.2856,
+      "step": 5368
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.68317277219934,
+      "learning_rate": 7.652223249874098e-07,
+      "loss": 0.2053,
+      "step": 5369
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.44636271056922516,
+      "learning_rate": 7.651347720070777e-07,
+      "loss": 0.1098,
+      "step": 5370
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.35610538276850995,
+      "learning_rate": 7.650472077155634e-07,
+      "loss": 0.2497,
+      "step": 5371
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6266563931129285,
+      "learning_rate": 7.649596321166024e-07,
+      "loss": 0.3499,
+      "step": 5372
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16752537143108334,
+      "learning_rate": 7.648720452139308e-07,
+      "loss": 0.0047,
+      "step": 5373
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.633535854388574,
+      "learning_rate": 7.647844470112854e-07,
+      "loss": 0.2081,
+      "step": 5374
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6073753413259145,
+      "learning_rate": 7.646968375124032e-07,
+      "loss": 0.297,
+      "step": 5375
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.899192877962476,
+      "learning_rate": 7.646092167210216e-07,
+      "loss": 0.1677,
+      "step": 5376
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4810626287457248,
+      "learning_rate": 7.64521584640879e-07,
+      "loss": 0.2114,
+      "step": 5377
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.842217120586512,
+      "learning_rate": 7.644339412757138e-07,
+      "loss": 0.2486,
+      "step": 5378
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.236412666233146,
+      "learning_rate": 7.643462866292651e-07,
+      "loss": 0.1284,
+      "step": 5379
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.465679153010764,
+      "learning_rate": 7.642586207052726e-07,
+      "loss": 0.0819,
+      "step": 5380
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5932907918056111,
+      "learning_rate": 7.641709435074759e-07,
+      "loss": 0.025,
+      "step": 5381
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8587739825305312,
+      "learning_rate": 7.640832550396157e-07,
+      "loss": 0.1432,
+      "step": 5382
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7548112413610587,
+      "learning_rate": 7.639955553054331e-07,
+      "loss": 0.2997,
+      "step": 5383
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3740770436734397,
+      "learning_rate": 7.639078443086693e-07,
+      "loss": 0.0577,
+      "step": 5384
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6269091369820854,
+      "learning_rate": 7.638201220530663e-07,
+      "loss": 0.219,
+      "step": 5385
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.35717292554144603,
+      "learning_rate": 7.637323885423667e-07,
+      "loss": 0.1343,
+      "step": 5386
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3400871097740708,
+      "learning_rate": 7.63644643780313e-07,
+      "loss": 0.3339,
+      "step": 5387
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8314382332385768,
+      "learning_rate": 7.635568877706491e-07,
+      "loss": 0.1248,
+      "step": 5388
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7448008465431624,
+      "learning_rate": 7.634691205171185e-07,
+      "loss": 0.2172,
+      "step": 5389
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3429403669183726,
+      "learning_rate": 7.633813420234654e-07,
+      "loss": 0.1637,
+      "step": 5390
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5481036834158419,
+      "learning_rate": 7.632935522934349e-07,
+      "loss": 0.3444,
+      "step": 5391
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.26179972085160513,
+      "learning_rate": 7.632057513307721e-07,
+      "loss": 0.0127,
+      "step": 5392
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7934907645401681,
+      "learning_rate": 7.63117939139223e-07,
+      "loss": 0.3575,
+      "step": 5393
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4847344941338364,
+      "learning_rate": 7.630301157225335e-07,
+      "loss": 0.0863,
+      "step": 5394
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5144340151696197,
+      "learning_rate": 7.629422810844506e-07,
+      "loss": 0.058,
+      "step": 5395
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5197423783363038,
+      "learning_rate": 7.628544352287213e-07,
+      "loss": 0.1952,
+      "step": 5396
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9357355154274088,
+      "learning_rate": 7.627665781590936e-07,
+      "loss": 0.0964,
+      "step": 5397
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7982943321960051,
+      "learning_rate": 7.626787098793153e-07,
+      "loss": 0.2457,
+      "step": 5398
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7873596990257173,
+      "learning_rate": 7.625908303931352e-07,
+      "loss": 0.1931,
+      "step": 5399
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.03574043489678248,
+      "learning_rate": 7.625029397043024e-07,
+      "loss": 0.0002,
+      "step": 5400
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3735738747832986,
+      "learning_rate": 7.624150378165665e-07,
+      "loss": 0.3125,
+      "step": 5401
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6538545634079106,
+      "learning_rate": 7.623271247336776e-07,
+      "loss": 0.1629,
+      "step": 5402
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2481701307765698,
+      "learning_rate": 7.622392004593861e-07,
+      "loss": 0.2579,
+      "step": 5403
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1906190333031517,
+      "learning_rate": 7.621512649974434e-07,
+      "loss": 0.3404,
+      "step": 5404
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7951479589285008,
+      "learning_rate": 7.620633183516004e-07,
+      "loss": 0.2589,
+      "step": 5405
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6350964950267514,
+      "learning_rate": 7.619753605256096e-07,
+      "loss": 0.1353,
+      "step": 5406
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.14833055215320196,
+      "learning_rate": 7.618873915232233e-07,
+      "loss": 0.0045,
+      "step": 5407
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7273594822670096,
+      "learning_rate": 7.617994113481944e-07,
+      "loss": 0.2311,
+      "step": 5408
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3214943997490827,
+      "learning_rate": 7.617114200042764e-07,
+      "loss": 0.131,
+      "step": 5409
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7704313806013201,
+      "learning_rate": 7.61623417495223e-07,
+      "loss": 0.1477,
+      "step": 5410
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.48631190898778454,
+      "learning_rate": 7.615354038247887e-07,
+      "loss": 0.0818,
+      "step": 5411
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6814791525704665,
+      "learning_rate": 7.614473789967284e-07,
+      "loss": 0.3285,
+      "step": 5412
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5000201149916452,
+      "learning_rate": 7.613593430147973e-07,
+      "loss": 0.3307,
+      "step": 5413
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.086899312102441,
+      "learning_rate": 7.612712958827511e-07,
+      "loss": 0.2334,
+      "step": 5414
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.43445168035263737,
+      "learning_rate": 7.611832376043464e-07,
+      "loss": 0.1203,
+      "step": 5415
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5063496868375928,
+      "learning_rate": 7.610951681833397e-07,
+      "loss": 0.1298,
+      "step": 5416
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1891589968534975,
+      "learning_rate": 7.610070876234882e-07,
+      "loss": 0.1901,
+      "step": 5417
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5705667873301133,
+      "learning_rate": 7.609189959285497e-07,
+      "loss": 0.0862,
+      "step": 5418
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.165052832144716,
+      "learning_rate": 7.608308931022822e-07,
+      "loss": 0.2605,
+      "step": 5419
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0442779890428175,
+      "learning_rate": 7.607427791484447e-07,
+      "loss": 0.1809,
+      "step": 5420
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4496070838672782,
+      "learning_rate": 7.606546540707959e-07,
+      "loss": 0.1794,
+      "step": 5421
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6030649034819161,
+      "learning_rate": 7.605665178730956e-07,
+      "loss": 0.1341,
+      "step": 5422
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2806532213380455,
+      "learning_rate": 7.604783705591039e-07,
+      "loss": 0.1071,
+      "step": 5423
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.747730965098597,
+      "learning_rate": 7.603902121325811e-07,
+      "loss": 0.262,
+      "step": 5424
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26019060436072894,
+      "learning_rate": 7.603020425972886e-07,
+      "loss": 0.119,
+      "step": 5425
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6580863305263909,
+      "learning_rate": 7.602138619569876e-07,
+      "loss": 0.0574,
+      "step": 5426
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.586572500302497,
+      "learning_rate": 7.601256702154402e-07,
+      "loss": 0.1988,
+      "step": 5427
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9429699165236303,
+      "learning_rate": 7.600374673764087e-07,
+      "loss": 0.1048,
+      "step": 5428
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6030129263368172,
+      "learning_rate": 7.599492534436562e-07,
+      "loss": 0.1095,
+      "step": 5429
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3970639490193903,
+      "learning_rate": 7.598610284209459e-07,
+      "loss": 0.123,
+      "step": 5430
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8875642730833075,
+      "learning_rate": 7.597727923120419e-07,
+      "loss": 0.3333,
+      "step": 5431
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7935667196936881,
+      "learning_rate": 7.596845451207081e-07,
+      "loss": 0.1299,
+      "step": 5432
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8845982663534275,
+      "learning_rate": 7.595962868507098e-07,
+      "loss": 0.0206,
+      "step": 5433
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6958167365070135,
+      "learning_rate": 7.595080175058119e-07,
+      "loss": 0.4635,
+      "step": 5434
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8434032033922428,
+      "learning_rate": 7.594197370897806e-07,
+      "loss": 0.138,
+      "step": 5435
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2340305106813205,
+      "learning_rate": 7.593314456063815e-07,
+      "loss": 0.1236,
+      "step": 5436
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3510436566896165,
+      "learning_rate": 7.592431430593818e-07,
+      "loss": 0.2035,
+      "step": 5437
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.113828575830915,
+      "learning_rate": 7.591548294525482e-07,
+      "loss": 0.2367,
+      "step": 5438
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6365103004291444,
+      "learning_rate": 7.590665047896489e-07,
+      "loss": 0.0133,
+      "step": 5439
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5961683705584258,
+      "learning_rate": 7.589781690744515e-07,
+      "loss": 0.0773,
+      "step": 5440
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7132547038521998,
+      "learning_rate": 7.588898223107249e-07,
+      "loss": 0.1689,
+      "step": 5441
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4314675017187079,
+      "learning_rate": 7.588014645022381e-07,
+      "loss": 0.109,
+      "step": 5442
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.498069862488393,
+      "learning_rate": 7.587130956527605e-07,
+      "loss": 0.3384,
+      "step": 5443
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4025239974395892,
+      "learning_rate": 7.586247157660623e-07,
+      "loss": 0.0816,
+      "step": 5444
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.006154927034557,
+      "learning_rate": 7.585363248459138e-07,
+      "loss": 0.0924,
+      "step": 5445
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5946390055624198,
+      "learning_rate": 7.584479228960858e-07,
+      "loss": 0.4267,
+      "step": 5446
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1102941873243526,
+      "learning_rate": 7.583595099203499e-07,
+      "loss": 0.1461,
+      "step": 5447
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8735929599987753,
+      "learning_rate": 7.582710859224779e-07,
+      "loss": 0.1238,
+      "step": 5448
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5119734516248617,
+      "learning_rate": 7.581826509062422e-07,
+      "loss": 0.0727,
+      "step": 5449
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5092960882392609,
+      "learning_rate": 7.580942048754158e-07,
+      "loss": 0.218,
+      "step": 5450
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.48091907921323257,
+      "learning_rate": 7.580057478337716e-07,
+      "loss": 0.135,
+      "step": 5451
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6804820882221856,
+      "learning_rate": 7.579172797850835e-07,
+      "loss": 0.1032,
+      "step": 5452
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5761016630247517,
+      "learning_rate": 7.578288007331259e-07,
+      "loss": 0.2332,
+      "step": 5453
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.148788013242087,
+      "learning_rate": 7.577403106816733e-07,
+      "loss": 0.3384,
+      "step": 5454
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.577660597165397,
+      "learning_rate": 7.576518096345008e-07,
+      "loss": 0.2448,
+      "step": 5455
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1541021511978142,
+      "learning_rate": 7.575632975953844e-07,
+      "loss": 0.1887,
+      "step": 5456
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9556198088872199,
+      "learning_rate": 7.574747745680998e-07,
+      "loss": 0.4103,
+      "step": 5457
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.011746950565354,
+      "learning_rate": 7.573862405564238e-07,
+      "loss": 0.235,
+      "step": 5458
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.867669886562826,
+      "learning_rate": 7.572976955641333e-07,
+      "loss": 0.0455,
+      "step": 5459
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1562166733665182,
+      "learning_rate": 7.57209139595006e-07,
+      "loss": 0.2442,
+      "step": 5460
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9807165096759748,
+      "learning_rate": 7.571205726528196e-07,
+      "loss": 0.2519,
+      "step": 5461
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.19813535693980622,
+      "learning_rate": 7.570319947413528e-07,
+      "loss": 0.0902,
+      "step": 5462
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.33052386627680735,
+      "learning_rate": 7.569434058643843e-07,
+      "loss": 0.1445,
+      "step": 5463
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.195067167890204,
+      "learning_rate": 7.568548060256937e-07,
+      "loss": 0.1691,
+      "step": 5464
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7083560570882403,
+      "learning_rate": 7.567661952290607e-07,
+      "loss": 0.1655,
+      "step": 5465
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6414384746879106,
+      "learning_rate": 7.566775734782656e-07,
+      "loss": 0.2678,
+      "step": 5466
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.237016991388835,
+      "learning_rate": 7.565889407770891e-07,
+      "loss": 0.1284,
+      "step": 5467
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1814716207331748,
+      "learning_rate": 7.565002971293127e-07,
+      "loss": 0.3381,
+      "step": 5468
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.222609216444924,
+      "learning_rate": 7.564116425387181e-07,
+      "loss": 0.1326,
+      "step": 5469
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3834778682885506,
+      "learning_rate": 7.563229770090873e-07,
+      "loss": 0.4125,
+      "step": 5470
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8188635420752655,
+      "learning_rate": 7.562343005442031e-07,
+      "loss": 0.0042,
+      "step": 5471
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3317451274996792,
+      "learning_rate": 7.561456131478486e-07,
+      "loss": 0.1304,
+      "step": 5472
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7369737169901256,
+      "learning_rate": 7.56056914823807e-07,
+      "loss": 0.2119,
+      "step": 5473
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7048416421697092,
+      "learning_rate": 7.55968205575863e-07,
+      "loss": 0.2323,
+      "step": 5474
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8939613133381339,
+      "learning_rate": 7.558794854078006e-07,
+      "loss": 0.1165,
+      "step": 5475
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9367642014346794,
+      "learning_rate": 7.55790754323405e-07,
+      "loss": 0.2342,
+      "step": 5476
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8430705717483711,
+      "learning_rate": 7.557020123264615e-07,
+      "loss": 0.2196,
+      "step": 5477
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5027771057273157,
+      "learning_rate": 7.556132594207564e-07,
+      "loss": 0.0873,
+      "step": 5478
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.45055304957519526,
+      "learning_rate": 7.555244956100757e-07,
+      "loss": 0.1712,
+      "step": 5479
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5384406883428974,
+      "learning_rate": 7.554357208982063e-07,
+      "loss": 0.2173,
+      "step": 5480
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5123732613510469,
+      "learning_rate": 7.553469352889355e-07,
+      "loss": 0.2544,
+      "step": 5481
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6622802529235681,
+      "learning_rate": 7.552581387860513e-07,
+      "loss": 0.172,
+      "step": 5482
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.246609448365193,
+      "learning_rate": 7.551693313933416e-07,
+      "loss": 0.3406,
+      "step": 5483
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6176914522877467,
+      "learning_rate": 7.550805131145954e-07,
+      "loss": 0.2154,
+      "step": 5484
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5857405908457256,
+      "learning_rate": 7.549916839536017e-07,
+      "loss": 0.1204,
+      "step": 5485
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9430934698659614,
+      "learning_rate": 7.549028439141502e-07,
+      "loss": 0.1439,
+      "step": 5486
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5069939926317235,
+      "learning_rate": 7.548139930000308e-07,
+      "loss": 0.132,
+      "step": 5487
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9469093464280597,
+      "learning_rate": 7.547251312150344e-07,
+      "loss": 0.1164,
+      "step": 5488
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.38428685088550063,
+      "learning_rate": 7.546362585629517e-07,
+      "loss": 0.1245,
+      "step": 5489
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4850082210504356,
+      "learning_rate": 7.545473750475744e-07,
+      "loss": 0.0876,
+      "step": 5490
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.814698988719398,
+      "learning_rate": 7.544584806726944e-07,
+      "loss": 0.2527,
+      "step": 5491
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8193787101128907,
+      "learning_rate": 7.54369575442104e-07,
+      "loss": 0.0961,
+      "step": 5492
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6859605299798347,
+      "learning_rate": 7.542806593595961e-07,
+      "loss": 0.2168,
+      "step": 5493
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.46656504276993166,
+      "learning_rate": 7.541917324289644e-07,
+      "loss": 0.0417,
+      "step": 5494
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6236552184627319,
+      "learning_rate": 7.541027946540022e-07,
+      "loss": 0.1295,
+      "step": 5495
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8537973831497453,
+      "learning_rate": 7.540138460385039e-07,
+      "loss": 0.1608,
+      "step": 5496
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2793862954463233,
+      "learning_rate": 7.539248865862644e-07,
+      "loss": 0.1617,
+      "step": 5497
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6398392521187805,
+      "learning_rate": 7.538359163010789e-07,
+      "loss": 0.1793,
+      "step": 5498
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4287984261789411,
+      "learning_rate": 7.537469351867429e-07,
+      "loss": 0.0101,
+      "step": 5499
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6166844168596135,
+      "learning_rate": 7.536579432470525e-07,
+      "loss": 0.3597,
+      "step": 5500
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4182742071772365,
+      "learning_rate": 7.535689404858041e-07,
+      "loss": 0.2603,
+      "step": 5501
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3854342258052628,
+      "learning_rate": 7.534799269067951e-07,
+      "loss": 0.1638,
+      "step": 5502
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8940115750825707,
+      "learning_rate": 7.53390902513823e-07,
+      "loss": 0.2861,
+      "step": 5503
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.30640763425162,
+      "learning_rate": 7.533018673106855e-07,
+      "loss": 0.2263,
+      "step": 5504
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7424977747555532,
+      "learning_rate": 7.532128213011813e-07,
+      "loss": 0.0734,
+      "step": 5505
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7383830668521845,
+      "learning_rate": 7.531237644891089e-07,
+      "loss": 0.3017,
+      "step": 5506
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6425962515759647,
+      "learning_rate": 7.530346968782679e-07,
+      "loss": 0.1868,
+      "step": 5507
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8891034027526099,
+      "learning_rate": 7.529456184724582e-07,
+      "loss": 0.2093,
+      "step": 5508
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6308586312027443,
+      "learning_rate": 7.528565292754798e-07,
+      "loss": 0.2022,
+      "step": 5509
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2966416101669163,
+      "learning_rate": 7.527674292911337e-07,
+      "loss": 0.346,
+      "step": 5510
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.565257841276505,
+      "learning_rate": 7.526783185232207e-07,
+      "loss": 0.2415,
+      "step": 5511
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6378618019879686,
+      "learning_rate": 7.525891969755429e-07,
+      "loss": 0.2266,
+      "step": 5512
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7370565367478511,
+      "learning_rate": 7.525000646519022e-07,
+      "loss": 0.0854,
+      "step": 5513
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.13803143539431034,
+      "learning_rate": 7.52410921556101e-07,
+      "loss": 0.0438,
+      "step": 5514
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8618364879921652,
+      "learning_rate": 7.523217676919427e-07,
+      "loss": 0.3591,
+      "step": 5515
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3538662856414543,
+      "learning_rate": 7.522326030632303e-07,
+      "loss": 0.3528,
+      "step": 5516
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4175646798417754,
+      "learning_rate": 7.521434276737682e-07,
+      "loss": 0.2288,
+      "step": 5517
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7815260591885742,
+      "learning_rate": 7.520542415273605e-07,
+      "loss": 0.2603,
+      "step": 5518
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6129913870949957,
+      "learning_rate": 7.51965044627812e-07,
+      "loss": 0.079,
+      "step": 5519
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9206101002044337,
+      "learning_rate": 7.518758369789284e-07,
+      "loss": 0.0695,
+      "step": 5520
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9850240890306616,
+      "learning_rate": 7.517866185845152e-07,
+      "loss": 0.3234,
+      "step": 5521
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4658518357104681,
+      "learning_rate": 7.516973894483788e-07,
+      "loss": 0.3014,
+      "step": 5522
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5990441004132712,
+      "learning_rate": 7.516081495743258e-07,
+      "loss": 0.0746,
+      "step": 5523
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9125143811157321,
+      "learning_rate": 7.515188989661631e-07,
+      "loss": 0.098,
+      "step": 5524
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6742846516022356,
+      "learning_rate": 7.514296376276988e-07,
+      "loss": 0.1169,
+      "step": 5525
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9410200416897807,
+      "learning_rate": 7.513403655627407e-07,
+      "loss": 0.2208,
+      "step": 5526
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1831979924336038,
+      "learning_rate": 7.512510827750973e-07,
+      "loss": 0.336,
+      "step": 5527
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6164531950287837,
+      "learning_rate": 7.511617892685775e-07,
+      "loss": 0.2676,
+      "step": 5528
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0073977437619148,
+      "learning_rate": 7.51072485046991e-07,
+      "loss": 0.0981,
+      "step": 5529
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5920372762894701,
+      "learning_rate": 7.509831701141476e-07,
+      "loss": 0.2343,
+      "step": 5530
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5246433394510883,
+      "learning_rate": 7.508938444738575e-07,
+      "loss": 0.2047,
+      "step": 5531
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.41019274446527026,
+      "learning_rate": 7.508045081299317e-07,
+      "loss": 0.3106,
+      "step": 5532
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9708985114565645,
+      "learning_rate": 7.507151610861815e-07,
+      "loss": 0.2222,
+      "step": 5533
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4165937334867051,
+      "learning_rate": 7.506258033464183e-07,
+      "loss": 0.3681,
+      "step": 5534
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.665154380980617,
+      "learning_rate": 7.505364349144547e-07,
+      "loss": 0.0905,
+      "step": 5535
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3393190699864382,
+      "learning_rate": 7.504470557941032e-07,
+      "loss": 0.1616,
+      "step": 5536
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3690012066375732,
+      "learning_rate": 7.503576659891767e-07,
+      "loss": 0.3203,
+      "step": 5537
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8934533090948807,
+      "learning_rate": 7.502682655034889e-07,
+      "loss": 0.2346,
+      "step": 5538
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.39530251328778704,
+      "learning_rate": 7.501788543408538e-07,
+      "loss": 0.2292,
+      "step": 5539
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7845545391057316,
+      "learning_rate": 7.50089432505086e-07,
+      "loss": 0.4175,
+      "step": 5540
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6618787956586863,
+      "learning_rate": 7.5e-07,
+      "loss": 0.3286,
+      "step": 5541
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7129523944298292,
+      "learning_rate": 7.499105568294117e-07,
+      "loss": 0.3006,
+      "step": 5542
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.34117821356147626,
+      "learning_rate": 7.498211029971364e-07,
+      "loss": 0.0433,
+      "step": 5543
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.31330844388871015,
+      "learning_rate": 7.497316385069907e-07,
+      "loss": 0.0605,
+      "step": 5544
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.38548290144148534,
+      "learning_rate": 7.496421633627914e-07,
+      "loss": 0.2189,
+      "step": 5545
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0065995220738866,
+      "learning_rate": 7.495526775683555e-07,
+      "loss": 0.2171,
+      "step": 5546
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5711721107906526,
+      "learning_rate": 7.494631811275007e-07,
+      "loss": 0.1296,
+      "step": 5547
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3489269810369782,
+      "learning_rate": 7.493736740440451e-07,
+      "loss": 0.2914,
+      "step": 5548
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.951365099114088,
+      "learning_rate": 7.492841563218073e-07,
+      "loss": 0.3966,
+      "step": 5549
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8814859294325017,
+      "learning_rate": 7.491946279646063e-07,
+      "loss": 0.2021,
+      "step": 5550
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7517411318173365,
+      "learning_rate": 7.491050889762615e-07,
+      "loss": 0.0111,
+      "step": 5551
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.061138678911251,
+      "learning_rate": 7.490155393605928e-07,
+      "loss": 0.1306,
+      "step": 5552
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0569228540036513,
+      "learning_rate": 7.489259791214207e-07,
+      "loss": 0.1079,
+      "step": 5553
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9278994889987051,
+      "learning_rate": 7.488364082625658e-07,
+      "loss": 0.3206,
+      "step": 5554
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8308183209066643,
+      "learning_rate": 7.487468267878496e-07,
+      "loss": 0.3183,
+      "step": 5555
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.509597860789319,
+      "learning_rate": 7.486572347010936e-07,
+      "loss": 0.2289,
+      "step": 5556
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1048010166411033,
+      "learning_rate": 7.485676320061203e-07,
+      "loss": 0.214,
+      "step": 5557
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.39917487915009037,
+      "learning_rate": 7.48478018706752e-07,
+      "loss": 0.1675,
+      "step": 5558
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9550023711294141,
+      "learning_rate": 7.48388394806812e-07,
+      "loss": 0.2732,
+      "step": 5559
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.29584155540136137,
+      "learning_rate": 7.482987603101236e-07,
+      "loss": 0.0789,
+      "step": 5560
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6734566651477544,
+      "learning_rate": 7.482091152205111e-07,
+      "loss": 0.1139,
+      "step": 5561
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2411371701448759,
+      "learning_rate": 7.481194595417987e-07,
+      "loss": 0.0923,
+      "step": 5562
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5948113824946445,
+      "learning_rate": 7.480297932778115e-07,
+      "loss": 0.3879,
+      "step": 5563
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2976684576031656,
+      "learning_rate": 7.479401164323744e-07,
+      "loss": 0.0527,
+      "step": 5564
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6931557936213576,
+      "learning_rate": 7.478504290093137e-07,
+      "loss": 0.3137,
+      "step": 5565
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.93424600996979,
+      "learning_rate": 7.477607310124556e-07,
+      "loss": 0.0462,
+      "step": 5566
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7233646757544797,
+      "learning_rate": 7.476710224456267e-07,
+      "loss": 0.2318,
+      "step": 5567
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7154052090167797,
+      "learning_rate": 7.475813033126539e-07,
+      "loss": 0.2672,
+      "step": 5568
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.36258857569078284,
+      "learning_rate": 7.47491573617365e-07,
+      "loss": 0.153,
+      "step": 5569
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.030209076820536,
+      "learning_rate": 7.474018333635881e-07,
+      "loss": 0.2758,
+      "step": 5570
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9478448659022397,
+      "learning_rate": 7.473120825551516e-07,
+      "loss": 0.2027,
+      "step": 5571
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7993747337189498,
+      "learning_rate": 7.472223211958845e-07,
+      "loss": 0.0878,
+      "step": 5572
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.318048620247762,
+      "learning_rate": 7.471325492896163e-07,
+      "loss": 0.2963,
+      "step": 5573
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7229115735007899,
+      "learning_rate": 7.470427668401766e-07,
+      "loss": 0.1638,
+      "step": 5574
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8616759970056704,
+      "learning_rate": 7.469529738513959e-07,
+      "loss": 0.3767,
+      "step": 5575
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2840704025590366,
+      "learning_rate": 7.468631703271049e-07,
+      "loss": 0.1306,
+      "step": 5576
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.24844192009676036,
+      "learning_rate": 7.467733562711349e-07,
+      "loss": 0.0068,
+      "step": 5577
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7622158886912915,
+      "learning_rate": 7.466835316873173e-07,
+      "loss": 0.1614,
+      "step": 5578
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5074933879665838,
+      "learning_rate": 7.465936965794844e-07,
+      "loss": 0.0773,
+      "step": 5579
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2259056590904156,
+      "learning_rate": 7.465038509514687e-07,
+      "loss": 0.3512,
+      "step": 5580
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.346760402131009,
+      "learning_rate": 7.464139948071032e-07,
+      "loss": 0.1872,
+      "step": 5581
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4691340629334052,
+      "learning_rate": 7.463241281502213e-07,
+      "loss": 0.2107,
+      "step": 5582
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.627102704587824,
+      "learning_rate": 7.462342509846569e-07,
+      "loss": 0.1586,
+      "step": 5583
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2568205072347259,
+      "learning_rate": 7.461443633142445e-07,
+      "loss": 0.0881,
+      "step": 5584
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4491214798196706,
+      "learning_rate": 7.460544651428186e-07,
+      "loss": 0.3717,
+      "step": 5585
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5255583347139716,
+      "learning_rate": 7.459645564742147e-07,
+      "loss": 0.192,
+      "step": 5586
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7139963882199662,
+      "learning_rate": 7.458746373122682e-07,
+      "loss": 0.2344,
+      "step": 5587
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8754339092552733,
+      "learning_rate": 7.457847076608154e-07,
+      "loss": 0.2091,
+      "step": 5588
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.737726411806768,
+      "learning_rate": 7.456947675236931e-07,
+      "loss": 0.227,
+      "step": 5589
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.943797248690084,
+      "learning_rate": 7.45604816904738e-07,
+      "loss": 0.4343,
+      "step": 5590
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5659900317588396,
+      "learning_rate": 7.455148558077875e-07,
+      "loss": 0.107,
+      "step": 5591
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.007616680055804,
+      "learning_rate": 7.454248842366799e-07,
+      "loss": 0.1941,
+      "step": 5592
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.41591064194367167,
+      "learning_rate": 7.453349021952533e-07,
+      "loss": 0.0516,
+      "step": 5593
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3143605152013824,
+      "learning_rate": 7.452449096873467e-07,
+      "loss": 0.0915,
+      "step": 5594
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8047655118185578,
+      "learning_rate": 7.451549067167993e-07,
+      "loss": 0.179,
+      "step": 5595
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.744229908964181,
+      "learning_rate": 7.450648932874506e-07,
+      "loss": 0.0828,
+      "step": 5596
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6446813391299382,
+      "learning_rate": 7.449748694031411e-07,
+      "loss": 0.1499,
+      "step": 5597
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9662495016901049,
+      "learning_rate": 7.44884835067711e-07,
+      "loss": 0.3574,
+      "step": 5598
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7946516559915121,
+      "learning_rate": 7.447947902850015e-07,
+      "loss": 0.3775,
+      "step": 5599
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6918885825520503,
+      "learning_rate": 7.447047350588542e-07,
+      "loss": 0.2905,
+      "step": 5600
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9162217735464907,
+      "learning_rate": 7.44614669393111e-07,
+      "loss": 0.2766,
+      "step": 5601
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5856221977423952,
+      "learning_rate": 7.445245932916145e-07,
+      "loss": 0.1782,
+      "step": 5602
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.130971782366612,
+      "learning_rate": 7.44434506758207e-07,
+      "loss": 0.3936,
+      "step": 5603
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.49956457392869424,
+      "learning_rate": 7.443444097967322e-07,
+      "loss": 0.1684,
+      "step": 5604
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.442124914605853,
+      "learning_rate": 7.442543024110336e-07,
+      "loss": 0.1308,
+      "step": 5605
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1012306541307841,
+      "learning_rate": 7.441641846049556e-07,
+      "loss": 0.157,
+      "step": 5606
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8128936184046678,
+      "learning_rate": 7.440740563823424e-07,
+      "loss": 0.2805,
+      "step": 5607
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0958500784278856,
+      "learning_rate": 7.439839177470395e-07,
+      "loss": 0.0014,
+      "step": 5608
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0622095849821385,
+      "learning_rate": 7.438937687028922e-07,
+      "loss": 0.0322,
+      "step": 5609
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.82240783570984,
+      "learning_rate": 7.438036092537464e-07,
+      "loss": 0.1379,
+      "step": 5610
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0064401223486799,
+      "learning_rate": 7.437134394034486e-07,
+      "loss": 0.271,
+      "step": 5611
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5824506377365629,
+      "learning_rate": 7.436232591558453e-07,
+      "loss": 0.1038,
+      "step": 5612
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6278024259582801,
+      "learning_rate": 7.435330685147842e-07,
+      "loss": 0.3068,
+      "step": 5613
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.263689566913914,
+      "learning_rate": 7.434428674841129e-07,
+      "loss": 0.4099,
+      "step": 5614
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.536778118306459,
+      "learning_rate": 7.433526560676795e-07,
+      "loss": 0.2519,
+      "step": 5615
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.22056175117818586,
+      "learning_rate": 7.432624342693325e-07,
+      "loss": 0.0012,
+      "step": 5616
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0107420371992206,
+      "learning_rate": 7.431722020929209e-07,
+      "loss": 0.3657,
+      "step": 5617
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.38133028570314137,
+      "learning_rate": 7.430819595422944e-07,
+      "loss": 0.0995,
+      "step": 5618
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3274933073862944,
+      "learning_rate": 7.429917066213029e-07,
+      "loss": 0.1068,
+      "step": 5619
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7391526671153332,
+      "learning_rate": 7.429014433337968e-07,
+      "loss": 0.2051,
+      "step": 5620
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6382017648543542,
+      "learning_rate": 7.428111696836268e-07,
+      "loss": 0.0841,
+      "step": 5621
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5479588869795209,
+      "learning_rate": 7.427208856746443e-07,
+      "loss": 0.0117,
+      "step": 5622
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4681924472219252,
+      "learning_rate": 7.426305913107007e-07,
+      "loss": 0.2606,
+      "step": 5623
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4349478615381459,
+      "learning_rate": 7.425402865956484e-07,
+      "loss": 0.2065,
+      "step": 5624
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.48133390750990807,
+      "learning_rate": 7.424499715333398e-07,
+      "loss": 0.2311,
+      "step": 5625
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.43393289337501645,
+      "learning_rate": 7.42359646127628e-07,
+      "loss": 0.0131,
+      "step": 5626
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9901231319832863,
+      "learning_rate": 7.422693103823667e-07,
+      "loss": 0.3646,
+      "step": 5627
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8897654773571841,
+      "learning_rate": 7.421789643014095e-07,
+      "loss": 0.2595,
+      "step": 5628
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.08700536088908316,
+      "learning_rate": 7.420886078886109e-07,
+      "loss": 0.0035,
+      "step": 5629
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5091640443719467,
+      "learning_rate": 7.419982411478255e-07,
+      "loss": 0.2817,
+      "step": 5630
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6308944245010688,
+      "learning_rate": 7.419078640829087e-07,
+      "loss": 0.3434,
+      "step": 5631
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.325752180559273,
+      "learning_rate": 7.418174766977161e-07,
+      "loss": 0.0618,
+      "step": 5632
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7016784882125594,
+      "learning_rate": 7.417270789961039e-07,
+      "loss": 0.2734,
+      "step": 5633
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7981139381820834,
+      "learning_rate": 7.416366709819286e-07,
+      "loss": 0.2542,
+      "step": 5634
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.570908425410927,
+      "learning_rate": 7.415462526590471e-07,
+      "loss": 0.315,
+      "step": 5635
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6039784607978905,
+      "learning_rate": 7.414558240313169e-07,
+      "loss": 0.0198,
+      "step": 5636
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8918662710400138,
+      "learning_rate": 7.413653851025958e-07,
+      "loss": 0.0737,
+      "step": 5637
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6018072937725762,
+      "learning_rate": 7.412749358767422e-07,
+      "loss": 0.1989,
+      "step": 5638
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5927248964447291,
+      "learning_rate": 7.41184476357615e-07,
+      "loss": 0.1069,
+      "step": 5639
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8822401615052278,
+      "learning_rate": 7.410940065490731e-07,
+      "loss": 0.1038,
+      "step": 5640
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9121659841341959,
+      "learning_rate": 7.410035264549761e-07,
+      "loss": 0.1719,
+      "step": 5641
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2428128379739538,
+      "learning_rate": 7.409130360791842e-07,
+      "loss": 0.2085,
+      "step": 5642
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9114363804896441,
+      "learning_rate": 7.408225354255579e-07,
+      "loss": 0.2893,
+      "step": 5643
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.56843863463519,
+      "learning_rate": 7.407320244979581e-07,
+      "loss": 0.1195,
+      "step": 5644
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5664476519349504,
+      "learning_rate": 7.406415033002463e-07,
+      "loss": 0.1656,
+      "step": 5645
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8187165423109634,
+      "learning_rate": 7.405509718362841e-07,
+      "loss": 0.0665,
+      "step": 5646
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7366153668689416,
+      "learning_rate": 7.404604301099339e-07,
+      "loss": 0.2392,
+      "step": 5647
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8249155852385768,
+      "learning_rate": 7.403698781250586e-07,
+      "loss": 0.4827,
+      "step": 5648
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.18021192622080656,
+      "learning_rate": 7.402793158855209e-07,
+      "loss": 0.0045,
+      "step": 5649
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6927277860411692,
+      "learning_rate": 7.401887433951847e-07,
+      "loss": 0.1217,
+      "step": 5650
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3272280206798197,
+      "learning_rate": 7.400981606579138e-07,
+      "loss": 0.0934,
+      "step": 5651
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2544410262624939,
+      "learning_rate": 7.400075676775724e-07,
+      "loss": 0.1007,
+      "step": 5652
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2623609796207327,
+      "learning_rate": 7.39916964458026e-07,
+      "loss": 0.2208,
+      "step": 5653
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3304452078577227,
+      "learning_rate": 7.398263510031395e-07,
+      "loss": 0.2133,
+      "step": 5654
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6820556533519123,
+      "learning_rate": 7.397357273167788e-07,
+      "loss": 0.2927,
+      "step": 5655
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.988366638314936,
+      "learning_rate": 7.396450934028101e-07,
+      "loss": 0.2815,
+      "step": 5656
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.792737750103525,
+      "learning_rate": 7.395544492650999e-07,
+      "loss": 0.077,
+      "step": 5657
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1168716330631443,
+      "learning_rate": 7.394637949075154e-07,
+      "loss": 0.3105,
+      "step": 5658
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6715781478947498,
+      "learning_rate": 7.393731303339239e-07,
+      "loss": 0.2448,
+      "step": 5659
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4642124289382994,
+      "learning_rate": 7.392824555481935e-07,
+      "loss": 0.0388,
+      "step": 5660
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3997446258465748,
+      "learning_rate": 7.391917705541925e-07,
+      "loss": 0.1816,
+      "step": 5661
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.179714691245469,
+      "learning_rate": 7.391010753557898e-07,
+      "loss": 0.1698,
+      "step": 5662
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.5078712312561215,
+      "learning_rate": 7.390103699568546e-07,
+      "loss": 0.0489,
+      "step": 5663
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8097312463539856,
+      "learning_rate": 7.389196543612566e-07,
+      "loss": 0.2355,
+      "step": 5664
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.612219173660267,
+      "learning_rate": 7.388289285728657e-07,
+      "loss": 0.1088,
+      "step": 5665
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4675509341118895,
+      "learning_rate": 7.387381925955527e-07,
+      "loss": 0.1119,
+      "step": 5666
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.36521656834890354,
+      "learning_rate": 7.386474464331884e-07,
+      "loss": 0.0157,
+      "step": 5667
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6613329380880284,
+      "learning_rate": 7.385566900896444e-07,
+      "loss": 0.4098,
+      "step": 5668
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.308144804980323,
+      "learning_rate": 7.384659235687923e-07,
+      "loss": 0.1153,
+      "step": 5669
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2843589636687059,
+      "learning_rate": 7.383751468745045e-07,
+      "loss": 0.2318,
+      "step": 5670
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9875939890431915,
+      "learning_rate": 7.382843600106539e-07,
+      "loss": 0.3433,
+      "step": 5671
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4268691867817186,
+      "learning_rate": 7.381935629811133e-07,
+      "loss": 0.1946,
+      "step": 5672
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5107892003036639,
+      "learning_rate": 7.381027557897567e-07,
+      "loss": 0.0937,
+      "step": 5673
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.27515494830063,
+      "learning_rate": 7.380119384404578e-07,
+      "loss": 0.0747,
+      "step": 5674
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6790032271875316,
+      "learning_rate": 7.379211109370911e-07,
+      "loss": 0.2643,
+      "step": 5675
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26761219533808595,
+      "learning_rate": 7.378302732835316e-07,
+      "loss": 0.1076,
+      "step": 5676
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.500631776750934,
+      "learning_rate": 7.377394254836547e-07,
+      "loss": 0.1016,
+      "step": 5677
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5614007688444749,
+      "learning_rate": 7.376485675413356e-07,
+      "loss": 0.1787,
+      "step": 5678
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3439127880289776,
+      "learning_rate": 7.375576994604511e-07,
+      "loss": 0.0058,
+      "step": 5679
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.158611270169246,
+      "learning_rate": 7.374668212448776e-07,
+      "loss": 0.2079,
+      "step": 5680
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9345290844976936,
+      "learning_rate": 7.373759328984921e-07,
+      "loss": 0.371,
+      "step": 5681
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6145023831308265,
+      "learning_rate": 7.37285034425172e-07,
+      "loss": 0.1397,
+      "step": 5682
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2459527759727258,
+      "learning_rate": 7.371941258287955e-07,
+      "loss": 0.0973,
+      "step": 5683
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9089091616003668,
+      "learning_rate": 7.371032071132408e-07,
+      "loss": 0.2132,
+      "step": 5684
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1794201041196493,
+      "learning_rate": 7.370122782823866e-07,
+      "loss": 0.3221,
+      "step": 5685
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.886339245117712,
+      "learning_rate": 7.36921339340112e-07,
+      "loss": 0.0274,
+      "step": 5686
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6004655532216744,
+      "learning_rate": 7.368303902902969e-07,
+      "loss": 0.2572,
+      "step": 5687
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5554489179893731,
+      "learning_rate": 7.367394311368212e-07,
+      "loss": 0.1274,
+      "step": 5688
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8241326705184715,
+      "learning_rate": 7.366484618835656e-07,
+      "loss": 0.115,
+      "step": 5689
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.236697076427162,
+      "learning_rate": 7.36557482534411e-07,
+      "loss": 0.0819,
+      "step": 5690
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0167983250243307,
+      "learning_rate": 7.364664930932384e-07,
+      "loss": 0.0316,
+      "step": 5691
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6488208347208061,
+      "learning_rate": 7.3637549356393e-07,
+      "loss": 0.3033,
+      "step": 5692
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6161355625411944,
+      "learning_rate": 7.362844839503677e-07,
+      "loss": 0.2882,
+      "step": 5693
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2956359600938927,
+      "learning_rate": 7.361934642564345e-07,
+      "loss": 0.5045,
+      "step": 5694
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3618528380705309,
+      "learning_rate": 7.361024344860132e-07,
+      "loss": 0.332,
+      "step": 5695
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1060362685573237,
+      "learning_rate": 7.360113946429873e-07,
+      "loss": 0.1582,
+      "step": 5696
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5879530344047207,
+      "learning_rate": 7.35920344731241e-07,
+      "loss": 0.1334,
+      "step": 5697
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.851411422399082,
+      "learning_rate": 7.358292847546585e-07,
+      "loss": 0.1131,
+      "step": 5698
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5272105328609933,
+      "learning_rate": 7.357382147171247e-07,
+      "loss": 0.2389,
+      "step": 5699
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5566793042569607,
+      "learning_rate": 7.356471346225248e-07,
+      "loss": 0.0871,
+      "step": 5700
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5260969300371358,
+      "learning_rate": 7.355560444747444e-07,
+      "loss": 0.1068,
+      "step": 5701
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.43081534978532504,
+      "learning_rate": 7.354649442776696e-07,
+      "loss": 0.0796,
+      "step": 5702
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9839624499303066,
+      "learning_rate": 7.35373834035187e-07,
+      "loss": 0.1926,
+      "step": 5703
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4249842854795375,
+      "learning_rate": 7.352827137511835e-07,
+      "loss": 0.048,
+      "step": 5704
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4826094211081446,
+      "learning_rate": 7.351915834295462e-07,
+      "loss": 0.2404,
+      "step": 5705
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.46693183919675874,
+      "learning_rate": 7.351004430741633e-07,
+      "loss": 0.1472,
+      "step": 5706
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4642519926183546,
+      "learning_rate": 7.350092926889229e-07,
+      "loss": 0.2596,
+      "step": 5707
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3382804651693839,
+      "learning_rate": 7.349181322777137e-07,
+      "loss": 0.0568,
+      "step": 5708
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2676612799329265,
+      "learning_rate": 7.348269618444247e-07,
+      "loss": 0.0702,
+      "step": 5709
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9893205741285498,
+      "learning_rate": 7.347357813929454e-07,
+      "loss": 0.3175,
+      "step": 5710
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8570067933630025,
+      "learning_rate": 7.346445909271658e-07,
+      "loss": 0.2645,
+      "step": 5711
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9840025305888707,
+      "learning_rate": 7.345533904509763e-07,
+      "loss": 0.0598,
+      "step": 5712
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4118979936901152,
+      "learning_rate": 7.344621799682675e-07,
+      "loss": 0.1159,
+      "step": 5713
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.171307740778579,
+      "learning_rate": 7.343709594829311e-07,
+      "loss": 0.0209,
+      "step": 5714
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6712994625698699,
+      "learning_rate": 7.34279728998858e-07,
+      "loss": 0.1095,
+      "step": 5715
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0261338381513776,
+      "learning_rate": 7.34188488519941e-07,
+      "loss": 0.28,
+      "step": 5716
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4612067449269904,
+      "learning_rate": 7.340972380500722e-07,
+      "loss": 0.244,
+      "step": 5717
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6316223035935755,
+      "learning_rate": 7.340059775931447e-07,
+      "loss": 0.46,
+      "step": 5718
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3100217606040487,
+      "learning_rate": 7.339147071530518e-07,
+      "loss": 0.1125,
+      "step": 5719
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.087385815884791,
+      "learning_rate": 7.338234267336872e-07,
+      "loss": 0.26,
+      "step": 5720
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.678636488255624,
+      "learning_rate": 7.337321363389452e-07,
+      "loss": 0.2709,
+      "step": 5721
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.62897584509711,
+      "learning_rate": 7.336408359727203e-07,
+      "loss": 0.2842,
+      "step": 5722
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.18081937509925167,
+      "learning_rate": 7.335495256389077e-07,
+      "loss": 0.0797,
+      "step": 5723
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6151101410382191,
+      "learning_rate": 7.334582053414029e-07,
+      "loss": 0.1554,
+      "step": 5724
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6673374149633937,
+      "learning_rate": 7.333668750841016e-07,
+      "loss": 0.1558,
+      "step": 5725
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1063472558336795,
+      "learning_rate": 7.332755348709005e-07,
+      "loss": 0.1048,
+      "step": 5726
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.206564300336258,
+      "learning_rate": 7.331841847056961e-07,
+      "loss": 0.0903,
+      "step": 5727
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5047740906055621,
+      "learning_rate": 7.330928245923856e-07,
+      "loss": 0.0925,
+      "step": 5728
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3432459942569583,
+      "learning_rate": 7.330014545348665e-07,
+      "loss": 0.058,
+      "step": 5729
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5192291835658449,
+      "learning_rate": 7.329100745370371e-07,
+      "loss": 0.1255,
+      "step": 5730
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.37345549224087526,
+      "learning_rate": 7.328186846027958e-07,
+      "loss": 0.1841,
+      "step": 5731
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7453052128979122,
+      "learning_rate": 7.327272847360411e-07,
+      "loss": 0.3567,
+      "step": 5732
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0458302309578975,
+      "learning_rate": 7.326358749406729e-07,
+      "loss": 0.2815,
+      "step": 5733
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9210884382714356,
+      "learning_rate": 7.325444552205903e-07,
+      "loss": 0.1767,
+      "step": 5734
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4306498508361315,
+      "learning_rate": 7.32453025579694e-07,
+      "loss": 0.2736,
+      "step": 5735
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6874778263152205,
+      "learning_rate": 7.323615860218842e-07,
+      "loss": 0.271,
+      "step": 5736
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.233931078266332,
+      "learning_rate": 7.322701365510622e-07,
+      "loss": 0.3315,
+      "step": 5737
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8323186982992874,
+      "learning_rate": 7.321786771711291e-07,
+      "loss": 0.3319,
+      "step": 5738
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7889333812689606,
+      "learning_rate": 7.32087207885987e-07,
+      "loss": 0.3679,
+      "step": 5739
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7308759383840684,
+      "learning_rate": 7.31995728699538e-07,
+      "loss": 0.3541,
+      "step": 5740
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.686361723910737,
+      "learning_rate": 7.319042396156848e-07,
+      "loss": 0.2275,
+      "step": 5741
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5245958672103136,
+      "learning_rate": 7.318127406383307e-07,
+      "loss": 0.3172,
+      "step": 5742
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.226124069878523,
+      "learning_rate": 7.317212317713789e-07,
+      "loss": 0.1312,
+      "step": 5743
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.120338026666057,
+      "learning_rate": 7.316297130187336e-07,
+      "loss": 0.2483,
+      "step": 5744
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.689993816657338,
+      "learning_rate": 7.315381843842994e-07,
+      "loss": 0.2007,
+      "step": 5745
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.22364874179971245,
+      "learning_rate": 7.314466458719805e-07,
+      "loss": 0.0673,
+      "step": 5746
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3927315160954293,
+      "learning_rate": 7.313550974856824e-07,
+      "loss": 0.1532,
+      "step": 5747
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0515841969074804,
+      "learning_rate": 7.312635392293108e-07,
+      "loss": 0.3764,
+      "step": 5748
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3917197689690394,
+      "learning_rate": 7.311719711067716e-07,
+      "loss": 0.1184,
+      "step": 5749
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6463061072605664,
+      "learning_rate": 7.310803931219717e-07,
+      "loss": 0.0841,
+      "step": 5750
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9927125562299818,
+      "learning_rate": 7.309888052788174e-07,
+      "loss": 0.2835,
+      "step": 5751
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1886007504177757,
+      "learning_rate": 7.308972075812165e-07,
+      "loss": 0.0942,
+      "step": 5752
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9151289201691571,
+      "learning_rate": 7.308056000330766e-07,
+      "loss": 0.1421,
+      "step": 5753
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.900235705381781,
+      "learning_rate": 7.307139826383058e-07,
+      "loss": 0.1209,
+      "step": 5754
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5837265002658238,
+      "learning_rate": 7.306223554008126e-07,
+      "loss": 0.1662,
+      "step": 5755
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.064992461090195,
+      "learning_rate": 7.305307183245062e-07,
+      "loss": 0.0825,
+      "step": 5756
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.42029654362577756,
+      "learning_rate": 7.304390714132958e-07,
+      "loss": 0.1253,
+      "step": 5757
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.22787478007896503,
+      "learning_rate": 7.303474146710915e-07,
+      "loss": 0.0995,
+      "step": 5758
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4476001216086352,
+      "learning_rate": 7.302557481018034e-07,
+      "loss": 0.1563,
+      "step": 5759
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.20644091302951,
+      "learning_rate": 7.301640717093423e-07,
+      "loss": 0.0606,
+      "step": 5760
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.932890582717694,
+      "learning_rate": 7.30072385497619e-07,
+      "loss": 0.1974,
+      "step": 5761
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.775157928985653,
+      "learning_rate": 7.299806894705455e-07,
+      "loss": 0.1684,
+      "step": 5762
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5930366332677323,
+      "learning_rate": 7.298889836320334e-07,
+      "loss": 0.1152,
+      "step": 5763
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5754587056495242,
+      "learning_rate": 7.29797267985995e-07,
+      "loss": 0.1185,
+      "step": 5764
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4145190021337893,
+      "learning_rate": 7.297055425363432e-07,
+      "loss": 0.0865,
+      "step": 5765
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.459475753926773,
+      "learning_rate": 7.296138072869913e-07,
+      "loss": 0.1672,
+      "step": 5766
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.327574824334071,
+      "learning_rate": 7.295220622418527e-07,
+      "loss": 0.0507,
+      "step": 5767
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.345594717162949,
+      "learning_rate": 7.294303074048415e-07,
+      "loss": 0.0843,
+      "step": 5768
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0252372310202145,
+      "learning_rate": 7.293385427798721e-07,
+      "loss": 0.2572,
+      "step": 5769
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6216556517457695,
+      "learning_rate": 7.292467683708596e-07,
+      "loss": 0.2336,
+      "step": 5770
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6782792319158881,
+      "learning_rate": 7.291549841817192e-07,
+      "loss": 0.2633,
+      "step": 5771
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5688118833042158,
+      "learning_rate": 7.290631902163664e-07,
+      "loss": 0.0075,
+      "step": 5772
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3895485852110559,
+      "learning_rate": 7.289713864787175e-07,
+      "loss": 0.4064,
+      "step": 5773
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.506150204514353,
+      "learning_rate": 7.288795729726889e-07,
+      "loss": 0.0723,
+      "step": 5774
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0517219096754018,
+      "learning_rate": 7.287877497021977e-07,
+      "loss": 0.1546,
+      "step": 5775
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8877197643238348,
+      "learning_rate": 7.28695916671161e-07,
+      "loss": 0.0975,
+      "step": 5776
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6454200695293008,
+      "learning_rate": 7.286040738834968e-07,
+      "loss": 0.1935,
+      "step": 5777
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.715357209162237,
+      "learning_rate": 7.285122213431233e-07,
+      "loss": 0.0033,
+      "step": 5778
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4877099297085567,
+      "learning_rate": 7.28420359053959e-07,
+      "loss": 0.1364,
+      "step": 5779
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.888481172298206,
+      "learning_rate": 7.283284870199231e-07,
+      "loss": 0.231,
+      "step": 5780
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7195130961315708,
+      "learning_rate": 7.28236605244935e-07,
+      "loss": 0.301,
+      "step": 5781
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.998004839873859,
+      "learning_rate": 7.281447137329144e-07,
+      "loss": 0.2926,
+      "step": 5782
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0589799086919884,
+      "learning_rate": 7.280528124877817e-07,
+      "loss": 0.0335,
+      "step": 5783
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.246847115705105,
+      "learning_rate": 7.279609015134577e-07,
+      "loss": 0.2298,
+      "step": 5784
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5127667301057278,
+      "learning_rate": 7.278689808138632e-07,
+      "loss": 0.0162,
+      "step": 5785
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5686118131564267,
+      "learning_rate": 7.2777705039292e-07,
+      "loss": 0.3105,
+      "step": 5786
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5504845596655386,
+      "learning_rate": 7.276851102545499e-07,
+      "loss": 0.1512,
+      "step": 5787
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3507589777277134,
+      "learning_rate": 7.275931604026752e-07,
+      "loss": 0.2006,
+      "step": 5788
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6363302877756208,
+      "learning_rate": 7.275012008412191e-07,
+      "loss": 0.2754,
+      "step": 5789
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7886439160855335,
+      "learning_rate": 7.27409231574104e-07,
+      "loss": 0.0081,
+      "step": 5790
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9040161559453388,
+      "learning_rate": 7.273172526052542e-07,
+      "loss": 0.3219,
+      "step": 5791
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.417099807574797,
+      "learning_rate": 7.272252639385935e-07,
+      "loss": 0.3511,
+      "step": 5792
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.803623343303704,
+      "learning_rate": 7.27133265578046e-07,
+      "loss": 0.2845,
+      "step": 5793
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.176265351119403,
+      "learning_rate": 7.270412575275368e-07,
+      "loss": 0.1648,
+      "step": 5794
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.553487019276648,
+      "learning_rate": 7.269492397909913e-07,
+      "loss": 0.2341,
+      "step": 5795
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.20324779002863755,
+      "learning_rate": 7.268572123723351e-07,
+      "loss": 0.0031,
+      "step": 5796
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.21900193556265898,
+      "learning_rate": 7.267651752754939e-07,
+      "loss": 0.2145,
+      "step": 5797
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8876393913576334,
+      "learning_rate": 7.266731285043948e-07,
+      "loss": 0.2838,
+      "step": 5798
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.9612015987432265,
+      "learning_rate": 7.265810720629642e-07,
+      "loss": 0.0532,
+      "step": 5799
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6036271570510925,
+      "learning_rate": 7.264890059551295e-07,
+      "loss": 0.1828,
+      "step": 5800
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8418700313466142,
+      "learning_rate": 7.263969301848187e-07,
+      "loss": 0.2084,
+      "step": 5801
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3243730146921566,
+      "learning_rate": 7.263048447559596e-07,
+      "loss": 0.1082,
+      "step": 5802
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4295623478554113,
+      "learning_rate": 7.262127496724809e-07,
+      "loss": 0.1655,
+      "step": 5803
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.42034912858230894,
+      "learning_rate": 7.261206449383115e-07,
+      "loss": 0.1166,
+      "step": 5804
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.792203867453036,
+      "learning_rate": 7.260285305573809e-07,
+      "loss": 0.3044,
+      "step": 5805
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1834953034758922,
+      "learning_rate": 7.259364065336188e-07,
+      "loss": 0.2387,
+      "step": 5806
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1441268848969504,
+      "learning_rate": 7.258442728709554e-07,
+      "loss": 0.082,
+      "step": 5807
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.547858405168101,
+      "learning_rate": 7.257521295733213e-07,
+      "loss": 0.203,
+      "step": 5808
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6498259210423912,
+      "learning_rate": 7.256599766446476e-07,
+      "loss": 0.1696,
+      "step": 5809
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.589797651636558,
+      "learning_rate": 7.255678140888657e-07,
+      "loss": 0.203,
+      "step": 5810
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.295665441714581,
+      "learning_rate": 7.254756419099073e-07,
+      "loss": 0.0987,
+      "step": 5811
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1615737150670548,
+      "learning_rate": 7.253834601117048e-07,
+      "loss": 0.0033,
+      "step": 5812
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.44912207714477587,
+      "learning_rate": 7.252912686981907e-07,
+      "loss": 0.1508,
+      "step": 5813
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.963656407308307,
+      "learning_rate": 7.251990676732984e-07,
+      "loss": 0.137,
+      "step": 5814
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7939091592666753,
+      "learning_rate": 7.251068570409611e-07,
+      "loss": 0.1983,
+      "step": 5815
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.24188510757011006,
+      "learning_rate": 7.250146368051126e-07,
+      "loss": 0.0978,
+      "step": 5816
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0629619612234744,
+      "learning_rate": 7.249224069696876e-07,
+      "loss": 0.1746,
+      "step": 5817
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9571145455150913,
+      "learning_rate": 7.248301675386204e-07,
+      "loss": 0.18,
+      "step": 5818
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0775381212771094,
+      "learning_rate": 7.247379185158463e-07,
+      "loss": 0.1911,
+      "step": 5819
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4392557445096235,
+      "learning_rate": 7.246456599053008e-07,
+      "loss": 0.2503,
+      "step": 5820
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9725169780983117,
+      "learning_rate": 7.245533917109198e-07,
+      "loss": 0.3065,
+      "step": 5821
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.966288576702168,
+      "learning_rate": 7.244611139366398e-07,
+      "loss": 0.2274,
+      "step": 5822
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7998015479631919,
+      "learning_rate": 7.243688265863974e-07,
+      "loss": 0.318,
+      "step": 5823
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1673036168895607,
+      "learning_rate": 7.2427652966413e-07,
+      "loss": 0.2272,
+      "step": 5824
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2907306020411378,
+      "learning_rate": 7.241842231737748e-07,
+      "loss": 0.08,
+      "step": 5825
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9708427215621421,
+      "learning_rate": 7.2409190711927e-07,
+      "loss": 0.3598,
+      "step": 5826
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6238604424541833,
+      "learning_rate": 7.239995815045541e-07,
+      "loss": 0.5109,
+      "step": 5827
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3065175720140885,
+      "learning_rate": 7.239072463335657e-07,
+      "loss": 0.1389,
+      "step": 5828
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0029626689565476,
+      "learning_rate": 7.238149016102439e-07,
+      "loss": 0.3771,
+      "step": 5829
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.19797498975210448,
+      "learning_rate": 7.237225473385286e-07,
+      "loss": 0.0397,
+      "step": 5830
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.432444886934499,
+      "learning_rate": 7.236301835223597e-07,
+      "loss": 0.179,
+      "step": 5831
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7383596119254882,
+      "learning_rate": 7.235378101656775e-07,
+      "loss": 0.2657,
+      "step": 5832
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.9716360262510553,
+      "learning_rate": 7.234454272724231e-07,
+      "loss": 0.0608,
+      "step": 5833
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8742398751449905,
+      "learning_rate": 7.233530348465376e-07,
+      "loss": 0.3217,
+      "step": 5834
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.296015653419755,
+      "learning_rate": 7.232606328919626e-07,
+      "loss": 0.2541,
+      "step": 5835
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0472261813186758,
+      "learning_rate": 7.231682214126401e-07,
+      "loss": 0.2328,
+      "step": 5836
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7530299828248986,
+      "learning_rate": 7.230758004125127e-07,
+      "loss": 0.1346,
+      "step": 5837
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4179811070545931,
+      "learning_rate": 7.229833698955232e-07,
+      "loss": 0.0977,
+      "step": 5838
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5539109258430226,
+      "learning_rate": 7.228909298656149e-07,
+      "loss": 0.1684,
+      "step": 5839
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.685320802630804,
+      "learning_rate": 7.227984803267315e-07,
+      "loss": 0.1562,
+      "step": 5840
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.46337319748195305,
+      "learning_rate": 7.22706021282817e-07,
+      "loss": 0.1288,
+      "step": 5841
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.22438321986461,
+      "learning_rate": 7.226135527378161e-07,
+      "loss": 0.0241,
+      "step": 5842
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6880371677332253,
+      "learning_rate": 7.225210746956733e-07,
+      "loss": 0.2242,
+      "step": 5843
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3904455356259318,
+      "learning_rate": 7.224285871603342e-07,
+      "loss": 0.1122,
+      "step": 5844
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.80368536327891,
+      "learning_rate": 7.223360901357445e-07,
+      "loss": 0.0082,
+      "step": 5845
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.12360576823331884,
+      "learning_rate": 7.222435836258503e-07,
+      "loss": 0.0081,
+      "step": 5846
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.072400722055551,
+      "learning_rate": 7.221510676345979e-07,
+      "loss": 0.482,
+      "step": 5847
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7497709231459546,
+      "learning_rate": 7.220585421659344e-07,
+      "loss": 0.2764,
+      "step": 5848
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1114486730556554,
+      "learning_rate": 7.21966007223807e-07,
+      "loss": 0.1383,
+      "step": 5849
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9980044127105586,
+      "learning_rate": 7.218734628121638e-07,
+      "loss": 0.1708,
+      "step": 5850
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0969380583317159,
+      "learning_rate": 7.217809089349524e-07,
+      "loss": 0.2902,
+      "step": 5851
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6849231119905266,
+      "learning_rate": 7.216883455961218e-07,
+      "loss": 0.0865,
+      "step": 5852
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.243019603258892,
+      "learning_rate": 7.215957727996207e-07,
+      "loss": 0.1,
+      "step": 5853
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.385221654833512,
+      "learning_rate": 7.215031905493983e-07,
+      "loss": 0.1145,
+      "step": 5854
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5605233878714039,
+      "learning_rate": 7.214105988494045e-07,
+      "loss": 0.1167,
+      "step": 5855
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6201025446143363,
+      "learning_rate": 7.213179977035897e-07,
+      "loss": 0.0043,
+      "step": 5856
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6588082719258256,
+      "learning_rate": 7.212253871159041e-07,
+      "loss": 0.3062,
+      "step": 5857
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.448959617441987,
+      "learning_rate": 7.211327670902988e-07,
+      "loss": 0.2483,
+      "step": 5858
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7645063728979504,
+      "learning_rate": 7.210401376307252e-07,
+      "loss": 0.3142,
+      "step": 5859
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3454183970448466,
+      "learning_rate": 7.209474987411346e-07,
+      "loss": 0.0788,
+      "step": 5860
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.489328950945489,
+      "learning_rate": 7.208548504254799e-07,
+      "loss": 0.0732,
+      "step": 5861
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.0839647157281016,
+      "learning_rate": 7.207621926877133e-07,
+      "loss": 0.2387,
+      "step": 5862
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6501698188666848,
+      "learning_rate": 7.206695255317876e-07,
+      "loss": 0.456,
+      "step": 5863
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7720724698021013,
+      "learning_rate": 7.205768489616565e-07,
+      "loss": 0.3358,
+      "step": 5864
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7459872200113052,
+      "learning_rate": 7.204841629812734e-07,
+      "loss": 0.2244,
+      "step": 5865
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6919461914751762,
+      "learning_rate": 7.203914675945928e-07,
+      "loss": 0.2421,
+      "step": 5866
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.528182902739734,
+      "learning_rate": 7.202987628055693e-07,
+      "loss": 0.2035,
+      "step": 5867
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.46861676408355124,
+      "learning_rate": 7.202060486181575e-07,
+      "loss": 0.0723,
+      "step": 5868
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.879367632376332,
+      "learning_rate": 7.201133250363132e-07,
+      "loss": 0.1566,
+      "step": 5869
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38307353665079696,
+      "learning_rate": 7.200205920639918e-07,
+      "loss": 0.0125,
+      "step": 5870
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.473801322198328,
+      "learning_rate": 7.199278497051497e-07,
+      "loss": 0.3076,
+      "step": 5871
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.39376928629463165,
+      "learning_rate": 7.198350979637434e-07,
+      "loss": 0.1229,
+      "step": 5872
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.502550421033535,
+      "learning_rate": 7.197423368437299e-07,
+      "loss": 0.3354,
+      "step": 5873
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5897333491847417,
+      "learning_rate": 7.196495663490665e-07,
+      "loss": 0.1523,
+      "step": 5874
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.159565031612828,
+      "learning_rate": 7.195567864837111e-07,
+      "loss": 0.2187,
+      "step": 5875
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.502572097228924,
+      "learning_rate": 7.194639972516218e-07,
+      "loss": 0.2899,
+      "step": 5876
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5176055018510678,
+      "learning_rate": 7.193711986567573e-07,
+      "loss": 0.0791,
+      "step": 5877
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3059301873472071,
+      "learning_rate": 7.192783907030765e-07,
+      "loss": 0.1164,
+      "step": 5878
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7189273682628163,
+      "learning_rate": 7.191855733945386e-07,
+      "loss": 0.1507,
+      "step": 5879
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6557773111881142,
+      "learning_rate": 7.190927467351037e-07,
+      "loss": 0.3814,
+      "step": 5880
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1147535581280337,
+      "learning_rate": 7.189999107287317e-07,
+      "loss": 0.1315,
+      "step": 5881
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9928493129046877,
+      "learning_rate": 7.189070653793833e-07,
+      "loss": 0.0101,
+      "step": 5882
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.37211700710997386,
+      "learning_rate": 7.188142106910193e-07,
+      "loss": 0.1829,
+      "step": 5883
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.165946089905001,
+      "learning_rate": 7.187213466676013e-07,
+      "loss": 0.3081,
+      "step": 5884
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4124492140865974,
+      "learning_rate": 7.18628473313091e-07,
+      "loss": 0.0811,
+      "step": 5885
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8954038009244101,
+      "learning_rate": 7.185355906314505e-07,
+      "loss": 0.1387,
+      "step": 5886
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3092639590667032,
+      "learning_rate": 7.184426986266423e-07,
+      "loss": 0.2749,
+      "step": 5887
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6949233311868814,
+      "learning_rate": 7.183497973026296e-07,
+      "loss": 0.0865,
+      "step": 5888
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2710940156925137,
+      "learning_rate": 7.182568866633756e-07,
+      "loss": 0.1948,
+      "step": 5889
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1978573352474637,
+      "learning_rate": 7.18163966712844e-07,
+      "loss": 0.2208,
+      "step": 5890
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6818154892923289,
+      "learning_rate": 7.18071037454999e-07,
+      "loss": 0.2935,
+      "step": 5891
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4738591744705973,
+      "learning_rate": 7.179780988938051e-07,
+      "loss": 0.1539,
+      "step": 5892
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1680365976824745,
+      "learning_rate": 7.178851510332274e-07,
+      "loss": 0.2514,
+      "step": 5893
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5345444115984965,
+      "learning_rate": 7.177921938772311e-07,
+      "loss": 0.3865,
+      "step": 5894
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4549782245854918,
+      "learning_rate": 7.17699227429782e-07,
+      "loss": 0.2943,
+      "step": 5895
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5654763469609636,
+      "learning_rate": 7.176062516948463e-07,
+      "loss": 0.0977,
+      "step": 5896
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.46045765447375575,
+      "learning_rate": 7.175132666763905e-07,
+      "loss": 0.1541,
+      "step": 5897
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5610590384086495,
+      "learning_rate": 7.174202723783814e-07,
+      "loss": 0.0925,
+      "step": 5898
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2928975090718757,
+      "learning_rate": 7.173272688047865e-07,
+      "loss": 0.121,
+      "step": 5899
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9139770998882433,
+      "learning_rate": 7.172342559595732e-07,
+      "loss": 0.066,
+      "step": 5900
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.485855768140534,
+      "learning_rate": 7.1714123384671e-07,
+      "loss": 0.1278,
+      "step": 5901
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.444138561462059,
+      "learning_rate": 7.170482024701651e-07,
+      "loss": 0.0571,
+      "step": 5902
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6205883371938983,
+      "learning_rate": 7.169551618339078e-07,
+      "loss": 0.243,
+      "step": 5903
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6260682375338533,
+      "learning_rate": 7.168621119419072e-07,
+      "loss": 0.1723,
+      "step": 5904
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1867454917890057,
+      "learning_rate": 7.167690527981327e-07,
+      "loss": 0.3286,
+      "step": 5905
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.407792610972802,
+      "learning_rate": 7.166759844065548e-07,
+      "loss": 0.1464,
+      "step": 5906
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7570667036092302,
+      "learning_rate": 7.165829067711439e-07,
+      "loss": 0.1031,
+      "step": 5907
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.360062034053246,
+      "learning_rate": 7.164898198958706e-07,
+      "loss": 0.1911,
+      "step": 5908
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6199722405916185,
+      "learning_rate": 7.163967237847066e-07,
+      "loss": 0.159,
+      "step": 5909
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3338296173886413,
+      "learning_rate": 7.163036184416231e-07,
+      "loss": 0.366,
+      "step": 5910
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.213067992527338,
+      "learning_rate": 7.162105038705926e-07,
+      "loss": 0.1768,
+      "step": 5911
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.480109946037491,
+      "learning_rate": 7.161173800755874e-07,
+      "loss": 0.1274,
+      "step": 5912
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6505464364524993,
+      "learning_rate": 7.160242470605803e-07,
+      "loss": 0.0872,
+      "step": 5913
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6077104443243688,
+      "learning_rate": 7.159311048295444e-07,
+      "loss": 0.2574,
+      "step": 5914
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7781919438857764,
+      "learning_rate": 7.158379533864537e-07,
+      "loss": 0.167,
+      "step": 5915
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0321930163705826,
+      "learning_rate": 7.15744792735282e-07,
+      "loss": 0.1722,
+      "step": 5916
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5674156331427114,
+      "learning_rate": 7.156516228800035e-07,
+      "loss": 0.1605,
+      "step": 5917
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7950157651962791,
+      "learning_rate": 7.155584438245935e-07,
+      "loss": 0.1945,
+      "step": 5918
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.091492675135481,
+      "learning_rate": 7.154652555730267e-07,
+      "loss": 0.0534,
+      "step": 5919
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6350082291984965,
+      "learning_rate": 7.153720581292793e-07,
+      "loss": 0.2207,
+      "step": 5920
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6751077283045518,
+      "learning_rate": 7.152788514973267e-07,
+      "loss": 0.1156,
+      "step": 5921
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7838490625511402,
+      "learning_rate": 7.151856356811456e-07,
+      "loss": 0.2122,
+      "step": 5922
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6838821417234526,
+      "learning_rate": 7.150924106847127e-07,
+      "loss": 0.2095,
+      "step": 5923
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6116515680506596,
+      "learning_rate": 7.149991765120054e-07,
+      "loss": 0.4388,
+      "step": 5924
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0175319634749764,
+      "learning_rate": 7.149059331670008e-07,
+      "loss": 0.1162,
+      "step": 5925
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4889298794243602,
+      "learning_rate": 7.14812680653677e-07,
+      "loss": 0.0855,
+      "step": 5926
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1329174315251007,
+      "learning_rate": 7.147194189760124e-07,
+      "loss": 0.2025,
+      "step": 5927
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.38708904205903855,
+      "learning_rate": 7.146261481379858e-07,
+      "loss": 0.0041,
+      "step": 5928
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4835130500903364,
+      "learning_rate": 7.145328681435764e-07,
+      "loss": 0.1677,
+      "step": 5929
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7288205730665984,
+      "learning_rate": 7.144395789967635e-07,
+      "loss": 0.3148,
+      "step": 5930
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9432187522235403,
+      "learning_rate": 7.14346280701527e-07,
+      "loss": 0.0094,
+      "step": 5931
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0728331646521974,
+      "learning_rate": 7.142529732618474e-07,
+      "loss": 0.2973,
+      "step": 5932
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.564085645081182,
+      "learning_rate": 7.141596566817052e-07,
+      "loss": 0.365,
+      "step": 5933
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.912222774090835,
+      "learning_rate": 7.140663309650816e-07,
+      "loss": 0.326,
+      "step": 5934
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7604164267103952,
+      "learning_rate": 7.13972996115958e-07,
+      "loss": 0.2514,
+      "step": 5935
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.436229042738186,
+      "learning_rate": 7.138796521383162e-07,
+      "loss": 0.112,
+      "step": 5936
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.998368663674957,
+      "learning_rate": 7.137862990361382e-07,
+      "loss": 0.0441,
+      "step": 5937
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4927003810543219,
+      "learning_rate": 7.136929368134074e-07,
+      "loss": 0.0082,
+      "step": 5938
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7354966554476302,
+      "learning_rate": 7.135995654741062e-07,
+      "loss": 0.4271,
+      "step": 5939
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.74617529396601,
+      "learning_rate": 7.135061850222179e-07,
+      "loss": 0.1556,
+      "step": 5940
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.435093383695401,
+      "learning_rate": 7.134127954617268e-07,
+      "loss": 0.178,
+      "step": 5941
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.477204625878189,
+      "learning_rate": 7.133193967966168e-07,
+      "loss": 0.2543,
+      "step": 5942
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.43863566913845464,
+      "learning_rate": 7.132259890308725e-07,
+      "loss": 0.183,
+      "step": 5943
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.44991637776734,
+      "learning_rate": 7.131325721684788e-07,
+      "loss": 0.378,
+      "step": 5944
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5281526357455267,
+      "learning_rate": 7.130391462134211e-07,
+      "loss": 0.2169,
+      "step": 5945
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8521604255758248,
+      "learning_rate": 7.129457111696852e-07,
+      "loss": 0.174,
+      "step": 5946
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.68207597659719,
+      "learning_rate": 7.128522670412571e-07,
+      "loss": 0.0522,
+      "step": 5947
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2533660514279892,
+      "learning_rate": 7.127588138321235e-07,
+      "loss": 0.0686,
+      "step": 5948
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2090779232570901,
+      "learning_rate": 7.126653515462713e-07,
+      "loss": 0.0796,
+      "step": 5949
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2050745694824858,
+      "learning_rate": 7.125718801876876e-07,
+      "loss": 0.2794,
+      "step": 5950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6509178274863041,
+      "learning_rate": 7.1247839976036e-07,
+      "loss": 0.0936,
+      "step": 5951
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8250706157425587,
+      "learning_rate": 7.123849102682771e-07,
+      "loss": 0.265,
+      "step": 5952
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8781829729621038,
+      "learning_rate": 7.122914117154267e-07,
+      "loss": 0.3591,
+      "step": 5953
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.066038436900582,
+      "learning_rate": 7.12197904105798e-07,
+      "loss": 0.1235,
+      "step": 5954
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3810299623815854,
+      "learning_rate": 7.121043874433801e-07,
+      "loss": 0.1342,
+      "step": 5955
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1710195322965773,
+      "learning_rate": 7.120108617321627e-07,
+      "loss": 0.1984,
+      "step": 5956
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.49842423271635744,
+      "learning_rate": 7.119173269761357e-07,
+      "loss": 0.2331,
+      "step": 5957
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.023827614331166,
+      "learning_rate": 7.118237831792895e-07,
+      "loss": 0.3535,
+      "step": 5958
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25458583003373897,
+      "learning_rate": 7.117302303456149e-07,
+      "loss": 0.0158,
+      "step": 5959
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6364437863317047,
+      "learning_rate": 7.116366684791032e-07,
+      "loss": 0.0611,
+      "step": 5960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.823260968022451,
+      "learning_rate": 7.115430975837456e-07,
+      "loss": 0.1022,
+      "step": 5961
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.751828298169778,
+      "learning_rate": 7.114495176635343e-07,
+      "loss": 0.0226,
+      "step": 5962
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.630816834039371,
+      "learning_rate": 7.113559287224614e-07,
+      "loss": 0.1887,
+      "step": 5963
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.45860696416686153,
+      "learning_rate": 7.112623307645198e-07,
+      "loss": 0.2074,
+      "step": 5964
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6064617076623829,
+      "learning_rate": 7.111687237937024e-07,
+      "loss": 0.2096,
+      "step": 5965
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9400640168041525,
+      "learning_rate": 7.11075107814003e-07,
+      "loss": 0.276,
+      "step": 5966
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9060897632718069,
+      "learning_rate": 7.10981482829415e-07,
+      "loss": 0.0865,
+      "step": 5967
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7413702713438625,
+      "learning_rate": 7.108878488439327e-07,
+      "loss": 0.2822,
+      "step": 5968
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7268002955728479,
+      "learning_rate": 7.10794205861551e-07,
+      "loss": 0.1332,
+      "step": 5969
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5774531281108808,
+      "learning_rate": 7.107005538862646e-07,
+      "loss": 0.0733,
+      "step": 5970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.129077254375828,
+      "learning_rate": 7.10606892922069e-07,
+      "loss": 0.0939,
+      "step": 5971
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.471993788958997,
+      "learning_rate": 7.1051322297296e-07,
+      "loss": 0.2458,
+      "step": 5972
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.21795629225975438,
+      "learning_rate": 7.104195440429338e-07,
+      "loss": 0.0039,
+      "step": 5973
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0986424896095943,
+      "learning_rate": 7.103258561359868e-07,
+      "loss": 0.3232,
+      "step": 5974
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5780048420019451,
+      "learning_rate": 7.102321592561161e-07,
+      "loss": 0.2073,
+      "step": 5975
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2990008380893296,
+      "learning_rate": 7.101384534073186e-07,
+      "loss": 0.1052,
+      "step": 5976
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3446679506575092,
+      "learning_rate": 7.100447385935924e-07,
+      "loss": 0.3446,
+      "step": 5977
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7417729866746887,
+      "learning_rate": 7.099510148189353e-07,
+      "loss": 0.3005,
+      "step": 5978
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.380731448893932,
+      "learning_rate": 7.098572820873461e-07,
+      "loss": 0.0407,
+      "step": 5979
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6480951396834793,
+      "learning_rate": 7.097635404028233e-07,
+      "loss": 0.2887,
+      "step": 5980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.659078677383878,
+      "learning_rate": 7.096697897693661e-07,
+      "loss": 0.3541,
+      "step": 5981
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1218545597779377,
+      "learning_rate": 7.095760301909742e-07,
+      "loss": 0.3356,
+      "step": 5982
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9847545022936296,
+      "learning_rate": 7.094822616716476e-07,
+      "loss": 0.1656,
+      "step": 5983
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1206415516885497,
+      "learning_rate": 7.093884842153866e-07,
+      "loss": 0.5928,
+      "step": 5984
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.9162260514925755,
+      "learning_rate": 7.092946978261918e-07,
+      "loss": 0.1928,
+      "step": 5985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7143376489897715,
+      "learning_rate": 7.092009025080647e-07,
+      "loss": 0.1116,
+      "step": 5986
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5437738671263492,
+      "learning_rate": 7.091070982650063e-07,
+      "loss": 0.2653,
+      "step": 5987
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2406481074654894,
+      "learning_rate": 7.090132851010189e-07,
+      "loss": 0.3299,
+      "step": 5988
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9467414831579638,
+      "learning_rate": 7.089194630201045e-07,
+      "loss": 0.2875,
+      "step": 5989
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9279443760306558,
+      "learning_rate": 7.088256320262658e-07,
+      "loss": 0.0794,
+      "step": 5990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5185733427488309,
+      "learning_rate": 7.087317921235059e-07,
+      "loss": 0.1306,
+      "step": 5991
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7948092221675529,
+      "learning_rate": 7.086379433158282e-07,
+      "loss": 0.0462,
+      "step": 5992
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6098031442702017,
+      "learning_rate": 7.085440856072364e-07,
+      "loss": 0.1068,
+      "step": 5993
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.106216605920332,
+      "learning_rate": 7.084502190017346e-07,
+      "loss": 0.0095,
+      "step": 5994
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2519067909669138,
+      "learning_rate": 7.083563435033275e-07,
+      "loss": 0.1246,
+      "step": 5995
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1356303064312276,
+      "learning_rate": 7.0826245911602e-07,
+      "loss": 0.2158,
+      "step": 5996
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5893033155310382,
+      "learning_rate": 7.081685658438172e-07,
+      "loss": 0.236,
+      "step": 5997
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5284526499896604,
+      "learning_rate": 7.080746636907249e-07,
+      "loss": 0.1444,
+      "step": 5998
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.62541020470911,
+      "learning_rate": 7.079807526607492e-07,
+      "loss": 0.1926,
+      "step": 5999
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6633854058116887,
+      "learning_rate": 7.078868327578965e-07,
+      "loss": 0.3817,
+      "step": 6000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5390019204594863,
+      "learning_rate": 7.077929039861737e-07,
+      "loss": 0.1851,
+      "step": 6001
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1361359268069435,
+      "learning_rate": 7.076989663495877e-07,
+      "loss": 0.1746,
+      "step": 6002
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.10970779392084333,
+      "learning_rate": 7.076050198521464e-07,
+      "loss": 0.0664,
+      "step": 6003
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0860967601909939,
+      "learning_rate": 7.075110644978577e-07,
+      "loss": 0.4086,
+      "step": 6004
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5685774998213641,
+      "learning_rate": 7.074171002907296e-07,
+      "loss": 0.3996,
+      "step": 6005
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6201711996331208,
+      "learning_rate": 7.073231272347713e-07,
+      "loss": 0.3049,
+      "step": 6006
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6000797086416605,
+      "learning_rate": 7.072291453339915e-07,
+      "loss": 0.3039,
+      "step": 6007
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.45780747046117626,
+      "learning_rate": 7.071351545923998e-07,
+      "loss": 0.2359,
+      "step": 6008
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3257811079413989,
+      "learning_rate": 7.07041155014006e-07,
+      "loss": 0.1005,
+      "step": 6009
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0872698006568964,
+      "learning_rate": 7.069471466028203e-07,
+      "loss": 0.1744,
+      "step": 6010
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.280298811781577,
+      "learning_rate": 7.068531293628533e-07,
+      "loss": 0.0576,
+      "step": 6011
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0526762193027266,
+      "learning_rate": 7.06759103298116e-07,
+      "loss": 0.1753,
+      "step": 6012
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.288961530840641,
+      "learning_rate": 7.066650684126198e-07,
+      "loss": 0.3042,
+      "step": 6013
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.37518201346987406,
+      "learning_rate": 7.065710247103762e-07,
+      "loss": 0.1116,
+      "step": 6014
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2338419063587356,
+      "learning_rate": 7.064769721953975e-07,
+      "loss": 0.0568,
+      "step": 6015
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3789297878621667,
+      "learning_rate": 7.06382910871696e-07,
+      "loss": 0.1485,
+      "step": 6016
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.220911917100636,
+      "learning_rate": 7.062888407432847e-07,
+      "loss": 0.145,
+      "step": 6017
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.44760179866017075,
+      "learning_rate": 7.061947618141768e-07,
+      "loss": 0.2021,
+      "step": 6018
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3631122735822044,
+      "learning_rate": 7.061006740883858e-07,
+      "loss": 0.0854,
+      "step": 6019
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6197074357619311,
+      "learning_rate": 7.060065775699257e-07,
+      "loss": 0.1229,
+      "step": 6020
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.303806414473023,
+      "learning_rate": 7.059124722628112e-07,
+      "loss": 0.1706,
+      "step": 6021
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5110289241404611,
+      "learning_rate": 7.058183581710564e-07,
+      "loss": 0.0506,
+      "step": 6022
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2294421418777262,
+      "learning_rate": 7.057242352986767e-07,
+      "loss": 0.0219,
+      "step": 6023
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.587852148027603,
+      "learning_rate": 7.056301036496874e-07,
+      "loss": 0.1163,
+      "step": 6024
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6857314604523593,
+      "learning_rate": 7.055359632281048e-07,
+      "loss": 0.5359,
+      "step": 6025
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.923768138429377,
+      "learning_rate": 7.054418140379448e-07,
+      "loss": 0.2454,
+      "step": 6026
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.47052950296915724,
+      "learning_rate": 7.053476560832239e-07,
+      "loss": 0.1317,
+      "step": 6027
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4121696819836364,
+      "learning_rate": 7.052534893679593e-07,
+      "loss": 0.2421,
+      "step": 6028
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.44023866777409837,
+      "learning_rate": 7.051593138961681e-07,
+      "loss": 0.2292,
+      "step": 6029
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.696574520204921,
+      "learning_rate": 7.050651296718683e-07,
+      "loss": 0.0751,
+      "step": 6030
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7402758210906683,
+      "learning_rate": 7.049709366990777e-07,
+      "loss": 0.2056,
+      "step": 6031
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4702207701856258,
+      "learning_rate": 7.04876734981815e-07,
+      "loss": 0.0254,
+      "step": 6032
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.999792419930243,
+      "learning_rate": 7.047825245240988e-07,
+      "loss": 0.1965,
+      "step": 6033
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4877844417113965,
+      "learning_rate": 7.046883053299486e-07,
+      "loss": 0.116,
+      "step": 6034
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.479980954195937,
+      "learning_rate": 7.045940774033838e-07,
+      "loss": 0.0313,
+      "step": 6035
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6086161206512105,
+      "learning_rate": 7.044998407484243e-07,
+      "loss": 0.1707,
+      "step": 6036
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.187171812083103,
+      "learning_rate": 7.044055953690905e-07,
+      "loss": 0.2214,
+      "step": 6037
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0984291500373335,
+      "learning_rate": 7.043113412694031e-07,
+      "loss": 0.3816,
+      "step": 6038
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3096972089826218,
+      "learning_rate": 7.042170784533832e-07,
+      "loss": 0.1404,
+      "step": 6039
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4580931602231195,
+      "learning_rate": 7.041228069250522e-07,
+      "loss": 0.1525,
+      "step": 6040
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8939244281693258,
+      "learning_rate": 7.040285266884318e-07,
+      "loss": 0.1209,
+      "step": 6041
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.39145821050631974,
+      "learning_rate": 7.039342377475443e-07,
+      "loss": 0.0901,
+      "step": 6042
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4035249113873034,
+      "learning_rate": 7.038399401064124e-07,
+      "loss": 0.0596,
+      "step": 6043
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4781807350668425,
+      "learning_rate": 7.037456337690588e-07,
+      "loss": 0.1057,
+      "step": 6044
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5610597985770469,
+      "learning_rate": 7.03651318739507e-07,
+      "loss": 0.0974,
+      "step": 6045
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.77816916366953,
+      "learning_rate": 7.035569950217806e-07,
+      "loss": 0.1531,
+      "step": 6046
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5232881218806097,
+      "learning_rate": 7.034626626199034e-07,
+      "loss": 0.322,
+      "step": 6047
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.38390568318509016,
+      "learning_rate": 7.033683215379002e-07,
+      "loss": 0.0462,
+      "step": 6048
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.47140231927303317,
+      "learning_rate": 7.032739717797954e-07,
+      "loss": 0.2785,
+      "step": 6049
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.22127341481379223,
+      "learning_rate": 7.031796133496144e-07,
+      "loss": 0.1747,
+      "step": 6050
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5056716221063638,
+      "learning_rate": 7.030852462513826e-07,
+      "loss": 0.3887,
+      "step": 6051
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6106568240598306,
+      "learning_rate": 7.02990870489126e-07,
+      "loss": 0.4211,
+      "step": 6052
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6996628918092257,
+      "learning_rate": 7.028964860668706e-07,
+      "loss": 0.3672,
+      "step": 6053
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8179165141596455,
+      "learning_rate": 7.028020929886436e-07,
+      "loss": 0.2225,
+      "step": 6054
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6077679931879819,
+      "learning_rate": 7.027076912584714e-07,
+      "loss": 0.1728,
+      "step": 6055
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.433169476401399,
+      "learning_rate": 7.026132808803816e-07,
+      "loss": 0.0204,
+      "step": 6056
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7197853818572455,
+      "learning_rate": 7.02518861858402e-07,
+      "loss": 0.2011,
+      "step": 6057
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6432525605362424,
+      "learning_rate": 7.024244341965606e-07,
+      "loss": 0.3238,
+      "step": 6058
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6331576072836341,
+      "learning_rate": 7.023299978988859e-07,
+      "loss": 0.2177,
+      "step": 6059
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.527315811383846,
+      "learning_rate": 7.022355529694068e-07,
+      "loss": 0.0996,
+      "step": 6060
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5933556118618397,
+      "learning_rate": 7.021410994121524e-07,
+      "loss": 0.0674,
+      "step": 6061
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0699772808004508,
+      "learning_rate": 7.020466372311525e-07,
+      "loss": 0.0994,
+      "step": 6062
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8203860558455764,
+      "learning_rate": 7.019521664304369e-07,
+      "loss": 0.1917,
+      "step": 6063
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7259338690054316,
+      "learning_rate": 7.018576870140357e-07,
+      "loss": 0.1829,
+      "step": 6064
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7431332243039687,
+      "learning_rate": 7.017631989859799e-07,
+      "loss": 0.194,
+      "step": 6065
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9385818688452157,
+      "learning_rate": 7.016687023503004e-07,
+      "loss": 0.1326,
+      "step": 6066
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6791666522907975,
+      "learning_rate": 7.015741971110287e-07,
+      "loss": 0.1821,
+      "step": 6067
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7781131213115616,
+      "learning_rate": 7.014796832721965e-07,
+      "loss": 0.1181,
+      "step": 6068
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5342813351196428,
+      "learning_rate": 7.013851608378358e-07,
+      "loss": 0.1441,
+      "step": 6069
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4942220602028391,
+      "learning_rate": 7.012906298119796e-07,
+      "loss": 0.0107,
+      "step": 6070
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.515678796282028,
+      "learning_rate": 7.011960901986603e-07,
+      "loss": 0.1908,
+      "step": 6071
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5427986787855831,
+      "learning_rate": 7.011015420019115e-07,
+      "loss": 0.1059,
+      "step": 6072
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3379644858261285,
+      "learning_rate": 7.010069852257665e-07,
+      "loss": 0.013,
+      "step": 6073
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.744299123554008,
+      "learning_rate": 7.009124198742595e-07,
+      "loss": 0.0254,
+      "step": 6074
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9154454539648509,
+      "learning_rate": 7.008178459514249e-07,
+      "loss": 0.2579,
+      "step": 6075
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0282065266041691,
+      "learning_rate": 7.007232634612972e-07,
+      "loss": 0.1421,
+      "step": 6076
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5030692551752733,
+      "learning_rate": 7.006286724079115e-07,
+      "loss": 0.1216,
+      "step": 6077
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7340609318070733,
+      "learning_rate": 7.005340727953034e-07,
+      "loss": 0.3215,
+      "step": 6078
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7045726026405857,
+      "learning_rate": 7.004394646275086e-07,
+      "loss": 0.3236,
+      "step": 6079
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2988262294681189,
+      "learning_rate": 7.003448479085634e-07,
+      "loss": 0.096,
+      "step": 6080
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.053010225464449,
+      "learning_rate": 7.002502226425041e-07,
+      "loss": 0.116,
+      "step": 6081
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8059010644650134,
+      "learning_rate": 7.001555888333679e-07,
+      "loss": 0.2569,
+      "step": 6082
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5653028761079574,
+      "learning_rate": 7.000609464851918e-07,
+      "loss": 0.2474,
+      "step": 6083
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2069537818478193,
+      "learning_rate": 6.999662956020136e-07,
+      "loss": 0.3237,
+      "step": 6084
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.44003547281224015,
+      "learning_rate": 6.998716361878712e-07,
+      "loss": 0.4035,
+      "step": 6085
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9190976304451878,
+      "learning_rate": 6.99776968246803e-07,
+      "loss": 0.1739,
+      "step": 6086
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1500311324465582,
+      "learning_rate": 6.996822917828476e-07,
+      "loss": 0.2147,
+      "step": 6087
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.03477222308533221,
+      "learning_rate": 6.995876068000445e-07,
+      "loss": 0.0008,
+      "step": 6088
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5832092753911268,
+      "learning_rate": 6.994929133024329e-07,
+      "loss": 0.3132,
+      "step": 6089
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.524669557833631,
+      "learning_rate": 6.993982112940525e-07,
+      "loss": 0.1944,
+      "step": 6090
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5149018659822298,
+      "learning_rate": 6.993035007789434e-07,
+      "loss": 0.3219,
+      "step": 6091
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1683874690062729,
+      "learning_rate": 6.992087817611466e-07,
+      "loss": 0.0672,
+      "step": 6092
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6868417573536876,
+      "learning_rate": 6.991140542447024e-07,
+      "loss": 0.1113,
+      "step": 6093
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.306435480036509,
+      "learning_rate": 6.990193182336524e-07,
+      "loss": 0.1352,
+      "step": 6094
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.981389326213064,
+      "learning_rate": 6.989245737320383e-07,
+      "loss": 0.1822,
+      "step": 6095
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5881073089154503,
+      "learning_rate": 6.988298207439021e-07,
+      "loss": 0.4433,
+      "step": 6096
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.862365536486719,
+      "learning_rate": 6.98735059273286e-07,
+      "loss": 0.2503,
+      "step": 6097
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.137412049406784,
+      "learning_rate": 6.986402893242326e-07,
+      "loss": 0.0186,
+      "step": 6098
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.629167955438788,
+      "learning_rate": 6.985455109007853e-07,
+      "loss": 0.0771,
+      "step": 6099
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7568199747826508,
+      "learning_rate": 6.984507240069873e-07,
+      "loss": 0.0125,
+      "step": 6100
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4294833206324395,
+      "learning_rate": 6.983559286468825e-07,
+      "loss": 0.1637,
+      "step": 6101
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0130177508255256,
+      "learning_rate": 6.982611248245152e-07,
+      "loss": 0.4066,
+      "step": 6102
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6991349913435341,
+      "learning_rate": 6.981663125439295e-07,
+      "loss": 0.3924,
+      "step": 6103
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8832818984152153,
+      "learning_rate": 6.980714918091706e-07,
+      "loss": 0.3424,
+      "step": 6104
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3879801519145525,
+      "learning_rate": 6.979766626242838e-07,
+      "loss": 0.0624,
+      "step": 6105
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.195726083938898,
+      "learning_rate": 6.978818249933145e-07,
+      "loss": 0.1041,
+      "step": 6106
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0888581955067818,
+      "learning_rate": 6.977869789203088e-07,
+      "loss": 0.247,
+      "step": 6107
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1743268342712296,
+      "learning_rate": 6.97692124409313e-07,
+      "loss": 0.0383,
+      "step": 6108
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0650161679738679,
+      "learning_rate": 6.975972614643737e-07,
+      "loss": 0.1798,
+      "step": 6109
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9407940742659258,
+      "learning_rate": 6.97502390089538e-07,
+      "loss": 0.3623,
+      "step": 6110
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9933576601280183,
+      "learning_rate": 6.974075102888535e-07,
+      "loss": 0.2004,
+      "step": 6111
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9204704042847702,
+      "learning_rate": 6.973126220663675e-07,
+      "loss": 0.3363,
+      "step": 6112
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.229552263715744,
+      "learning_rate": 6.972177254261285e-07,
+      "loss": 0.3976,
+      "step": 6113
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7396302309870308,
+      "learning_rate": 6.971228203721848e-07,
+      "loss": 0.1982,
+      "step": 6114
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5209664358944699,
+      "learning_rate": 6.970279069085855e-07,
+      "loss": 0.144,
+      "step": 6115
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4425491786711748,
+      "learning_rate": 6.969329850393795e-07,
+      "loss": 0.1312,
+      "step": 6116
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5263695534515982,
+      "learning_rate": 6.968380547686166e-07,
+      "loss": 0.3222,
+      "step": 6117
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.575979932582315,
+      "learning_rate": 6.967431161003465e-07,
+      "loss": 0.0941,
+      "step": 6118
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.222430021897893,
+      "learning_rate": 6.966481690386195e-07,
+      "loss": 0.0992,
+      "step": 6119
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2576809504348785,
+      "learning_rate": 6.965532135874863e-07,
+      "loss": 0.1708,
+      "step": 6120
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.854626612892088,
+      "learning_rate": 6.96458249750998e-07,
+      "loss": 0.0214,
+      "step": 6121
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2937559918030428,
+      "learning_rate": 6.963632775332055e-07,
+      "loss": 0.3294,
+      "step": 6122
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6462685025279785,
+      "learning_rate": 6.962682969381613e-07,
+      "loss": 0.1446,
+      "step": 6123
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.144011777287466,
+      "learning_rate": 6.961733079699168e-07,
+      "loss": 0.3612,
+      "step": 6124
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.762259905785336,
+      "learning_rate": 6.960783106325246e-07,
+      "loss": 0.4583,
+      "step": 6125
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5472923641710189,
+      "learning_rate": 6.959833049300375e-07,
+      "loss": 0.1341,
+      "step": 6126
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.618147425453013,
+      "learning_rate": 6.958882908665087e-07,
+      "loss": 0.2312,
+      "step": 6127
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.039500895902682,
+      "learning_rate": 6.957932684459915e-07,
+      "loss": 0.2508,
+      "step": 6128
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7293971604524371,
+      "learning_rate": 6.9569823767254e-07,
+      "loss": 0.2275,
+      "step": 6129
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8612548679002621,
+      "learning_rate": 6.956031985502084e-07,
+      "loss": 0.0322,
+      "step": 6130
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8993388064077994,
+      "learning_rate": 6.955081510830509e-07,
+      "loss": 0.1982,
+      "step": 6131
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8921071696384599,
+      "learning_rate": 6.954130952751227e-07,
+      "loss": 0.1631,
+      "step": 6132
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.44758189450209807,
+      "learning_rate": 6.953180311304792e-07,
+      "loss": 0.2095,
+      "step": 6133
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0596157588948227,
+      "learning_rate": 6.952229586531756e-07,
+      "loss": 0.2035,
+      "step": 6134
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.43274357682038095,
+      "learning_rate": 6.951278778472682e-07,
+      "loss": 0.0727,
+      "step": 6135
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.32998110733984454,
+      "learning_rate": 6.950327887168133e-07,
+      "loss": 0.2334,
+      "step": 6136
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6791760685309348,
+      "learning_rate": 6.949376912658678e-07,
+      "loss": 0.2407,
+      "step": 6137
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4853614325615889,
+      "learning_rate": 6.948425854984883e-07,
+      "loss": 0.1118,
+      "step": 6138
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2479386879988221,
+      "learning_rate": 6.947474714187324e-07,
+      "loss": 0.1549,
+      "step": 6139
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3361662390743687,
+      "learning_rate": 6.946523490306578e-07,
+      "loss": 0.0888,
+      "step": 6140
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6868660305286954,
+      "learning_rate": 6.945572183383229e-07,
+      "loss": 0.1573,
+      "step": 6141
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2057914628517277,
+      "learning_rate": 6.944620793457857e-07,
+      "loss": 0.3429,
+      "step": 6142
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7712588457012892,
+      "learning_rate": 6.943669320571055e-07,
+      "loss": 0.0063,
+      "step": 6143
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.214776392904596,
+      "learning_rate": 6.942717764763412e-07,
+      "loss": 0.2139,
+      "step": 6144
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.824250979215668,
+      "learning_rate": 6.941766126075524e-07,
+      "loss": 0.1055,
+      "step": 6145
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.429547489762958,
+      "learning_rate": 6.94081440454799e-07,
+      "loss": 0.1091,
+      "step": 6146
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9529996714874591,
+      "learning_rate": 6.939862600221411e-07,
+      "loss": 0.2029,
+      "step": 6147
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5255961179150546,
+      "learning_rate": 6.938910713136393e-07,
+      "loss": 0.405,
+      "step": 6148
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6362585419101171,
+      "learning_rate": 6.937958743333548e-07,
+      "loss": 0.1283,
+      "step": 6149
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.465940031687446,
+      "learning_rate": 6.937006690853486e-07,
+      "loss": 0.0199,
+      "step": 6150
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7828230822884971,
+      "learning_rate": 6.936054555736825e-07,
+      "loss": 0.2407,
+      "step": 6151
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.25168434736024076,
+      "learning_rate": 6.935102338024185e-07,
+      "loss": 0.174,
+      "step": 6152
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8794224204941419,
+      "learning_rate": 6.93415003775619e-07,
+      "loss": 0.3587,
+      "step": 6153
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9803000681134058,
+      "learning_rate": 6.933197654973466e-07,
+      "loss": 0.1463,
+      "step": 6154
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.20458817097033785,
+      "learning_rate": 6.932245189716643e-07,
+      "loss": 0.0215,
+      "step": 6155
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1878195279602228,
+      "learning_rate": 6.931292642026356e-07,
+      "loss": 0.1178,
+      "step": 6156
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9650560839732295,
+      "learning_rate": 6.930340011943244e-07,
+      "loss": 0.3217,
+      "step": 6157
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6340945096257533,
+      "learning_rate": 6.929387299507944e-07,
+      "loss": 0.4892,
+      "step": 6158
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.655075799508818,
+      "learning_rate": 6.928434504761105e-07,
+      "loss": 0.0638,
+      "step": 6159
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.61875768161717,
+      "learning_rate": 6.927481627743373e-07,
+      "loss": 0.1523,
+      "step": 6160
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.033864607713213,
+      "learning_rate": 6.926528668495399e-07,
+      "loss": 0.2278,
+      "step": 6161
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7826141802236259,
+      "learning_rate": 6.925575627057841e-07,
+      "loss": 0.0551,
+      "step": 6162
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28373260478851425,
+      "learning_rate": 6.924622503471356e-07,
+      "loss": 0.0812,
+      "step": 6163
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8340090495261956,
+      "learning_rate": 6.923669297776604e-07,
+      "loss": 0.1287,
+      "step": 6164
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.672714393111128,
+      "learning_rate": 6.922716010014255e-07,
+      "loss": 0.1098,
+      "step": 6165
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.871746698251231,
+      "learning_rate": 6.921762640224974e-07,
+      "loss": 0.0985,
+      "step": 6166
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7736099022362137,
+      "learning_rate": 6.920809188449435e-07,
+      "loss": 0.0043,
+      "step": 6167
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8113946228779848,
+      "learning_rate": 6.919855654728316e-07,
+      "loss": 0.1173,
+      "step": 6168
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9167850203021984,
+      "learning_rate": 6.918902039102296e-07,
+      "loss": 0.2546,
+      "step": 6169
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7498944438835116,
+      "learning_rate": 6.917948341612056e-07,
+      "loss": 0.2329,
+      "step": 6170
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.063631914034943,
+      "learning_rate": 6.916994562298285e-07,
+      "loss": 0.1669,
+      "step": 6171
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7192687269786702,
+      "learning_rate": 6.916040701201674e-07,
+      "loss": 0.2198,
+      "step": 6172
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7899638685431615,
+      "learning_rate": 6.915086758362914e-07,
+      "loss": 0.2831,
+      "step": 6173
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9522419273413391,
+      "learning_rate": 6.914132733822701e-07,
+      "loss": 0.0687,
+      "step": 6174
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3403160937259742,
+      "learning_rate": 6.913178627621739e-07,
+      "loss": 0.0447,
+      "step": 6175
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8785572537418758,
+      "learning_rate": 6.912224439800731e-07,
+      "loss": 0.08,
+      "step": 6176
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.0105327342004704,
+      "learning_rate": 6.911270170400384e-07,
+      "loss": 0.1012,
+      "step": 6177
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7582801548032934,
+      "learning_rate": 6.91031581946141e-07,
+      "loss": 0.2048,
+      "step": 6178
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4176320377241375,
+      "learning_rate": 6.909361387024522e-07,
+      "loss": 0.0817,
+      "step": 6179
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.719948092491281,
+      "learning_rate": 6.908406873130439e-07,
+      "loss": 0.1972,
+      "step": 6180
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4633957144557999,
+      "learning_rate": 6.907452277819883e-07,
+      "loss": 0.2768,
+      "step": 6181
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.384205034374505,
+      "learning_rate": 6.906497601133579e-07,
+      "loss": 0.1005,
+      "step": 6182
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5906550470369817,
+      "learning_rate": 6.905542843112253e-07,
+      "loss": 0.1545,
+      "step": 6183
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5678397989120448,
+      "learning_rate": 6.90458800379664e-07,
+      "loss": 0.2816,
+      "step": 6184
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.384474132225101,
+      "learning_rate": 6.903633083227474e-07,
+      "loss": 0.1302,
+      "step": 6185
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.982603229099718,
+      "learning_rate": 6.902678081445494e-07,
+      "loss": 0.0161,
+      "step": 6186
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2659266276101535,
+      "learning_rate": 6.901722998491441e-07,
+      "loss": 0.0176,
+      "step": 6187
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2677240536185697,
+      "learning_rate": 6.900767834406063e-07,
+      "loss": 0.1869,
+      "step": 6188
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5057187867373214,
+      "learning_rate": 6.899812589230108e-07,
+      "loss": 0.1465,
+      "step": 6189
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6191932612231814,
+      "learning_rate": 6.89885726300433e-07,
+      "loss": 0.3231,
+      "step": 6190
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2446369380745481,
+      "learning_rate": 6.897901855769483e-07,
+      "loss": 0.0954,
+      "step": 6191
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8173892522714576,
+      "learning_rate": 6.896946367566327e-07,
+      "loss": 0.1784,
+      "step": 6192
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5691305531822278,
+      "learning_rate": 6.895990798435625e-07,
+      "loss": 0.2155,
+      "step": 6193
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0434859744339169,
+      "learning_rate": 6.895035148418144e-07,
+      "loss": 0.5934,
+      "step": 6194
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.35781916837890215,
+      "learning_rate": 6.894079417554655e-07,
+      "loss": 0.0663,
+      "step": 6195
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.24015978873153349,
+      "learning_rate": 6.893123605885931e-07,
+      "loss": 0.0623,
+      "step": 6196
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.321693853670986,
+      "learning_rate": 6.892167713452748e-07,
+      "loss": 0.3533,
+      "step": 6197
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.48782737986281427,
+      "learning_rate": 6.891211740295887e-07,
+      "loss": 0.1524,
+      "step": 6198
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.11333829615908207,
+      "learning_rate": 6.890255686456133e-07,
+      "loss": 0.0053,
+      "step": 6199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9010655888481797,
+      "learning_rate": 6.889299551974268e-07,
+      "loss": 0.3647,
+      "step": 6200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5065096572806629,
+      "learning_rate": 6.888343336891087e-07,
+      "loss": 0.2117,
+      "step": 6201
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.14211318909038,
+      "learning_rate": 6.887387041247384e-07,
+      "loss": 0.1118,
+      "step": 6202
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3927341176803692,
+      "learning_rate": 6.886430665083955e-07,
+      "loss": 0.0819,
+      "step": 6203
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3924321143043787,
+      "learning_rate": 6.885474208441601e-07,
+      "loss": 0.127,
+      "step": 6204
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4337249428796377,
+      "learning_rate": 6.88451767136113e-07,
+      "loss": 0.2451,
+      "step": 6205
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6311639888403874,
+      "learning_rate": 6.883561053883344e-07,
+      "loss": 0.401,
+      "step": 6206
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5139037001185284,
+      "learning_rate": 6.882604356049058e-07,
+      "loss": 0.1032,
+      "step": 6207
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.137946904279486,
+      "learning_rate": 6.881647577899086e-07,
+      "loss": 0.2727,
+      "step": 6208
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.146831100302851,
+      "learning_rate": 6.880690719474245e-07,
+      "loss": 0.0998,
+      "step": 6209
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8310608478545772,
+      "learning_rate": 6.879733780815357e-07,
+      "loss": 0.1059,
+      "step": 6210
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4979772168898703,
+      "learning_rate": 6.878776761963248e-07,
+      "loss": 0.0443,
+      "step": 6211
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8899238443742646,
+      "learning_rate": 6.877819662958743e-07,
+      "loss": 0.292,
+      "step": 6212
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4890042180484366,
+      "learning_rate": 6.87686248384268e-07,
+      "loss": 0.3342,
+      "step": 6213
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6739477447958372,
+      "learning_rate": 6.875905224655889e-07,
+      "loss": 0.2706,
+      "step": 6214
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5369417685652146,
+      "learning_rate": 6.874947885439211e-07,
+      "loss": 0.0872,
+      "step": 6215
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.12216211068175181,
+      "learning_rate": 6.873990466233486e-07,
+      "loss": 0.0638,
+      "step": 6216
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9266919734418062,
+      "learning_rate": 6.87303296707956e-07,
+      "loss": 0.0298,
+      "step": 6217
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9563162868912487,
+      "learning_rate": 6.872075388018284e-07,
+      "loss": 0.072,
+      "step": 6218
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.86366765529282,
+      "learning_rate": 6.871117729090508e-07,
+      "loss": 0.2834,
+      "step": 6219
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.183708277852859,
+      "learning_rate": 6.870159990337086e-07,
+      "loss": 0.0882,
+      "step": 6220
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.285448526287848,
+      "learning_rate": 6.869202171798881e-07,
+      "loss": 0.2821,
+      "step": 6221
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9273322831659865,
+      "learning_rate": 6.868244273516755e-07,
+      "loss": 0.2233,
+      "step": 6222
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.179773481144634,
+      "learning_rate": 6.867286295531571e-07,
+      "loss": 0.2397,
+      "step": 6223
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8681641867231622,
+      "learning_rate": 6.8663282378842e-07,
+      "loss": 0.1859,
+      "step": 6224
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5727729885534764,
+      "learning_rate": 6.865370100615515e-07,
+      "loss": 0.265,
+      "step": 6225
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30902591923894296,
+      "learning_rate": 6.864411883766393e-07,
+      "loss": 0.1712,
+      "step": 6226
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1296518840080736,
+      "learning_rate": 6.863453587377711e-07,
+      "loss": 0.3299,
+      "step": 6227
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.940590276393374,
+      "learning_rate": 6.862495211490352e-07,
+      "loss": 0.1227,
+      "step": 6228
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4717924375144069,
+      "learning_rate": 6.861536756145205e-07,
+      "loss": 0.183,
+      "step": 6229
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.908760307610912,
+      "learning_rate": 6.860578221383155e-07,
+      "loss": 0.0143,
+      "step": 6230
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9651197902918841,
+      "learning_rate": 6.859619607245101e-07,
+      "loss": 0.1343,
+      "step": 6231
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9555572567907259,
+      "learning_rate": 6.858660913771934e-07,
+      "loss": 0.0517,
+      "step": 6232
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0019228992172688,
+      "learning_rate": 6.857702141004558e-07,
+      "loss": 0.1124,
+      "step": 6233
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.375353658408546,
+      "learning_rate": 6.856743288983873e-07,
+      "loss": 0.1988,
+      "step": 6234
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5454114092696687,
+      "learning_rate": 6.855784357750786e-07,
+      "loss": 0.4377,
+      "step": 6235
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2153189426097419,
+      "learning_rate": 6.854825347346209e-07,
+      "loss": 0.0625,
+      "step": 6236
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.931577506079022,
+      "learning_rate": 6.853866257811054e-07,
+      "loss": 0.0245,
+      "step": 6237
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.686939894574234,
+      "learning_rate": 6.852907089186236e-07,
+      "loss": 0.0939,
+      "step": 6238
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8131809197901665,
+      "learning_rate": 6.851947841512679e-07,
+      "loss": 0.2382,
+      "step": 6239
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.497959138251734,
+      "learning_rate": 6.850988514831303e-07,
+      "loss": 0.0311,
+      "step": 6240
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.275365788480474,
+      "learning_rate": 6.850029109183038e-07,
+      "loss": 0.1505,
+      "step": 6241
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2547152339384144,
+      "learning_rate": 6.84906962460881e-07,
+      "loss": 0.1546,
+      "step": 6242
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3994564711131799,
+      "learning_rate": 6.848110061149555e-07,
+      "loss": 0.0257,
+      "step": 6243
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0345026380540454,
+      "learning_rate": 6.84715041884621e-07,
+      "loss": 0.3793,
+      "step": 6244
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.43390028267813124,
+      "learning_rate": 6.846190697739714e-07,
+      "loss": 0.0355,
+      "step": 6245
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4501212776602618,
+      "learning_rate": 6.845230897871012e-07,
+      "loss": 0.1599,
+      "step": 6246
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.580929022002219,
+      "learning_rate": 6.84427101928105e-07,
+      "loss": 0.15,
+      "step": 6247
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7918955130027382,
+      "learning_rate": 6.84331106201078e-07,
+      "loss": 0.1475,
+      "step": 6248
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7569238348928278,
+      "learning_rate": 6.842351026101154e-07,
+      "loss": 0.107,
+      "step": 6249
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4995717604473063,
+      "learning_rate": 6.84139091159313e-07,
+      "loss": 0.2009,
+      "step": 6250
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9157557633717965,
+      "learning_rate": 6.840430718527667e-07,
+      "loss": 0.003,
+      "step": 6251
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7124468562585222,
+      "learning_rate": 6.839470446945732e-07,
+      "loss": 0.2808,
+      "step": 6252
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2051956401988941,
+      "learning_rate": 6.838510096888288e-07,
+      "loss": 0.2728,
+      "step": 6253
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5212681126804175,
+      "learning_rate": 6.837549668396309e-07,
+      "loss": 0.0165,
+      "step": 6254
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7271535467615731,
+      "learning_rate": 6.836589161510766e-07,
+      "loss": 0.1922,
+      "step": 6255
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9817213786470782,
+      "learning_rate": 6.835628576272637e-07,
+      "loss": 0.244,
+      "step": 6256
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0020861976946314,
+      "learning_rate": 6.834667912722904e-07,
+      "loss": 0.2997,
+      "step": 6257
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.44966370145242784,
+      "learning_rate": 6.833707170902549e-07,
+      "loss": 0.1575,
+      "step": 6258
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5248286096091686,
+      "learning_rate": 6.832746350852561e-07,
+      "loss": 0.2467,
+      "step": 6259
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6208019392122317,
+      "learning_rate": 6.831785452613927e-07,
+      "loss": 0.2825,
+      "step": 6260
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.057503488458636,
+      "learning_rate": 6.830824476227646e-07,
+      "loss": 0.4394,
+      "step": 6261
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4542964719792616,
+      "learning_rate": 6.82986342173471e-07,
+      "loss": 0.3187,
+      "step": 6262
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0216584528054076,
+      "learning_rate": 6.828902289176124e-07,
+      "loss": 0.0623,
+      "step": 6263
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3721115803693167,
+      "learning_rate": 6.827941078592888e-07,
+      "loss": 0.2275,
+      "step": 6264
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6876065045926866,
+      "learning_rate": 6.826979790026012e-07,
+      "loss": 0.2629,
+      "step": 6265
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5959189359618354,
+      "learning_rate": 6.826018423516505e-07,
+      "loss": 0.0692,
+      "step": 6266
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0987326793714454,
+      "learning_rate": 6.825056979105381e-07,
+      "loss": 0.3505,
+      "step": 6267
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5305220700049564,
+      "learning_rate": 6.82409545683366e-07,
+      "loss": 0.2905,
+      "step": 6268
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9076133219416157,
+      "learning_rate": 6.823133856742358e-07,
+      "loss": 0.3421,
+      "step": 6269
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0344516297215987,
+      "learning_rate": 6.822172178872501e-07,
+      "loss": 0.1861,
+      "step": 6270
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6141548408873324,
+      "learning_rate": 6.821210423265115e-07,
+      "loss": 0.1325,
+      "step": 6271
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.259281383914091,
+      "learning_rate": 6.820248589961231e-07,
+      "loss": 0.0573,
+      "step": 6272
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1936161211338814,
+      "learning_rate": 6.819286679001883e-07,
+      "loss": 0.0052,
+      "step": 6273
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4761750848766932,
+      "learning_rate": 6.818324690428109e-07,
+      "loss": 0.0708,
+      "step": 6274
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30141760885628427,
+      "learning_rate": 6.817362624280948e-07,
+      "loss": 0.1181,
+      "step": 6275
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.27387131059069,
+      "learning_rate": 6.816400480601444e-07,
+      "loss": 0.111,
+      "step": 6276
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.78459022870707,
+      "learning_rate": 6.815438259430645e-07,
+      "loss": 0.2472,
+      "step": 6277
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9573281104683642,
+      "learning_rate": 6.8144759608096e-07,
+      "loss": 0.1301,
+      "step": 6278
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.028878427849481,
+      "learning_rate": 6.813513584779362e-07,
+      "loss": 0.2653,
+      "step": 6279
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.381367328172131,
+      "learning_rate": 6.81255113138099e-07,
+      "loss": 0.0072,
+      "step": 6280
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4528935142791577,
+      "learning_rate": 6.811588600655542e-07,
+      "loss": 0.1284,
+      "step": 6281
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.827426492833447,
+      "learning_rate": 6.810625992644083e-07,
+      "loss": 0.156,
+      "step": 6282
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5734812805451183,
+      "learning_rate": 6.80966330738768e-07,
+      "loss": 0.0737,
+      "step": 6283
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4426379072669722,
+      "learning_rate": 6.808700544927402e-07,
+      "loss": 0.1289,
+      "step": 6284
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5192769835580945,
+      "learning_rate": 6.807737705304323e-07,
+      "loss": 0.2355,
+      "step": 6285
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6008560869470537,
+      "learning_rate": 6.806774788559519e-07,
+      "loss": 0.2676,
+      "step": 6286
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9184380307491457,
+      "learning_rate": 6.80581179473407e-07,
+      "loss": 0.4086,
+      "step": 6287
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.24984547522423572,
+      "learning_rate": 6.804848723869061e-07,
+      "loss": 0.2237,
+      "step": 6288
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5476870152107928,
+      "learning_rate": 6.803885576005577e-07,
+      "loss": 0.1659,
+      "step": 6289
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5843547615793472,
+      "learning_rate": 6.802922351184707e-07,
+      "loss": 0.1208,
+      "step": 6290
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.36539251437592996,
+      "learning_rate": 6.801959049447545e-07,
+      "loss": 0.1999,
+      "step": 6291
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.163480671211163,
+      "learning_rate": 6.800995670835188e-07,
+      "loss": 0.1565,
+      "step": 6292
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 15.468156813312886,
+      "learning_rate": 6.800032215388736e-07,
+      "loss": 0.1801,
+      "step": 6293
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5391643643765122,
+      "learning_rate": 6.79906868314929e-07,
+      "loss": 0.3423,
+      "step": 6294
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1876241664389382,
+      "learning_rate": 6.798105074157958e-07,
+      "loss": 0.1654,
+      "step": 6295
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8193904871120007,
+      "learning_rate": 6.797141388455851e-07,
+      "loss": 0.2495,
+      "step": 6296
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3866432513975203,
+      "learning_rate": 6.796177626084078e-07,
+      "loss": 0.164,
+      "step": 6297
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0882381860432704,
+      "learning_rate": 6.795213787083755e-07,
+      "loss": 0.1788,
+      "step": 6298
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6830235621905923,
+      "learning_rate": 6.794249871496005e-07,
+      "loss": 0.3599,
+      "step": 6299
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.695270607090052,
+      "learning_rate": 6.793285879361948e-07,
+      "loss": 0.0877,
+      "step": 6300
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7194837379626305,
+      "learning_rate": 6.79232181072271e-07,
+      "loss": 0.3207,
+      "step": 6301
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9357192862350031,
+      "learning_rate": 6.79135766561942e-07,
+      "loss": 0.3821,
+      "step": 6302
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5450760651607603,
+      "learning_rate": 6.790393444093213e-07,
+      "loss": 0.1295,
+      "step": 6303
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5605722159660916,
+      "learning_rate": 6.789429146185222e-07,
+      "loss": 0.0503,
+      "step": 6304
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.223088798425785,
+      "learning_rate": 6.788464771936585e-07,
+      "loss": 0.5387,
+      "step": 6305
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.428314502941871,
+      "learning_rate": 6.787500321388447e-07,
+      "loss": 0.004,
+      "step": 6306
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0015165823477454,
+      "learning_rate": 6.786535794581951e-07,
+      "loss": 0.1085,
+      "step": 6307
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.178132416984318,
+      "learning_rate": 6.785571191558247e-07,
+      "loss": 0.1638,
+      "step": 6308
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7832864868245555,
+      "learning_rate": 6.784606512358486e-07,
+      "loss": 0.1841,
+      "step": 6309
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2342575567020215,
+      "learning_rate": 6.783641757023825e-07,
+      "loss": 0.3512,
+      "step": 6310
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.02782269276322806,
+      "learning_rate": 6.78267692559542e-07,
+      "loss": 0.0007,
+      "step": 6311
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9632128514314175,
+      "learning_rate": 6.781712018114434e-07,
+      "loss": 0.2452,
+      "step": 6312
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9387459585892254,
+      "learning_rate": 6.780747034622032e-07,
+      "loss": 0.1231,
+      "step": 6313
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6726956733244531,
+      "learning_rate": 6.77978197515938e-07,
+      "loss": 0.1672,
+      "step": 6314
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8011892718369154,
+      "learning_rate": 6.778816839767654e-07,
+      "loss": 0.2529,
+      "step": 6315
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.654576615629492,
+      "learning_rate": 6.777851628488022e-07,
+      "loss": 0.1318,
+      "step": 6316
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.468742702068239,
+      "learning_rate": 6.776886341361668e-07,
+      "loss": 0.1593,
+      "step": 6317
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8942879143321072,
+      "learning_rate": 6.77592097842977e-07,
+      "loss": 0.3876,
+      "step": 6318
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6368684574518247,
+      "learning_rate": 6.774955539733514e-07,
+      "loss": 0.0998,
+      "step": 6319
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7566350153543336,
+      "learning_rate": 6.773990025314086e-07,
+      "loss": 0.1656,
+      "step": 6320
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.731691102948197,
+      "learning_rate": 6.773024435212677e-07,
+      "loss": 0.0195,
+      "step": 6321
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6343300683315546,
+      "learning_rate": 6.772058769470483e-07,
+      "loss": 0.2368,
+      "step": 6322
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.180929949863677,
+      "learning_rate": 6.771093028128699e-07,
+      "loss": 0.0249,
+      "step": 6323
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.20453988497507,
+      "learning_rate": 6.770127211228526e-07,
+      "loss": 0.1609,
+      "step": 6324
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.817803980307327,
+      "learning_rate": 6.769161318811165e-07,
+      "loss": 0.1692,
+      "step": 6325
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6893446552490462,
+      "learning_rate": 6.76819535091783e-07,
+      "loss": 0.2116,
+      "step": 6326
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.248398235457054,
+      "learning_rate": 6.767229307589723e-07,
+      "loss": 0.1265,
+      "step": 6327
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1731612265180096,
+      "learning_rate": 6.766263188868063e-07,
+      "loss": 0.4075,
+      "step": 6328
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1095975619987564,
+      "learning_rate": 6.765296994794065e-07,
+      "loss": 0.341,
+      "step": 6329
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.232819800035905,
+      "learning_rate": 6.764330725408948e-07,
+      "loss": 0.4152,
+      "step": 6330
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.40808648334532943,
+      "learning_rate": 6.763364380753936e-07,
+      "loss": 0.0156,
+      "step": 6331
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3680753344624959,
+      "learning_rate": 6.762397960870255e-07,
+      "loss": 0.1849,
+      "step": 6332
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9161980014580507,
+      "learning_rate": 6.761431465799133e-07,
+      "loss": 0.2478,
+      "step": 6333
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7829457137275232,
+      "learning_rate": 6.760464895581804e-07,
+      "loss": 0.3173,
+      "step": 6334
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.025450958864549,
+      "learning_rate": 6.759498250259503e-07,
+      "loss": 0.2282,
+      "step": 6335
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.622090011675234,
+      "learning_rate": 6.75853152987347e-07,
+      "loss": 0.2927,
+      "step": 6336
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5802494290831592,
+      "learning_rate": 6.757564734464948e-07,
+      "loss": 0.1875,
+      "step": 6337
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4361417755347666,
+      "learning_rate": 6.756597864075181e-07,
+      "loss": 0.05,
+      "step": 6338
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.166029941431076,
+      "learning_rate": 6.755630918745417e-07,
+      "loss": 0.166,
+      "step": 6339
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7587754413586166,
+      "learning_rate": 6.754663898516909e-07,
+      "loss": 0.2583,
+      "step": 6340
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.35510815215948655,
+      "learning_rate": 6.753696803430913e-07,
+      "loss": 0.1415,
+      "step": 6341
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.942727015461539,
+      "learning_rate": 6.752729633528684e-07,
+      "loss": 0.296,
+      "step": 6342
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9292487618931184,
+      "learning_rate": 6.751762388851486e-07,
+      "loss": 0.1719,
+      "step": 6343
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5921404678657156,
+      "learning_rate": 6.750795069440582e-07,
+      "loss": 0.0233,
+      "step": 6344
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.063019208120071,
+      "learning_rate": 6.749827675337243e-07,
+      "loss": 0.1282,
+      "step": 6345
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6059948271698828,
+      "learning_rate": 6.748860206582738e-07,
+      "loss": 0.3899,
+      "step": 6346
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.819963819050505,
+      "learning_rate": 6.747892663218339e-07,
+      "loss": 0.0624,
+      "step": 6347
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.087721441077157,
+      "learning_rate": 6.746925045285327e-07,
+      "loss": 0.0145,
+      "step": 6348
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1544841294131863,
+      "learning_rate": 6.745957352824979e-07,
+      "loss": 0.0223,
+      "step": 6349
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9026185574747579,
+      "learning_rate": 6.744989585878582e-07,
+      "loss": 0.0029,
+      "step": 6350
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.920157172626653,
+      "learning_rate": 6.744021744487422e-07,
+      "loss": 0.3463,
+      "step": 6351
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8754340127993847,
+      "learning_rate": 6.743053828692787e-07,
+      "loss": 0.2398,
+      "step": 6352
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.125188320794302,
+      "learning_rate": 6.742085838535972e-07,
+      "loss": 0.2402,
+      "step": 6353
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.09075343192105742,
+      "learning_rate": 6.741117774058271e-07,
+      "loss": 0.0033,
+      "step": 6354
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.591763171934437,
+      "learning_rate": 6.740149635300989e-07,
+      "loss": 0.3425,
+      "step": 6355
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9332673410017857,
+      "learning_rate": 6.739181422305424e-07,
+      "loss": 0.3002,
+      "step": 6356
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6561116635671101,
+      "learning_rate": 6.738213135112884e-07,
+      "loss": 0.1401,
+      "step": 6357
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5671755029146652,
+      "learning_rate": 6.737244773764677e-07,
+      "loss": 0.2599,
+      "step": 6358
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9630947590414323,
+      "learning_rate": 6.736276338302115e-07,
+      "loss": 0.2172,
+      "step": 6359
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1197194309821634,
+      "learning_rate": 6.735307828766514e-07,
+      "loss": 0.1828,
+      "step": 6360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4267255933308179,
+      "learning_rate": 6.734339245199194e-07,
+      "loss": 0.2289,
+      "step": 6361
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.061760439965799295,
+      "learning_rate": 6.733370587641473e-07,
+      "loss": 0.0005,
+      "step": 6362
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0401649801124897,
+      "learning_rate": 6.732401856134681e-07,
+      "loss": 0.2416,
+      "step": 6363
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1427021284904402,
+      "learning_rate": 6.731433050720143e-07,
+      "loss": 0.1625,
+      "step": 6364
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7584669024460573,
+      "learning_rate": 6.730464171439189e-07,
+      "loss": 0.5968,
+      "step": 6365
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.947389546604894,
+      "learning_rate": 6.729495218333156e-07,
+      "loss": 0.3292,
+      "step": 6366
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2671454166673617,
+      "learning_rate": 6.728526191443379e-07,
+      "loss": 0.3977,
+      "step": 6367
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2807435233821538,
+      "learning_rate": 6.7275570908112e-07,
+      "loss": 0.1983,
+      "step": 6368
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1223775443224318,
+      "learning_rate": 6.726587916477963e-07,
+      "loss": 0.2572,
+      "step": 6369
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7796094429732707,
+      "learning_rate": 6.725618668485015e-07,
+      "loss": 0.2362,
+      "step": 6370
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8569312131457432,
+      "learning_rate": 6.724649346873705e-07,
+      "loss": 0.0895,
+      "step": 6371
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3162810885051341,
+      "learning_rate": 6.723679951685387e-07,
+      "loss": 0.1953,
+      "step": 6372
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0160207518359947,
+      "learning_rate": 6.722710482961418e-07,
+      "loss": 0.1238,
+      "step": 6373
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.239643357815333,
+      "learning_rate": 6.721740940743155e-07,
+      "loss": 0.101,
+      "step": 6374
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7511605584213781,
+      "learning_rate": 6.720771325071965e-07,
+      "loss": 0.2265,
+      "step": 6375
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.875460261507989,
+      "learning_rate": 6.719801635989209e-07,
+      "loss": 0.1086,
+      "step": 6376
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.21785125615534,
+      "learning_rate": 6.71883187353626e-07,
+      "loss": 0.0764,
+      "step": 6377
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.68719826755472,
+      "learning_rate": 6.717862037754486e-07,
+      "loss": 0.2404,
+      "step": 6378
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.571688675166393,
+      "learning_rate": 6.716892128685264e-07,
+      "loss": 0.1843,
+      "step": 6379
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6185389766513376,
+      "learning_rate": 6.715922146369973e-07,
+      "loss": 0.3114,
+      "step": 6380
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3319873843100709,
+      "learning_rate": 6.714952090849995e-07,
+      "loss": 0.1966,
+      "step": 6381
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2561438784387966,
+      "learning_rate": 6.713981962166712e-07,
+      "loss": 0.1524,
+      "step": 6382
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28233751339411856,
+      "learning_rate": 6.713011760361513e-07,
+      "loss": 0.1737,
+      "step": 6383
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.182331344622677,
+      "learning_rate": 6.712041485475789e-07,
+      "loss": 0.1968,
+      "step": 6384
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.972702489962563,
+      "learning_rate": 6.711071137550934e-07,
+      "loss": 0.2662,
+      "step": 6385
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7964588870798529,
+      "learning_rate": 6.710100716628344e-07,
+      "loss": 0.0531,
+      "step": 6386
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4359035223098955,
+      "learning_rate": 6.709130222749419e-07,
+      "loss": 0.2233,
+      "step": 6387
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8217549713020542,
+      "learning_rate": 6.708159655955563e-07,
+      "loss": 0.4644,
+      "step": 6388
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3280261803432758,
+      "learning_rate": 6.707189016288184e-07,
+      "loss": 0.4257,
+      "step": 6389
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3041206893016233,
+      "learning_rate": 6.706218303788688e-07,
+      "loss": 0.1558,
+      "step": 6390
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7141351038072162,
+      "learning_rate": 6.70524751849849e-07,
+      "loss": 0.4258,
+      "step": 6391
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1079707205512248,
+      "learning_rate": 6.704276660459007e-07,
+      "loss": 0.3083,
+      "step": 6392
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2512500793703196,
+      "learning_rate": 6.703305729711652e-07,
+      "loss": 0.3654,
+      "step": 6393
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8928412086560439,
+      "learning_rate": 6.702334726297852e-07,
+      "loss": 0.0603,
+      "step": 6394
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5169301511702493,
+      "learning_rate": 6.70136365025903e-07,
+      "loss": 0.1779,
+      "step": 6395
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9494093079091375,
+      "learning_rate": 6.700392501636616e-07,
+      "loss": 0.2459,
+      "step": 6396
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3858153397375629,
+      "learning_rate": 6.699421280472037e-07,
+      "loss": 0.1034,
+      "step": 6397
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8402655369890025,
+      "learning_rate": 6.698449986806731e-07,
+      "loss": 0.0212,
+      "step": 6398
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4053758742325924,
+      "learning_rate": 6.697478620682136e-07,
+      "loss": 0.0908,
+      "step": 6399
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6037583497943138,
+      "learning_rate": 6.696507182139689e-07,
+      "loss": 0.0792,
+      "step": 6400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0995171185001902,
+      "learning_rate": 6.695535671220835e-07,
+      "loss": 0.2685,
+      "step": 6401
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7153908915965723,
+      "learning_rate": 6.694564087967022e-07,
+      "loss": 0.1978,
+      "step": 6402
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.480691566042909,
+      "learning_rate": 6.693592432419697e-07,
+      "loss": 0.0978,
+      "step": 6403
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3212146592714469,
+      "learning_rate": 6.692620704620315e-07,
+      "loss": 0.1111,
+      "step": 6404
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9104411249066503,
+      "learning_rate": 6.691648904610331e-07,
+      "loss": 0.223,
+      "step": 6405
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.725987599836831,
+      "learning_rate": 6.690677032431206e-07,
+      "loss": 0.1841,
+      "step": 6406
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.167513301779837,
+      "learning_rate": 6.689705088124397e-07,
+      "loss": 0.0675,
+      "step": 6407
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6746871248555427,
+      "learning_rate": 6.688733071731375e-07,
+      "loss": 0.2058,
+      "step": 6408
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4429761523137734,
+      "learning_rate": 6.687760983293605e-07,
+      "loss": 0.3075,
+      "step": 6409
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8222056307600253,
+      "learning_rate": 6.686788822852557e-07,
+      "loss": 0.2678,
+      "step": 6410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6403125674895418,
+      "learning_rate": 6.685816590449708e-07,
+      "loss": 0.1277,
+      "step": 6411
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.387080088376487,
+      "learning_rate": 6.684844286126534e-07,
+      "loss": 0.175,
+      "step": 6412
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8035832009531326,
+      "learning_rate": 6.683871909924516e-07,
+      "loss": 0.164,
+      "step": 6413
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7419355566774356,
+      "learning_rate": 6.682899461885136e-07,
+      "loss": 0.3492,
+      "step": 6414
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7583044093642305,
+      "learning_rate": 6.681926942049882e-07,
+      "loss": 0.1684,
+      "step": 6415
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.540524158474532,
+      "learning_rate": 6.680954350460247e-07,
+      "loss": 0.2001,
+      "step": 6416
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3589512671052704,
+      "learning_rate": 6.679981687157717e-07,
+      "loss": 0.1495,
+      "step": 6417
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6716513360157987,
+      "learning_rate": 6.679008952183791e-07,
+      "loss": 0.2796,
+      "step": 6418
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2655350732043996,
+      "learning_rate": 6.67803614557997e-07,
+      "loss": 0.1029,
+      "step": 6419
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7685597814453371,
+      "learning_rate": 6.677063267387752e-07,
+      "loss": 0.078,
+      "step": 6420
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.37207951166087705,
+      "learning_rate": 6.676090317648645e-07,
+      "loss": 0.1047,
+      "step": 6421
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6874716827952146,
+      "learning_rate": 6.675117296404155e-07,
+      "loss": 0.2572,
+      "step": 6422
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1641264249149006,
+      "learning_rate": 6.674144203695793e-07,
+      "loss": 0.0321,
+      "step": 6423
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.783936055476026,
+      "learning_rate": 6.673171039565075e-07,
+      "loss": 0.3694,
+      "step": 6424
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.06002758520689,
+      "learning_rate": 6.672197804053515e-07,
+      "loss": 0.3867,
+      "step": 6425
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1685342360811017,
+      "learning_rate": 6.671224497202636e-07,
+      "loss": 0.0897,
+      "step": 6426
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2502365801792226,
+      "learning_rate": 6.670251119053962e-07,
+      "loss": 0.2772,
+      "step": 6427
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8456925694631241,
+      "learning_rate": 6.669277669649017e-07,
+      "loss": 0.2723,
+      "step": 6428
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2186130984426744,
+      "learning_rate": 6.66830414902933e-07,
+      "loss": 0.1497,
+      "step": 6429
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.672158032240156,
+      "learning_rate": 6.667330557236435e-07,
+      "loss": 0.2074,
+      "step": 6430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8898479913339835,
+      "learning_rate": 6.666356894311866e-07,
+      "loss": 0.0757,
+      "step": 6431
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6753965207327566,
+      "learning_rate": 6.665383160297162e-07,
+      "loss": 0.2126,
+      "step": 6432
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6426171782885012,
+      "learning_rate": 6.664409355233867e-07,
+      "loss": 0.2228,
+      "step": 6433
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9163429384997365,
+      "learning_rate": 6.66343547916352e-07,
+      "loss": 0.2849,
+      "step": 6434
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.39900204553834123,
+      "learning_rate": 6.662461532127673e-07,
+      "loss": 0.133,
+      "step": 6435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5645169244933328,
+      "learning_rate": 6.661487514167874e-07,
+      "loss": 0.2357,
+      "step": 6436
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0206422537663962,
+      "learning_rate": 6.660513425325679e-07,
+      "loss": 0.116,
+      "step": 6437
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2466906121473285,
+      "learning_rate": 6.659539265642642e-07,
+      "loss": 0.0162,
+      "step": 6438
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.30870560313402473,
+      "learning_rate": 6.658565035160325e-07,
+      "loss": 0.1962,
+      "step": 6439
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.20218703221726,
+      "learning_rate": 6.657590733920289e-07,
+      "loss": 0.1525,
+      "step": 6440
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.23113188580448643,
+      "learning_rate": 6.656616361964099e-07,
+      "loss": 0.0881,
+      "step": 6441
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.49620995616366453,
+      "learning_rate": 6.655641919333325e-07,
+      "loss": 0.3741,
+      "step": 6442
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5141069284573254,
+      "learning_rate": 6.654667406069539e-07,
+      "loss": 0.0724,
+      "step": 6443
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7985520950661176,
+      "learning_rate": 6.653692822214316e-07,
+      "loss": 0.3261,
+      "step": 6444
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8930237310932487,
+      "learning_rate": 6.652718167809232e-07,
+      "loss": 0.0418,
+      "step": 6445
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8823951379759444,
+      "learning_rate": 6.65174344289587e-07,
+      "loss": 0.2082,
+      "step": 6446
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3241973155351798,
+      "learning_rate": 6.650768647515812e-07,
+      "loss": 0.0664,
+      "step": 6447
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6709408431913578,
+      "learning_rate": 6.649793781710644e-07,
+      "loss": 0.2816,
+      "step": 6448
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7087149931517718,
+      "learning_rate": 6.648818845521955e-07,
+      "loss": 0.2196,
+      "step": 6449
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7660689037529811,
+      "learning_rate": 6.647843838991342e-07,
+      "loss": 0.115,
+      "step": 6450
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.586727923499886,
+      "learning_rate": 6.646868762160398e-07,
+      "loss": 0.255,
+      "step": 6451
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2765990842381686,
+      "learning_rate": 6.645893615070722e-07,
+      "loss": 0.0083,
+      "step": 6452
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7482934174593467,
+      "learning_rate": 6.644918397763914e-07,
+      "loss": 0.2262,
+      "step": 6453
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3505831849518155,
+      "learning_rate": 6.643943110281583e-07,
+      "loss": 0.1289,
+      "step": 6454
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5051971188853868,
+      "learning_rate": 6.642967752665333e-07,
+      "loss": 0.1454,
+      "step": 6455
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.45116601841098775,
+      "learning_rate": 6.641992324956775e-07,
+      "loss": 0.1051,
+      "step": 6456
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.6467710453751,
+      "learning_rate": 6.641016827197526e-07,
+      "loss": 0.1847,
+      "step": 6457
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8112876594939268,
+      "learning_rate": 6.640041259429199e-07,
+      "loss": 0.5393,
+      "step": 6458
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8930198854204855,
+      "learning_rate": 6.639065621693414e-07,
+      "loss": 0.325,
+      "step": 6459
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7377636794694982,
+      "learning_rate": 6.638089914031794e-07,
+      "loss": 0.3848,
+      "step": 6460
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1674386532537577,
+      "learning_rate": 6.637114136485968e-07,
+      "loss": 0.298,
+      "step": 6461
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.598930227126228,
+      "learning_rate": 6.636138289097561e-07,
+      "loss": 0.1786,
+      "step": 6462
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.377681268256108,
+      "learning_rate": 6.635162371908205e-07,
+      "loss": 0.1953,
+      "step": 6463
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.51374160520077,
+      "learning_rate": 6.634186384959536e-07,
+      "loss": 0.1956,
+      "step": 6464
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5770264650314494,
+      "learning_rate": 6.63321032829319e-07,
+      "loss": 0.1165,
+      "step": 6465
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4165259930628278,
+      "learning_rate": 6.632234201950808e-07,
+      "loss": 0.0081,
+      "step": 6466
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5479311939948879,
+      "learning_rate": 6.631258005974034e-07,
+      "loss": 0.1396,
+      "step": 6467
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7595715487242493,
+      "learning_rate": 6.630281740404513e-07,
+      "loss": 0.3016,
+      "step": 6468
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.39108477600863556,
+      "learning_rate": 6.629305405283897e-07,
+      "loss": 0.3183,
+      "step": 6469
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.949277306145099,
+      "learning_rate": 6.628329000653837e-07,
+      "loss": 0.1225,
+      "step": 6470
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7967134536685371,
+      "learning_rate": 6.62735252655599e-07,
+      "loss": 0.2556,
+      "step": 6471
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3221524500716266,
+      "learning_rate": 6.62637598303201e-07,
+      "loss": 0.4512,
+      "step": 6472
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.538877138746729,
+      "learning_rate": 6.625399370123562e-07,
+      "loss": 0.3627,
+      "step": 6473
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.466976161261532,
+      "learning_rate": 6.624422687872311e-07,
+      "loss": 0.0839,
+      "step": 6474
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0441176222594615,
+      "learning_rate": 6.623445936319922e-07,
+      "loss": 0.3385,
+      "step": 6475
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9286328959903735,
+      "learning_rate": 6.622469115508065e-07,
+      "loss": 0.3544,
+      "step": 6476
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8921096068236626,
+      "learning_rate": 6.621492225478413e-07,
+      "loss": 0.217,
+      "step": 6477
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.527601776872973,
+      "learning_rate": 6.620515266272645e-07,
+      "loss": 0.2838,
+      "step": 6478
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5407391774707727,
+      "learning_rate": 6.619538237932437e-07,
+      "loss": 0.0094,
+      "step": 6479
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0682816568887366,
+      "learning_rate": 6.618561140499472e-07,
+      "loss": 0.1509,
+      "step": 6480
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.767621032382869,
+      "learning_rate": 6.617583974015436e-07,
+      "loss": 0.4076,
+      "step": 6481
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8183856655656446,
+      "learning_rate": 6.616606738522016e-07,
+      "loss": 0.2044,
+      "step": 6482
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.29348998718732544,
+      "learning_rate": 6.615629434060902e-07,
+      "loss": 0.0747,
+      "step": 6483
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.316376771796467,
+      "learning_rate": 6.614652060673789e-07,
+      "loss": 0.2296,
+      "step": 6484
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.49404427248597543,
+      "learning_rate": 6.613674618402373e-07,
+      "loss": 0.3097,
+      "step": 6485
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.635776322474302,
+      "learning_rate": 6.612697107288352e-07,
+      "loss": 0.1519,
+      "step": 6486
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4708852073170756,
+      "learning_rate": 6.611719527373433e-07,
+      "loss": 0.0105,
+      "step": 6487
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6560381143197433,
+      "learning_rate": 6.610741878699319e-07,
+      "loss": 0.2706,
+      "step": 6488
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.683823648475233,
+      "learning_rate": 6.609764161307718e-07,
+      "loss": 0.345,
+      "step": 6489
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5789976564007079,
+      "learning_rate": 6.608786375240342e-07,
+      "loss": 0.2959,
+      "step": 6490
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.47281055515844767,
+      "learning_rate": 6.607808520538904e-07,
+      "loss": 0.0035,
+      "step": 6491
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0032031279807376,
+      "learning_rate": 6.606830597245123e-07,
+      "loss": 0.1048,
+      "step": 6492
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6261579156018745,
+      "learning_rate": 6.605852605400719e-07,
+      "loss": 0.1656,
+      "step": 6493
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.732854958715936,
+      "learning_rate": 6.604874545047414e-07,
+      "loss": 0.2638,
+      "step": 6494
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.780507289556038,
+      "learning_rate": 6.603896416226935e-07,
+      "loss": 0.4101,
+      "step": 6495
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3972146105870629,
+      "learning_rate": 6.60291821898101e-07,
+      "loss": 0.2275,
+      "step": 6496
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9831244538642729,
+      "learning_rate": 6.601939953351373e-07,
+      "loss": 0.3621,
+      "step": 6497
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9525721240066384,
+      "learning_rate": 6.600961619379757e-07,
+      "loss": 0.1266,
+      "step": 6498
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.801474755548807,
+      "learning_rate": 6.599983217107899e-07,
+      "loss": 0.0536,
+      "step": 6499
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9934987509011964,
+      "learning_rate": 6.599004746577541e-07,
+      "loss": 0.1469,
+      "step": 6500
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.473780673185841,
+      "learning_rate": 6.598026207830427e-07,
+      "loss": 0.206,
+      "step": 6501
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5678875570797841,
+      "learning_rate": 6.597047600908301e-07,
+      "loss": 0.1189,
+      "step": 6502
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.981066779593897,
+      "learning_rate": 6.596068925852915e-07,
+      "loss": 0.2041,
+      "step": 6503
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9243978928505816,
+      "learning_rate": 6.59509018270602e-07,
+      "loss": 0.303,
+      "step": 6504
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6919267581567077,
+      "learning_rate": 6.594111371509371e-07,
+      "loss": 0.2919,
+      "step": 6505
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7933123534219113,
+      "learning_rate": 6.593132492304727e-07,
+      "loss": 0.3159,
+      "step": 6506
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.729456866562622,
+      "learning_rate": 6.592153545133847e-07,
+      "loss": 0.0478,
+      "step": 6507
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3389561586035512,
+      "learning_rate": 6.591174530038497e-07,
+      "loss": 0.163,
+      "step": 6508
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3966142880288903,
+      "learning_rate": 6.590195447060442e-07,
+      "loss": 0.1249,
+      "step": 6509
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.47325938802164785,
+      "learning_rate": 6.589216296241454e-07,
+      "loss": 0.1079,
+      "step": 6510
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.38768715086358946,
+      "learning_rate": 6.588237077623305e-07,
+      "loss": 0.1644,
+      "step": 6511
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9449971313198207,
+      "learning_rate": 6.587257791247767e-07,
+      "loss": 0.3849,
+      "step": 6512
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4173316413548671,
+      "learning_rate": 6.586278437156621e-07,
+      "loss": 0.1504,
+      "step": 6513
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7588325020050893,
+      "learning_rate": 6.585299015391648e-07,
+      "loss": 0.3681,
+      "step": 6514
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.643512989101004,
+      "learning_rate": 6.584319525994633e-07,
+      "loss": 0.1995,
+      "step": 6515
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.448581609866346,
+      "learning_rate": 6.583339969007363e-07,
+      "loss": 0.1167,
+      "step": 6516
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.179579918401583,
+      "learning_rate": 6.582360344471626e-07,
+      "loss": 0.03,
+      "step": 6517
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9476712906991313,
+      "learning_rate": 6.581380652429215e-07,
+      "loss": 0.0128,
+      "step": 6518
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5618076951985692,
+      "learning_rate": 6.580400892921928e-07,
+      "loss": 0.0394,
+      "step": 6519
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7208144385368301,
+      "learning_rate": 6.579421065991562e-07,
+      "loss": 0.1988,
+      "step": 6520
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6700607999321094,
+      "learning_rate": 6.578441171679916e-07,
+      "loss": 0.1638,
+      "step": 6521
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6036084279517975,
+      "learning_rate": 6.577461210028798e-07,
+      "loss": 0.104,
+      "step": 6522
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2579511343596985,
+      "learning_rate": 6.576481181080014e-07,
+      "loss": 0.2006,
+      "step": 6523
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.995596720419675,
+      "learning_rate": 6.575501084875373e-07,
+      "loss": 0.0842,
+      "step": 6524
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7921197772652121,
+      "learning_rate": 6.574520921456687e-07,
+      "loss": 0.1853,
+      "step": 6525
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8754186202166979,
+      "learning_rate": 6.573540690865777e-07,
+      "loss": 0.2886,
+      "step": 6526
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6026451594831287,
+      "learning_rate": 6.572560393144456e-07,
+      "loss": 0.1946,
+      "step": 6527
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9409002386965092,
+      "learning_rate": 6.571580028334546e-07,
+      "loss": 0.298,
+      "step": 6528
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6088540029934685,
+      "learning_rate": 6.570599596477874e-07,
+      "loss": 0.0904,
+      "step": 6529
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4002230965927323,
+      "learning_rate": 6.569619097616268e-07,
+      "loss": 0.0855,
+      "step": 6530
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3362088504980511,
+      "learning_rate": 6.568638531791554e-07,
+      "loss": 0.0082,
+      "step": 6531
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3475426728193038,
+      "learning_rate": 6.567657899045566e-07,
+      "loss": 0.0606,
+      "step": 6532
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.105330272258604,
+      "learning_rate": 6.566677199420142e-07,
+      "loss": 0.3037,
+      "step": 6533
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7076940168194243,
+      "learning_rate": 6.565696432957119e-07,
+      "loss": 0.2003,
+      "step": 6534
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9592712351857866,
+      "learning_rate": 6.564715599698338e-07,
+      "loss": 0.154,
+      "step": 6535
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0719706917957044,
+      "learning_rate": 6.563734699685646e-07,
+      "loss": 0.3982,
+      "step": 6536
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5285085284835996,
+      "learning_rate": 6.562753732960886e-07,
+      "loss": 0.2912,
+      "step": 6537
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.709800823003546,
+      "learning_rate": 6.56177269956591e-07,
+      "loss": 0.3394,
+      "step": 6538
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8931223092415999,
+      "learning_rate": 6.560791599542572e-07,
+      "loss": 0.1268,
+      "step": 6539
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.20830405696897,
+      "learning_rate": 6.559810432932727e-07,
+      "loss": 0.1853,
+      "step": 6540
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6305902483585847,
+      "learning_rate": 6.558829199778233e-07,
+      "loss": 0.2657,
+      "step": 6541
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7724616563658449,
+      "learning_rate": 6.557847900120952e-07,
+      "loss": 0.2717,
+      "step": 6542
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4791742121706651,
+      "learning_rate": 6.55686653400275e-07,
+      "loss": 0.134,
+      "step": 6543
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9442269367318035,
+      "learning_rate": 6.555885101465489e-07,
+      "loss": 0.2108,
+      "step": 6544
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4039040723449807,
+      "learning_rate": 6.554903602551043e-07,
+      "loss": 0.1719,
+      "step": 6545
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3915708401672107,
+      "learning_rate": 6.553922037301283e-07,
+      "loss": 0.3847,
+      "step": 6546
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.38248881133172413,
+      "learning_rate": 6.552940405758084e-07,
+      "loss": 0.0838,
+      "step": 6547
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5619254010366976,
+      "learning_rate": 6.551958707963328e-07,
+      "loss": 0.226,
+      "step": 6548
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2534025099978723,
+      "learning_rate": 6.550976943958891e-07,
+      "loss": 0.1677,
+      "step": 6549
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.831115820788772,
+      "learning_rate": 6.549995113786662e-07,
+      "loss": 0.1906,
+      "step": 6550
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.0313492870443195,
+      "learning_rate": 6.549013217488525e-07,
+      "loss": 0.2437,
+      "step": 6551
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.123844449278096,
+      "learning_rate": 6.54803125510637e-07,
+      "loss": 0.1735,
+      "step": 6552
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9805224328146542,
+      "learning_rate": 6.547049226682089e-07,
+      "loss": 0.2336,
+      "step": 6553
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6122253035320749,
+      "learning_rate": 6.546067132257579e-07,
+      "loss": 0.1658,
+      "step": 6554
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9216797554993842,
+      "learning_rate": 6.545084971874736e-07,
+      "loss": 0.2078,
+      "step": 6555
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6379449880822752,
+      "learning_rate": 6.544102745575463e-07,
+      "loss": 0.2793,
+      "step": 6556
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3316942329819206,
+      "learning_rate": 6.543120453401664e-07,
+      "loss": 0.0257,
+      "step": 6557
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3965400725147025,
+      "learning_rate": 6.542138095395243e-07,
+      "loss": 0.0984,
+      "step": 6558
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2872385639535027,
+      "learning_rate": 6.541155671598111e-07,
+      "loss": 0.248,
+      "step": 6559
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.246088391524185,
+      "learning_rate": 6.54017318205218e-07,
+      "loss": 0.1558,
+      "step": 6560
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6877516861733453,
+      "learning_rate": 6.539190626799364e-07,
+      "loss": 0.0134,
+      "step": 6561
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.443626408272202,
+      "learning_rate": 6.538208005881583e-07,
+      "loss": 0.1173,
+      "step": 6562
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.273995148271526,
+      "learning_rate": 6.537225319340757e-07,
+      "loss": 0.197,
+      "step": 6563
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5114725362639235,
+      "learning_rate": 6.536242567218807e-07,
+      "loss": 0.2292,
+      "step": 6564
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.243953362163476,
+      "learning_rate": 6.53525974955766e-07,
+      "loss": 0.0133,
+      "step": 6565
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5488081363315258,
+      "learning_rate": 6.534276866399247e-07,
+      "loss": 0.0538,
+      "step": 6566
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9886178392048715,
+      "learning_rate": 6.5332939177855e-07,
+      "loss": 0.2797,
+      "step": 6567
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8601829829661732,
+      "learning_rate": 6.53231090375835e-07,
+      "loss": 0.3236,
+      "step": 6568
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5067175979204344,
+      "learning_rate": 6.531327824359738e-07,
+      "loss": 0.2217,
+      "step": 6569
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.75280071450143,
+      "learning_rate": 6.530344679631602e-07,
+      "loss": 0.2105,
+      "step": 6570
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9963648393536968,
+      "learning_rate": 6.529361469615887e-07,
+      "loss": 0.2076,
+      "step": 6571
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.120861879069957,
+      "learning_rate": 6.528378194354536e-07,
+      "loss": 0.1174,
+      "step": 6572
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4431707215135499,
+      "learning_rate": 6.527394853889499e-07,
+      "loss": 0.0998,
+      "step": 6573
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.09461529127731924,
+      "learning_rate": 6.526411448262726e-07,
+      "loss": 0.005,
+      "step": 6574
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8224445636242497,
+      "learning_rate": 6.525427977516173e-07,
+      "loss": 0.4131,
+      "step": 6575
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6901181559297348,
+      "learning_rate": 6.524444441691795e-07,
+      "loss": 0.3649,
+      "step": 6576
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.35761199002053246,
+      "learning_rate": 6.523460840831554e-07,
+      "loss": 0.2469,
+      "step": 6577
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5334737367261122,
+      "learning_rate": 6.522477174977411e-07,
+      "loss": 0.0843,
+      "step": 6578
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5179009731488441,
+      "learning_rate": 6.52149344417133e-07,
+      "loss": 0.0151,
+      "step": 6579
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6488619418859829,
+      "learning_rate": 6.520509648455282e-07,
+      "loss": 0.2846,
+      "step": 6580
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3682507192636819,
+      "learning_rate": 6.519525787871234e-07,
+      "loss": 0.095,
+      "step": 6581
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1212866152391388,
+      "learning_rate": 6.518541862461162e-07,
+      "loss": 0.0726,
+      "step": 6582
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.54516622159933,
+      "learning_rate": 6.517557872267041e-07,
+      "loss": 0.1133,
+      "step": 6583
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6169215072686566,
+      "learning_rate": 6.516573817330851e-07,
+      "loss": 0.1888,
+      "step": 6584
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6833676158879121,
+      "learning_rate": 6.515589697694574e-07,
+      "loss": 0.1507,
+      "step": 6585
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5188710826658951,
+      "learning_rate": 6.514605513400193e-07,
+      "loss": 0.4211,
+      "step": 6586
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3199304249283735,
+      "learning_rate": 6.513621264489696e-07,
+      "loss": 0.2161,
+      "step": 6587
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9024744484290665,
+      "learning_rate": 6.512636951005073e-07,
+      "loss": 0.2223,
+      "step": 6588
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2358497207831372,
+      "learning_rate": 6.511652572988316e-07,
+      "loss": 0.2116,
+      "step": 6589
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.790714556211648,
+      "learning_rate": 6.510668130481423e-07,
+      "loss": 0.2035,
+      "step": 6590
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5837610043799196,
+      "learning_rate": 6.50968362352639e-07,
+      "loss": 0.2612,
+      "step": 6591
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1322110048786116,
+      "learning_rate": 6.508699052165218e-07,
+      "loss": 0.0731,
+      "step": 6592
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4869642528488591,
+      "learning_rate": 6.507714416439914e-07,
+      "loss": 0.1773,
+      "step": 6593
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.46931346360923487,
+      "learning_rate": 6.506729716392479e-07,
+      "loss": 0.1304,
+      "step": 6594
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6495464461000598,
+      "learning_rate": 6.505744952064927e-07,
+      "loss": 0.211,
+      "step": 6595
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6816189436042995,
+      "learning_rate": 6.504760123499269e-07,
+      "loss": 0.2026,
+      "step": 6596
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5359782743913805,
+      "learning_rate": 6.503775230737518e-07,
+      "loss": 0.1804,
+      "step": 6597
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6466741031098954,
+      "learning_rate": 6.502790273821694e-07,
+      "loss": 0.2872,
+      "step": 6598
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.059231198735174,
+      "learning_rate": 6.501805252793817e-07,
+      "loss": 0.2318,
+      "step": 6599
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1495449829824636,
+      "learning_rate": 6.500820167695905e-07,
+      "loss": 0.0118,
+      "step": 6600
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6556357119339661,
+      "learning_rate": 6.49983501856999e-07,
+      "loss": 0.2049,
+      "step": 6601
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.19084012033555592,
+      "learning_rate": 6.498849805458098e-07,
+      "loss": 0.0892,
+      "step": 6602
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7542708776617518,
+      "learning_rate": 6.49786452840226e-07,
+      "loss": 0.0684,
+      "step": 6603
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.130488531154757,
+      "learning_rate": 6.496879187444509e-07,
+      "loss": 0.0361,
+      "step": 6604
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.12030908783750106,
+      "learning_rate": 6.495893782626884e-07,
+      "loss": 0.0038,
+      "step": 6605
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.472643673466841,
+      "learning_rate": 6.494908313991424e-07,
+      "loss": 0.027,
+      "step": 6606
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.756578264745721,
+      "learning_rate": 6.493922781580169e-07,
+      "loss": 0.1752,
+      "step": 6607
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.47065007556631094,
+      "learning_rate": 6.492937185435165e-07,
+      "loss": 0.0365,
+      "step": 6608
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.37773434203172523,
+      "learning_rate": 6.491951525598461e-07,
+      "loss": 0.1043,
+      "step": 6609
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.170231038508277,
+      "learning_rate": 6.490965802112103e-07,
+      "loss": 0.0348,
+      "step": 6610
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2153154323505864,
+      "learning_rate": 6.489980015018147e-07,
+      "loss": 0.2384,
+      "step": 6611
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1887446464012499,
+      "learning_rate": 6.488994164358651e-07,
+      "loss": 0.4168,
+      "step": 6612
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5097010072238237,
+      "learning_rate": 6.488008250175669e-07,
+      "loss": 0.0474,
+      "step": 6613
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7804920096956791,
+      "learning_rate": 6.487022272511264e-07,
+      "loss": 0.1254,
+      "step": 6614
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5381319677601095,
+      "learning_rate": 6.486036231407499e-07,
+      "loss": 0.0628,
+      "step": 6615
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3956450165002046,
+      "learning_rate": 6.485050126906442e-07,
+      "loss": 0.1665,
+      "step": 6616
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6006757979680589,
+      "learning_rate": 6.48406395905016e-07,
+      "loss": 0.1678,
+      "step": 6617
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5749528585225581,
+      "learning_rate": 6.483077727880726e-07,
+      "loss": 0.2657,
+      "step": 6618
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.546013603761911,
+      "learning_rate": 6.482091433440215e-07,
+      "loss": 0.1231,
+      "step": 6619
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6563493718514495,
+      "learning_rate": 6.481105075770705e-07,
+      "loss": 0.3289,
+      "step": 6620
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3557252399707117,
+      "learning_rate": 6.480118654914275e-07,
+      "loss": 0.2553,
+      "step": 6621
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3836554168728923,
+      "learning_rate": 6.479132170913009e-07,
+      "loss": 0.2631,
+      "step": 6622
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.49305934072685575,
+      "learning_rate": 6.478145623808988e-07,
+      "loss": 0.2529,
+      "step": 6623
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9065870549860774,
+      "learning_rate": 6.477159013644306e-07,
+      "loss": 0.2566,
+      "step": 6624
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.220369398802296,
+      "learning_rate": 6.476172340461051e-07,
+      "loss": 0.3382,
+      "step": 6625
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5111977616318292,
+      "learning_rate": 6.475185604301314e-07,
+      "loss": 0.1724,
+      "step": 6626
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4111482409555583,
+      "learning_rate": 6.474198805207196e-07,
+      "loss": 0.1109,
+      "step": 6627
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.147696513130137,
+      "learning_rate": 6.473211943220792e-07,
+      "loss": 0.1227,
+      "step": 6628
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9306742720330724,
+      "learning_rate": 6.472225018384205e-07,
+      "loss": 0.2259,
+      "step": 6629
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26695730537907403,
+      "learning_rate": 6.471238030739541e-07,
+      "loss": 0.191,
+      "step": 6630
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1497587790002861,
+      "learning_rate": 6.470250980328903e-07,
+      "loss": 0.3335,
+      "step": 6631
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5381242770023748,
+      "learning_rate": 6.469263867194404e-07,
+      "loss": 0.0837,
+      "step": 6632
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1718839718459892,
+      "learning_rate": 6.468276691378154e-07,
+      "loss": 0.2356,
+      "step": 6633
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0734400553710504,
+      "learning_rate": 6.467289452922268e-07,
+      "loss": 0.229,
+      "step": 6634
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4721460541689104,
+      "learning_rate": 6.466302151868865e-07,
+      "loss": 0.3775,
+      "step": 6635
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.47354152135182404,
+      "learning_rate": 6.465314788260065e-07,
+      "loss": 0.3041,
+      "step": 6636
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.213183190688514,
+      "learning_rate": 6.46432736213799e-07,
+      "loss": 0.0726,
+      "step": 6637
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.691683528170075,
+      "learning_rate": 6.463339873544766e-07,
+      "loss": 0.0671,
+      "step": 6638
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7382855304101535,
+      "learning_rate": 6.462352322522523e-07,
+      "loss": 0.3539,
+      "step": 6639
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0412966731957,
+      "learning_rate": 6.461364709113389e-07,
+      "loss": 0.1656,
+      "step": 6640
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9203636723708595,
+      "learning_rate": 6.460377033359499e-07,
+      "loss": 0.2078,
+      "step": 6641
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5885890438419307,
+      "learning_rate": 6.459389295302989e-07,
+      "loss": 0.0815,
+      "step": 6642
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7677984394111165,
+      "learning_rate": 6.458401494985997e-07,
+      "loss": 0.3011,
+      "step": 6643
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2514408018419238,
+      "learning_rate": 6.457413632450666e-07,
+      "loss": 0.0802,
+      "step": 6644
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9141602880990805,
+      "learning_rate": 6.45642570773914e-07,
+      "loss": 0.2407,
+      "step": 6645
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0567396478833757,
+      "learning_rate": 6.455437720893564e-07,
+      "loss": 0.3146,
+      "step": 6646
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7719336200227167,
+      "learning_rate": 6.454449671956091e-07,
+      "loss": 0.1949,
+      "step": 6647
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.799251195275632,
+      "learning_rate": 6.45346156096887e-07,
+      "loss": 0.296,
+      "step": 6648
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.86904514165989,
+      "learning_rate": 6.452473387974058e-07,
+      "loss": 0.368,
+      "step": 6649
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5941697259217461,
+      "learning_rate": 6.45148515301381e-07,
+      "loss": 0.2574,
+      "step": 6650
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.471621454658878,
+      "learning_rate": 6.45049685613029e-07,
+      "loss": 0.1181,
+      "step": 6651
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5072800655176259,
+      "learning_rate": 6.449508497365656e-07,
+      "loss": 0.3983,
+      "step": 6652
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.44085371993939076,
+      "learning_rate": 6.448520076762076e-07,
+      "loss": 0.0789,
+      "step": 6653
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.275489586224218,
+      "learning_rate": 6.447531594361719e-07,
+      "loss": 0.0508,
+      "step": 6654
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.630060685326117,
+      "learning_rate": 6.446543050206752e-07,
+      "loss": 0.1589,
+      "step": 6655
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2328741430616827,
+      "learning_rate": 6.445554444339352e-07,
+      "loss": 0.0978,
+      "step": 6656
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5372530982254144,
+      "learning_rate": 6.444565776801693e-07,
+      "loss": 0.1315,
+      "step": 6657
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7599789551403485,
+      "learning_rate": 6.443577047635956e-07,
+      "loss": 0.1224,
+      "step": 6658
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.0296851642204246,
+      "learning_rate": 6.442588256884318e-07,
+      "loss": 0.2358,
+      "step": 6659
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.108297544744153,
+      "learning_rate": 6.441599404588966e-07,
+      "loss": 0.12,
+      "step": 6660
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.501292539605557,
+      "learning_rate": 6.440610490792084e-07,
+      "loss": 0.2431,
+      "step": 6661
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.34171415167564934,
+      "learning_rate": 6.439621515535863e-07,
+      "loss": 0.1894,
+      "step": 6662
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.682733308349673,
+      "learning_rate": 6.438632478862494e-07,
+      "loss": 0.3229,
+      "step": 6663
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.390017699026032,
+      "learning_rate": 6.437643380814171e-07,
+      "loss": 0.3022,
+      "step": 6664
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8681554634213695,
+      "learning_rate": 6.436654221433093e-07,
+      "loss": 0.0442,
+      "step": 6665
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.541158370693776,
+      "learning_rate": 6.435665000761458e-07,
+      "loss": 0.2862,
+      "step": 6666
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.390095187041583,
+      "learning_rate": 6.434675718841468e-07,
+      "loss": 0.2874,
+      "step": 6667
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4536828965465176,
+      "learning_rate": 6.433686375715327e-07,
+      "loss": 0.2102,
+      "step": 6668
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1446732248591567,
+      "learning_rate": 6.432696971425243e-07,
+      "loss": 0.2566,
+      "step": 6669
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.757703886679042,
+      "learning_rate": 6.431707506013426e-07,
+      "loss": 0.2267,
+      "step": 6670
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4140671835415346,
+      "learning_rate": 6.430717979522088e-07,
+      "loss": 0.2333,
+      "step": 6671
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6067960560052531,
+      "learning_rate": 6.429728391993445e-07,
+      "loss": 0.2523,
+      "step": 6672
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7960205313213627,
+      "learning_rate": 6.428738743469717e-07,
+      "loss": 0.2754,
+      "step": 6673
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7451427062897477,
+      "learning_rate": 6.427749033993119e-07,
+      "loss": 0.228,
+      "step": 6674
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0022757694742077,
+      "learning_rate": 6.42675926360588e-07,
+      "loss": 0.4864,
+      "step": 6675
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1426768082927397,
+      "learning_rate": 6.425769432350221e-07,
+      "loss": 0.1139,
+      "step": 6676
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.124319062949054,
+      "learning_rate": 6.424779540268372e-07,
+      "loss": 0.1503,
+      "step": 6677
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2308310821855686,
+      "learning_rate": 6.423789587402564e-07,
+      "loss": 0.1021,
+      "step": 6678
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5066897111623873,
+      "learning_rate": 6.422799573795031e-07,
+      "loss": 0.3278,
+      "step": 6679
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5059648558799121,
+      "learning_rate": 6.421809499488006e-07,
+      "loss": 0.1839,
+      "step": 6680
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27275453426585833,
+      "learning_rate": 6.420819364523731e-07,
+      "loss": 0.0828,
+      "step": 6681
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5761906999379414,
+      "learning_rate": 6.419829168944444e-07,
+      "loss": 0.1035,
+      "step": 6682
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7912710360378307,
+      "learning_rate": 6.418838912792393e-07,
+      "loss": 0.0861,
+      "step": 6683
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.572790723941205,
+      "learning_rate": 6.41784859610982e-07,
+      "loss": 0.1856,
+      "step": 6684
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2006324218480453,
+      "learning_rate": 6.416858218938975e-07,
+      "loss": 0.199,
+      "step": 6685
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.181230130596287,
+      "learning_rate": 6.415867781322112e-07,
+      "loss": 0.3226,
+      "step": 6686
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9367056371259241,
+      "learning_rate": 6.414877283301482e-07,
+      "loss": 0.3043,
+      "step": 6687
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.780971038545777,
+      "learning_rate": 6.413886724919343e-07,
+      "loss": 0.3924,
+      "step": 6688
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1464454703347422,
+      "learning_rate": 6.412896106217955e-07,
+      "loss": 0.0669,
+      "step": 6689
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6590059178969544,
+      "learning_rate": 6.411905427239577e-07,
+      "loss": 0.3126,
+      "step": 6690
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5853783605081505,
+      "learning_rate": 6.410914688026475e-07,
+      "loss": 0.2109,
+      "step": 6691
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1179016763262029,
+      "learning_rate": 6.409923888620918e-07,
+      "loss": 0.2518,
+      "step": 6692
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.581686640631824,
+      "learning_rate": 6.408933029065173e-07,
+      "loss": 0.2171,
+      "step": 6693
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8759997525338321,
+      "learning_rate": 6.407942109401514e-07,
+      "loss": 0.141,
+      "step": 6694
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5080766936981634,
+      "learning_rate": 6.406951129672212e-07,
+      "loss": 0.2806,
+      "step": 6695
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0399523992915547,
+      "learning_rate": 6.405960089919548e-07,
+      "loss": 0.0772,
+      "step": 6696
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6478695764754306,
+      "learning_rate": 6.404968990185799e-07,
+      "loss": 0.0619,
+      "step": 6697
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.273136086090483,
+      "learning_rate": 6.403977830513248e-07,
+      "loss": 0.2038,
+      "step": 6698
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7704261948601646,
+      "learning_rate": 6.402986610944182e-07,
+      "loss": 0.2672,
+      "step": 6699
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0132380244586805,
+      "learning_rate": 6.401995331520886e-07,
+      "loss": 0.1967,
+      "step": 6700
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9905749152200145,
+      "learning_rate": 6.401003992285652e-07,
+      "loss": 0.0433,
+      "step": 6701
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4633532929933657,
+      "learning_rate": 6.400012593280771e-07,
+      "loss": 0.1597,
+      "step": 6702
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9126576229653443,
+      "learning_rate": 6.399021134548537e-07,
+      "loss": 0.0904,
+      "step": 6703
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7200646969220672,
+      "learning_rate": 6.39802961613125e-07,
+      "loss": 0.3084,
+      "step": 6704
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1758572937507459,
+      "learning_rate": 6.39703803807121e-07,
+      "loss": 0.0533,
+      "step": 6705
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4498625526282167,
+      "learning_rate": 6.396046400410718e-07,
+      "loss": 0.0553,
+      "step": 6706
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6364404529115247,
+      "learning_rate": 6.39505470319208e-07,
+      "loss": 0.226,
+      "step": 6707
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7772069739561727,
+      "learning_rate": 6.394062946457604e-07,
+      "loss": 0.4006,
+      "step": 6708
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6028916535151102,
+      "learning_rate": 6.3930711302496e-07,
+      "loss": 0.3125,
+      "step": 6709
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6758249587863409,
+      "learning_rate": 6.39207925461038e-07,
+      "loss": 0.4262,
+      "step": 6710
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.127001295949069,
+      "learning_rate": 6.391087319582263e-07,
+      "loss": 0.2229,
+      "step": 6711
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3262722172903861,
+      "learning_rate": 6.390095325207564e-07,
+      "loss": 0.0041,
+      "step": 6712
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7507862581580848,
+      "learning_rate": 6.389103271528605e-07,
+      "loss": 0.145,
+      "step": 6713
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6605094691567882,
+      "learning_rate": 6.388111158587706e-07,
+      "loss": 0.2998,
+      "step": 6714
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5563965705038187,
+      "learning_rate": 6.387118986427195e-07,
+      "loss": 0.0762,
+      "step": 6715
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8925033905398138,
+      "learning_rate": 6.386126755089398e-07,
+      "loss": 0.2854,
+      "step": 6716
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.237973136924914,
+      "learning_rate": 6.385134464616648e-07,
+      "loss": 0.0446,
+      "step": 6717
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5150703661256893,
+      "learning_rate": 6.384142115051279e-07,
+      "loss": 0.1314,
+      "step": 6718
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4835942761407444,
+      "learning_rate": 6.383149706435625e-07,
+      "loss": 0.1088,
+      "step": 6719
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6085024289437382,
+      "learning_rate": 6.382157238812023e-07,
+      "loss": 0.1894,
+      "step": 6720
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6893796463969709,
+      "learning_rate": 6.381164712222814e-07,
+      "loss": 0.1025,
+      "step": 6721
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3454877105562411,
+      "learning_rate": 6.380172126710344e-07,
+      "loss": 0.0908,
+      "step": 6722
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4616723698580447,
+      "learning_rate": 6.379179482316954e-07,
+      "loss": 0.2442,
+      "step": 6723
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9968537970426457,
+      "learning_rate": 6.378186779084995e-07,
+      "loss": 0.3062,
+      "step": 6724
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6244526051281594,
+      "learning_rate": 6.377194017056819e-07,
+      "loss": 0.2975,
+      "step": 6725
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7509501104222818,
+      "learning_rate": 6.376201196274777e-07,
+      "loss": 0.1483,
+      "step": 6726
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3151371133817729,
+      "learning_rate": 6.375208316781226e-07,
+      "loss": 0.1096,
+      "step": 6727
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3383792185360287,
+      "learning_rate": 6.374215378618523e-07,
+      "loss": 0.0132,
+      "step": 6728
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6491694219196131,
+      "learning_rate": 6.373222381829031e-07,
+      "loss": 0.1896,
+      "step": 6729
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3156566306106663,
+      "learning_rate": 6.37222932645511e-07,
+      "loss": 0.2652,
+      "step": 6730
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8979187862266332,
+      "learning_rate": 6.371236212539129e-07,
+      "loss": 0.2058,
+      "step": 6731
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.278954544093297,
+      "learning_rate": 6.370243040123452e-07,
+      "loss": 0.3163,
+      "step": 6732
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6705358804955086,
+      "learning_rate": 6.369249809250454e-07,
+      "loss": 0.2729,
+      "step": 6733
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.706961289716489,
+      "learning_rate": 6.368256519962506e-07,
+      "loss": 0.0283,
+      "step": 6734
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1270196563346673,
+      "learning_rate": 6.367263172301984e-07,
+      "loss": 0.043,
+      "step": 6735
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27520589895700176,
+      "learning_rate": 6.366269766311269e-07,
+      "loss": 0.1172,
+      "step": 6736
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6130876314593725,
+      "learning_rate": 6.365276302032737e-07,
+      "loss": 0.4631,
+      "step": 6737
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9158684740988248,
+      "learning_rate": 6.364282779508774e-07,
+      "loss": 0.0629,
+      "step": 6738
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5385877544612692,
+      "learning_rate": 6.363289198781765e-07,
+      "loss": 0.1979,
+      "step": 6739
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7091496753358872,
+      "learning_rate": 6.362295559894099e-07,
+      "loss": 0.1839,
+      "step": 6740
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9930991041229553,
+      "learning_rate": 6.361301862888164e-07,
+      "loss": 0.3691,
+      "step": 6741
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5466540366244256,
+      "learning_rate": 6.360308107806357e-07,
+      "loss": 0.1623,
+      "step": 6742
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.13058829593568,
+      "learning_rate": 6.35931429469107e-07,
+      "loss": 0.266,
+      "step": 6743
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8362409781450588,
+      "learning_rate": 6.358320423584704e-07,
+      "loss": 0.3643,
+      "step": 6744
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8712235975431387,
+      "learning_rate": 6.357326494529657e-07,
+      "loss": 0.2947,
+      "step": 6745
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.790006196204553,
+      "learning_rate": 6.356332507568333e-07,
+      "loss": 0.1578,
+      "step": 6746
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5730066080214747,
+      "learning_rate": 6.355338462743138e-07,
+      "loss": 0.1548,
+      "step": 6747
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.454384903442159,
+      "learning_rate": 6.35434436009648e-07,
+      "loss": 0.3031,
+      "step": 6748
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6514055731437256,
+      "learning_rate": 6.353350199670771e-07,
+      "loss": 0.1687,
+      "step": 6749
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18309743262848088,
+      "learning_rate": 6.352355981508419e-07,
+      "loss": 0.1015,
+      "step": 6750
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8656871711227753,
+      "learning_rate": 6.351361705651842e-07,
+      "loss": 0.3523,
+      "step": 6751
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8071333822294513,
+      "learning_rate": 6.350367372143459e-07,
+      "loss": 0.261,
+      "step": 6752
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6646761374076501,
+      "learning_rate": 6.34937298102569e-07,
+      "loss": 0.2523,
+      "step": 6753
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5172036440857891,
+      "learning_rate": 6.348378532340957e-07,
+      "loss": 0.4282,
+      "step": 6754
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8127899949717996,
+      "learning_rate": 6.347384026131683e-07,
+      "loss": 0.2521,
+      "step": 6755
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1329997645487928,
+      "learning_rate": 6.3463894624403e-07,
+      "loss": 0.2012,
+      "step": 6756
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.088470560050196,
+      "learning_rate": 6.345394841309237e-07,
+      "loss": 0.0251,
+      "step": 6757
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8617685245678508,
+      "learning_rate": 6.344400162780923e-07,
+      "loss": 0.0105,
+      "step": 6758
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5893830867155143,
+      "learning_rate": 6.343405426897797e-07,
+      "loss": 0.2246,
+      "step": 6759
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7666749117852034,
+      "learning_rate": 6.342410633702294e-07,
+      "loss": 0.1976,
+      "step": 6760
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.43331517845402123,
+      "learning_rate": 6.341415783236854e-07,
+      "loss": 0.1695,
+      "step": 6761
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9868818836290725,
+      "learning_rate": 6.340420875543921e-07,
+      "loss": 0.0934,
+      "step": 6762
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6897306726627255,
+      "learning_rate": 6.33942591066594e-07,
+      "loss": 0.2653,
+      "step": 6763
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.0389408405008735,
+      "learning_rate": 6.338430888645356e-07,
+      "loss": 0.0828,
+      "step": 6764
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5074205963952,
+      "learning_rate": 6.33743580952462e-07,
+      "loss": 0.2088,
+      "step": 6765
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.47124719780596425,
+      "learning_rate": 6.336440673346184e-07,
+      "loss": 0.1126,
+      "step": 6766
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4583817460931665,
+      "learning_rate": 6.335445480152503e-07,
+      "loss": 0.1736,
+      "step": 6767
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3462864839528916,
+      "learning_rate": 6.334450229986031e-07,
+      "loss": 0.1249,
+      "step": 6768
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.36479249709676476,
+      "learning_rate": 6.33345492288923e-07,
+      "loss": 0.0074,
+      "step": 6769
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2588922594358886,
+      "learning_rate": 6.332459558904563e-07,
+      "loss": 0.1877,
+      "step": 6770
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.899914125571852,
+      "learning_rate": 6.331464138074491e-07,
+      "loss": 0.3181,
+      "step": 6771
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.934832204196358,
+      "learning_rate": 6.330468660441484e-07,
+      "loss": 0.2004,
+      "step": 6772
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9039667094154356,
+      "learning_rate": 6.329473126048008e-07,
+      "loss": 0.0988,
+      "step": 6773
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1220408080787108,
+      "learning_rate": 6.328477534936537e-07,
+      "loss": 0.2913,
+      "step": 6774
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7439464280070911,
+      "learning_rate": 6.327481887149542e-07,
+      "loss": 0.2032,
+      "step": 6775
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2175894949275574,
+      "learning_rate": 6.326486182729504e-07,
+      "loss": 0.217,
+      "step": 6776
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6617656036990464,
+      "learning_rate": 6.325490421718897e-07,
+      "loss": 0.0769,
+      "step": 6777
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6548727496560991,
+      "learning_rate": 6.324494604160205e-07,
+      "loss": 0.1635,
+      "step": 6778
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1961538380748058,
+      "learning_rate": 6.323498730095909e-07,
+      "loss": 0.34,
+      "step": 6779
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7573309400627748,
+      "learning_rate": 6.322502799568496e-07,
+      "loss": 0.1469,
+      "step": 6780
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4620618967612187,
+      "learning_rate": 6.321506812620457e-07,
+      "loss": 0.1377,
+      "step": 6781
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9670347054280445,
+      "learning_rate": 6.32051076929428e-07,
+      "loss": 0.2661,
+      "step": 6782
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6153038264053357,
+      "learning_rate": 6.319514669632459e-07,
+      "loss": 0.1095,
+      "step": 6783
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1576094055504325,
+      "learning_rate": 6.318518513677491e-07,
+      "loss": 0.2552,
+      "step": 6784
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3510418092236915,
+      "learning_rate": 6.317522301471872e-07,
+      "loss": 0.2902,
+      "step": 6785
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.609354472764255,
+      "learning_rate": 6.316526033058103e-07,
+      "loss": 0.2841,
+      "step": 6786
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3021652967977673,
+      "learning_rate": 6.315529708478685e-07,
+      "loss": 0.1702,
+      "step": 6787
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9888446225038058,
+      "learning_rate": 6.314533327776126e-07,
+      "loss": 0.2604,
+      "step": 6788
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0346951448584885,
+      "learning_rate": 6.313536890992935e-07,
+      "loss": 0.4024,
+      "step": 6789
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5675104587584527,
+      "learning_rate": 6.312540398171617e-07,
+      "loss": 0.0111,
+      "step": 6790
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5672563634103571,
+      "learning_rate": 6.311543849354689e-07,
+      "loss": 0.327,
+      "step": 6791
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.443557818405537,
+      "learning_rate": 6.310547244584663e-07,
+      "loss": 0.0626,
+      "step": 6792
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3011797489014841,
+      "learning_rate": 6.309550583904057e-07,
+      "loss": 0.1006,
+      "step": 6793
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.246214236196623,
+      "learning_rate": 6.308553867355391e-07,
+      "loss": 0.1736,
+      "step": 6794
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.446954270409426,
+      "learning_rate": 6.307557094981184e-07,
+      "loss": 0.1546,
+      "step": 6795
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7072287569490263,
+      "learning_rate": 6.306560266823966e-07,
+      "loss": 0.1176,
+      "step": 6796
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9312694871096633,
+      "learning_rate": 6.305563382926259e-07,
+      "loss": 0.1041,
+      "step": 6797
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1974276895069347,
+      "learning_rate": 6.304566443330594e-07,
+      "loss": 0.1647,
+      "step": 6798
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.719215210065318,
+      "learning_rate": 6.303569448079502e-07,
+      "loss": 0.1296,
+      "step": 6799
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.842410190704364,
+      "learning_rate": 6.302572397215516e-07,
+      "loss": 0.0253,
+      "step": 6800
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30814150918820393,
+      "learning_rate": 6.301575290781174e-07,
+      "loss": 0.0688,
+      "step": 6801
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6158718607868237,
+      "learning_rate": 6.300578128819015e-07,
+      "loss": 0.2403,
+      "step": 6802
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4780560018664006,
+      "learning_rate": 6.299580911371576e-07,
+      "loss": 0.1857,
+      "step": 6803
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9397332241649011,
+      "learning_rate": 6.298583638481403e-07,
+      "loss": 0.2389,
+      "step": 6804
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.923574127674357,
+      "learning_rate": 6.297586310191043e-07,
+      "loss": 0.0653,
+      "step": 6805
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5311185850515917,
+      "learning_rate": 6.296588926543042e-07,
+      "loss": 0.295,
+      "step": 6806
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6618138251404375,
+      "learning_rate": 6.29559148757995e-07,
+      "loss": 0.2583,
+      "step": 6807
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7704545834260043,
+      "learning_rate": 6.294593993344322e-07,
+      "loss": 0.2924,
+      "step": 6808
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.225263236600781,
+      "learning_rate": 6.29359644387871e-07,
+      "loss": 0.218,
+      "step": 6809
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5587407216636802,
+      "learning_rate": 6.292598839225674e-07,
+      "loss": 0.3032,
+      "step": 6810
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0127580991201515,
+      "learning_rate": 6.291601179427774e-07,
+      "loss": 0.3185,
+      "step": 6811
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.511789852653221,
+      "learning_rate": 6.290603464527571e-07,
+      "loss": 0.0501,
+      "step": 6812
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.280016632547844,
+      "learning_rate": 6.28960569456763e-07,
+      "loss": 0.2233,
+      "step": 6813
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7443266284461287,
+      "learning_rate": 6.288607869590517e-07,
+      "loss": 0.2224,
+      "step": 6814
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1274665943383784,
+      "learning_rate": 6.2876099896388e-07,
+      "loss": 0.3323,
+      "step": 6815
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5260091674169767,
+      "learning_rate": 6.286612054755054e-07,
+      "loss": 0.1671,
+      "step": 6816
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6868699193481335,
+      "learning_rate": 6.285614064981853e-07,
+      "loss": 0.02,
+      "step": 6817
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.732733327052659,
+      "learning_rate": 6.284616020361771e-07,
+      "loss": 0.0463,
+      "step": 6818
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.21229347455748196,
+      "learning_rate": 6.283617920937388e-07,
+      "loss": 0.0889,
+      "step": 6819
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7182442012876338,
+      "learning_rate": 6.282619766751282e-07,
+      "loss": 0.2928,
+      "step": 6820
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2240411603638657,
+      "learning_rate": 6.281621557846039e-07,
+      "loss": 0.1031,
+      "step": 6821
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.49941767733144304,
+      "learning_rate": 6.280623294264242e-07,
+      "loss": 0.2916,
+      "step": 6822
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4504738802027255,
+      "learning_rate": 6.279624976048483e-07,
+      "loss": 0.3901,
+      "step": 6823
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8574063668748844,
+      "learning_rate": 6.278626603241351e-07,
+      "loss": 0.1969,
+      "step": 6824
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6650372944583962,
+      "learning_rate": 6.277628175885435e-07,
+      "loss": 0.1371,
+      "step": 6825
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8725049317109095,
+      "learning_rate": 6.276629694023335e-07,
+      "loss": 0.0056,
+      "step": 6826
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7852613308837884,
+      "learning_rate": 6.275631157697646e-07,
+      "loss": 0.1149,
+      "step": 6827
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5960671203452759,
+      "learning_rate": 6.274632566950966e-07,
+      "loss": 0.2283,
+      "step": 6828
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3019755436344658,
+      "learning_rate": 6.273633921825899e-07,
+      "loss": 0.3012,
+      "step": 6829
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.425201121885749,
+      "learning_rate": 6.272635222365049e-07,
+      "loss": 0.2481,
+      "step": 6830
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8176663616259935,
+      "learning_rate": 6.271636468611021e-07,
+      "loss": 0.1553,
+      "step": 6831
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6115331284522859,
+      "learning_rate": 6.270637660606426e-07,
+      "loss": 0.0967,
+      "step": 6832
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9737249811992744,
+      "learning_rate": 6.269638798393875e-07,
+      "loss": 0.281,
+      "step": 6833
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 18.043001773579363,
+      "learning_rate": 6.268639882015978e-07,
+      "loss": 0.1687,
+      "step": 6834
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6132946727237054,
+      "learning_rate": 6.267640911515355e-07,
+      "loss": 0.2331,
+      "step": 6835
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5091497400318866,
+      "learning_rate": 6.266641886934622e-07,
+      "loss": 0.1694,
+      "step": 6836
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3970003631326264,
+      "learning_rate": 6.2656428083164e-07,
+      "loss": 0.3293,
+      "step": 6837
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.505052794904146,
+      "learning_rate": 6.264643675703312e-07,
+      "loss": 0.226,
+      "step": 6838
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.9167611183008,
+      "learning_rate": 6.263644489137982e-07,
+      "loss": 0.1118,
+      "step": 6839
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0811616980231809,
+      "learning_rate": 6.262645248663037e-07,
+      "loss": 0.2247,
+      "step": 6840
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6414905323992222,
+      "learning_rate": 6.261645954321108e-07,
+      "loss": 0.1955,
+      "step": 6841
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3250585746642969,
+      "learning_rate": 6.260646606154827e-07,
+      "loss": 0.1739,
+      "step": 6842
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7866168030369379,
+      "learning_rate": 6.259647204206827e-07,
+      "loss": 0.3972,
+      "step": 6843
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5581515988682273,
+      "learning_rate": 6.258647748519746e-07,
+      "loss": 0.2474,
+      "step": 6844
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7993384137029356,
+      "learning_rate": 6.257648239136221e-07,
+      "loss": 0.3265,
+      "step": 6845
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8276376844992642,
+      "learning_rate": 6.256648676098895e-07,
+      "loss": 0.2602,
+      "step": 6846
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3657207400181275,
+      "learning_rate": 6.25564905945041e-07,
+      "loss": 0.3829,
+      "step": 6847
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4485079253704959,
+      "learning_rate": 6.254649389233409e-07,
+      "loss": 0.176,
+      "step": 6848
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.43576652917857184,
+      "learning_rate": 6.253649665490545e-07,
+      "loss": 0.0893,
+      "step": 6849
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5289305109863717,
+      "learning_rate": 6.252649888264468e-07,
+      "loss": 0.1441,
+      "step": 6850
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.469843352456335,
+      "learning_rate": 6.251650057597826e-07,
+      "loss": 0.1063,
+      "step": 6851
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1768701925903495,
+      "learning_rate": 6.250650173533278e-07,
+      "loss": 0.2213,
+      "step": 6852
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9943437390577592,
+      "learning_rate": 6.24965023611348e-07,
+      "loss": 0.3129,
+      "step": 6853
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.057901605441251,
+      "learning_rate": 6.248650245381088e-07,
+      "loss": 0.0195,
+      "step": 6854
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7806514200502783,
+      "learning_rate": 6.247650201378769e-07,
+      "loss": 0.2214,
+      "step": 6855
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5671469742361622,
+      "learning_rate": 6.246650104149184e-07,
+      "loss": 0.2613,
+      "step": 6856
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.373765136142397,
+      "learning_rate": 6.245649953734998e-07,
+      "loss": 0.1473,
+      "step": 6857
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4461208500038242,
+      "learning_rate": 6.244649750178882e-07,
+      "loss": 0.2095,
+      "step": 6858
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.724439562233869,
+      "learning_rate": 6.243649493523505e-07,
+      "loss": 0.127,
+      "step": 6859
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.835199059229101,
+      "learning_rate": 6.242649183811541e-07,
+      "loss": 0.0996,
+      "step": 6860
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7136121296431818,
+      "learning_rate": 6.241648821085665e-07,
+      "loss": 0.1202,
+      "step": 6861
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1481135542597556,
+      "learning_rate": 6.240648405388555e-07,
+      "loss": 0.1732,
+      "step": 6862
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.84286234223769,
+      "learning_rate": 6.239647936762888e-07,
+      "loss": 0.21,
+      "step": 6863
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.559212080866238,
+      "learning_rate": 6.238647415251349e-07,
+      "loss": 0.1487,
+      "step": 6864
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8333486084871736,
+      "learning_rate": 6.237646840896622e-07,
+      "loss": 0.0493,
+      "step": 6865
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9564629195000102,
+      "learning_rate": 6.236646213741393e-07,
+      "loss": 0.2068,
+      "step": 6866
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6018376526384829,
+      "learning_rate": 6.235645533828348e-07,
+      "loss": 0.2356,
+      "step": 6867
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6077695408432908,
+      "learning_rate": 6.234644801200182e-07,
+      "loss": 0.378,
+      "step": 6868
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.25863719910906324,
+      "learning_rate": 6.233644015899587e-07,
+      "loss": 0.1129,
+      "step": 6869
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.29980844732213,
+      "learning_rate": 6.232643177969258e-07,
+      "loss": 0.092,
+      "step": 6870
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4036243777816879,
+      "learning_rate": 6.231642287451894e-07,
+      "loss": 0.137,
+      "step": 6871
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.531938766838832,
+      "learning_rate": 6.230641344390193e-07,
+      "loss": 0.1709,
+      "step": 6872
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.626332250713213,
+      "learning_rate": 6.22964034882686e-07,
+      "loss": 0.259,
+      "step": 6873
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3131201119663751,
+      "learning_rate": 6.228639300804597e-07,
+      "loss": 0.1008,
+      "step": 6874
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.948858701528825,
+      "learning_rate": 6.227638200366111e-07,
+      "loss": 0.0435,
+      "step": 6875
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.34621796552462913,
+      "learning_rate": 6.226637047554112e-07,
+      "loss": 0.1809,
+      "step": 6876
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6545483903692884,
+      "learning_rate": 6.22563584241131e-07,
+      "loss": 0.2285,
+      "step": 6877
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4983342353946147,
+      "learning_rate": 6.224634584980419e-07,
+      "loss": 0.2559,
+      "step": 6878
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6262553504638929,
+      "learning_rate": 6.223633275304157e-07,
+      "loss": 0.2534,
+      "step": 6879
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0476521647089876,
+      "learning_rate": 6.222631913425237e-07,
+      "loss": 0.1286,
+      "step": 6880
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7532988373024432,
+      "learning_rate": 6.221630499386383e-07,
+      "loss": 0.1562,
+      "step": 6881
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2812288963283702,
+      "learning_rate": 6.220629033230317e-07,
+      "loss": 0.0915,
+      "step": 6882
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5943660575348206,
+      "learning_rate": 6.219627514999761e-07,
+      "loss": 0.1219,
+      "step": 6883
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5110541196172584,
+      "learning_rate": 6.218625944737444e-07,
+      "loss": 0.0744,
+      "step": 6884
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1145811824476093,
+      "learning_rate": 6.217624322486094e-07,
+      "loss": 0.361,
+      "step": 6885
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0813975123631303,
+      "learning_rate": 6.216622648288443e-07,
+      "loss": 0.1015,
+      "step": 6886
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5614615087155922,
+      "learning_rate": 6.215620922187226e-07,
+      "loss": 0.2179,
+      "step": 6887
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.666469936246563,
+      "learning_rate": 6.214619144225175e-07,
+      "loss": 0.1969,
+      "step": 6888
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8435504813603971,
+      "learning_rate": 6.21361731444503e-07,
+      "loss": 0.1713,
+      "step": 6889
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.12098429969003,
+      "learning_rate": 6.212615432889529e-07,
+      "loss": 0.1041,
+      "step": 6890
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6345391001430074,
+      "learning_rate": 6.211613499601418e-07,
+      "loss": 0.327,
+      "step": 6891
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1628990424915013,
+      "learning_rate": 6.210611514623439e-07,
+      "loss": 0.1649,
+      "step": 6892
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8765090102836677,
+      "learning_rate": 6.209609477998338e-07,
+      "loss": 0.2759,
+      "step": 6893
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2374649525278218,
+      "learning_rate": 6.208607389768866e-07,
+      "loss": 0.3338,
+      "step": 6894
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.199880028380958,
+      "learning_rate": 6.207605249977773e-07,
+      "loss": 0.3526,
+      "step": 6895
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5493822708776075,
+      "learning_rate": 6.206603058667814e-07,
+      "loss": 0.3173,
+      "step": 6896
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2408383608057978,
+      "learning_rate": 6.20560081588174e-07,
+      "loss": 0.2257,
+      "step": 6897
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.28977942430802,
+      "learning_rate": 6.204598521662315e-07,
+      "loss": 0.0809,
+      "step": 6898
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5550949076424141,
+      "learning_rate": 6.203596176052293e-07,
+      "loss": 0.0761,
+      "step": 6899
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2417739785051891,
+      "learning_rate": 6.20259377909444e-07,
+      "loss": 0.0807,
+      "step": 6900
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.39630038415248925,
+      "learning_rate": 6.201591330831517e-07,
+      "loss": 0.0536,
+      "step": 6901
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9842101392684679,
+      "learning_rate": 6.200588831306293e-07,
+      "loss": 0.3377,
+      "step": 6902
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5111251463979756,
+      "learning_rate": 6.199586280561538e-07,
+      "loss": 0.1592,
+      "step": 6903
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3409038537179123,
+      "learning_rate": 6.198583678640019e-07,
+      "loss": 0.1111,
+      "step": 6904
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3043629387571401,
+      "learning_rate": 6.197581025584511e-07,
+      "loss": 0.0168,
+      "step": 6905
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7300653946375809,
+      "learning_rate": 6.196578321437789e-07,
+      "loss": 0.0399,
+      "step": 6906
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8141902535050116,
+      "learning_rate": 6.19557556624263e-07,
+      "loss": 0.1366,
+      "step": 6907
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.07619674392620761,
+      "learning_rate": 6.194572760041815e-07,
+      "loss": 0.0012,
+      "step": 6908
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8270895485880623,
+      "learning_rate": 6.193569902878124e-07,
+      "loss": 0.2732,
+      "step": 6909
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8055134513779834,
+      "learning_rate": 6.192566994794342e-07,
+      "loss": 0.219,
+      "step": 6910
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3787247546125327,
+      "learning_rate": 6.191564035833253e-07,
+      "loss": 0.4333,
+      "step": 6911
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9173091222520319,
+      "learning_rate": 6.190561026037648e-07,
+      "loss": 0.2145,
+      "step": 6912
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8747517670691473,
+      "learning_rate": 6.189557965450316e-07,
+      "loss": 0.1481,
+      "step": 6913
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5975429868199008,
+      "learning_rate": 6.188554854114052e-07,
+      "loss": 0.137,
+      "step": 6914
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7264176118036235,
+      "learning_rate": 6.187551692071648e-07,
+      "loss": 0.1171,
+      "step": 6915
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.098857129560535,
+      "learning_rate": 6.1865484793659e-07,
+      "loss": 0.487,
+      "step": 6916
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5252791025312935,
+      "learning_rate": 6.185545216039609e-07,
+      "loss": 0.2031,
+      "step": 6917
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6138002838529008,
+      "learning_rate": 6.184541902135576e-07,
+      "loss": 0.1736,
+      "step": 6918
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5698825485798151,
+      "learning_rate": 6.183538537696604e-07,
+      "loss": 0.2127,
+      "step": 6919
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5815916642846686,
+      "learning_rate": 6.182535122765498e-07,
+      "loss": 0.1407,
+      "step": 6920
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9493237980265676,
+      "learning_rate": 6.181531657385068e-07,
+      "loss": 0.1887,
+      "step": 6921
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7150951635851865,
+      "learning_rate": 6.180528141598121e-07,
+      "loss": 0.1532,
+      "step": 6922
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.39603092612381635,
+      "learning_rate": 6.179524575447471e-07,
+      "loss": 0.0771,
+      "step": 6923
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0938817803981258,
+      "learning_rate": 6.178520958975932e-07,
+      "loss": 0.2842,
+      "step": 6924
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8009649953234502,
+      "learning_rate": 6.17751729222632e-07,
+      "loss": 0.0132,
+      "step": 6925
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.33108503279882795,
+      "learning_rate": 6.176513575241452e-07,
+      "loss": 0.0082,
+      "step": 6926
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.31729671294644823,
+      "learning_rate": 6.175509808064149e-07,
+      "loss": 0.1053,
+      "step": 6927
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5492720653003984,
+      "learning_rate": 6.174505990737238e-07,
+      "loss": 0.1087,
+      "step": 6928
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.41360439446294106,
+      "learning_rate": 6.173502123303538e-07,
+      "loss": 0.2309,
+      "step": 6929
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5401488246862701,
+      "learning_rate": 6.172498205805878e-07,
+      "loss": 0.2762,
+      "step": 6930
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1357691343779526,
+      "learning_rate": 6.171494238287088e-07,
+      "loss": 0.26,
+      "step": 6931
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.34405876677592406,
+      "learning_rate": 6.170490220789999e-07,
+      "loss": 0.2515,
+      "step": 6932
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0478404241618138,
+      "learning_rate": 6.169486153357444e-07,
+      "loss": 0.2009,
+      "step": 6933
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.40073446946851377,
+      "learning_rate": 6.168482036032259e-07,
+      "loss": 0.0367,
+      "step": 6934
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6573471808166629,
+      "learning_rate": 6.167477868857281e-07,
+      "loss": 0.2296,
+      "step": 6935
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.43918478126204147,
+      "learning_rate": 6.166473651875351e-07,
+      "loss": 0.0706,
+      "step": 6936
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.002656753065787,
+      "learning_rate": 6.165469385129309e-07,
+      "loss": 0.1135,
+      "step": 6937
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.494979637287609,
+      "learning_rate": 6.164465068662e-07,
+      "loss": 0.4066,
+      "step": 6938
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.47883964075679697,
+      "learning_rate": 6.163460702516271e-07,
+      "loss": 0.1891,
+      "step": 6939
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1409255996678267,
+      "learning_rate": 6.162456286734969e-07,
+      "loss": 0.3263,
+      "step": 6940
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3621007983991172,
+      "learning_rate": 6.161451821360947e-07,
+      "loss": 0.1045,
+      "step": 6941
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0862254194163052,
+      "learning_rate": 6.160447306437054e-07,
+      "loss": 0.2759,
+      "step": 6942
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5946015766783351,
+      "learning_rate": 6.159442742006147e-07,
+      "loss": 0.0527,
+      "step": 6943
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5582335389228525,
+      "learning_rate": 6.158438128111081e-07,
+      "loss": 0.0515,
+      "step": 6944
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6744063246468665,
+      "learning_rate": 6.157433464794716e-07,
+      "loss": 0.0112,
+      "step": 6945
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7471634695884373,
+      "learning_rate": 6.156428752099912e-07,
+      "loss": 0.1808,
+      "step": 6946
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.383769180786806,
+      "learning_rate": 6.155423990069532e-07,
+      "loss": 0.1272,
+      "step": 6947
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.464658661457626,
+      "learning_rate": 6.154419178746443e-07,
+      "loss": 0.2879,
+      "step": 6948
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.586826095802796,
+      "learning_rate": 6.153414318173511e-07,
+      "loss": 0.1035,
+      "step": 6949
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.42072790409513783,
+      "learning_rate": 6.152409408393606e-07,
+      "loss": 0.0064,
+      "step": 6950
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3812237351074588,
+      "learning_rate": 6.151404449449599e-07,
+      "loss": 0.1688,
+      "step": 6951
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0321606069137594,
+      "learning_rate": 6.150399441384363e-07,
+      "loss": 0.1237,
+      "step": 6952
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.606937099338738,
+      "learning_rate": 6.149394384240775e-07,
+      "loss": 0.205,
+      "step": 6953
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.982863396210204,
+      "learning_rate": 6.148389278061711e-07,
+      "loss": 0.2688,
+      "step": 6954
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5434452744793399,
+      "learning_rate": 6.147384122890052e-07,
+      "loss": 0.281,
+      "step": 6955
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0847305758437145,
+      "learning_rate": 6.146378918768681e-07,
+      "loss": 0.2153,
+      "step": 6956
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8470400376818763,
+      "learning_rate": 6.14537366574048e-07,
+      "loss": 0.3103,
+      "step": 6957
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7175605445892274,
+      "learning_rate": 6.144368363848335e-07,
+      "loss": 0.2604,
+      "step": 6958
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.565787560280558,
+      "learning_rate": 6.143363013135136e-07,
+      "loss": 0.2501,
+      "step": 6959
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3774440405473363,
+      "learning_rate": 6.142357613643773e-07,
+      "loss": 0.2021,
+      "step": 6960
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2299796206338582,
+      "learning_rate": 6.141352165417137e-07,
+      "loss": 0.1713,
+      "step": 6961
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7901467738433083,
+      "learning_rate": 6.140346668498124e-07,
+      "loss": 0.1122,
+      "step": 6962
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6819889165777708,
+      "learning_rate": 6.139341122929629e-07,
+      "loss": 0.1528,
+      "step": 6963
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8651701321719657,
+      "learning_rate": 6.13833552875455e-07,
+      "loss": 0.2012,
+      "step": 6964
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17737635653643177,
+      "learning_rate": 6.137329886015791e-07,
+      "loss": 0.019,
+      "step": 6965
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2778729695799618,
+      "learning_rate": 6.136324194756252e-07,
+      "loss": 0.3287,
+      "step": 6966
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5028356491136653,
+      "learning_rate": 6.135318455018839e-07,
+      "loss": 0.185,
+      "step": 6967
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.49877890473115016,
+      "learning_rate": 6.134312666846459e-07,
+      "loss": 0.1714,
+      "step": 6968
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7981573156656027,
+      "learning_rate": 6.13330683028202e-07,
+      "loss": 0.0811,
+      "step": 6969
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6005895831721726,
+      "learning_rate": 6.132300945368433e-07,
+      "loss": 0.3987,
+      "step": 6970
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.31192909204643715,
+      "learning_rate": 6.131295012148612e-07,
+      "loss": 0.2181,
+      "step": 6971
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3545310906924588,
+      "learning_rate": 6.130289030665468e-07,
+      "loss": 0.12,
+      "step": 6972
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.025570772503253,
+      "learning_rate": 6.129283000961926e-07,
+      "loss": 0.3388,
+      "step": 6973
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5429665518394462,
+      "learning_rate": 6.1282769230809e-07,
+      "loss": 0.2221,
+      "step": 6974
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0744699445752983,
+      "learning_rate": 6.127270797065312e-07,
+      "loss": 0.0902,
+      "step": 6975
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2929872500332653,
+      "learning_rate": 6.126264622958086e-07,
+      "loss": 0.038,
+      "step": 6976
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7455266878558893,
+      "learning_rate": 6.125258400802146e-07,
+      "loss": 0.3655,
+      "step": 6977
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.627917352512731,
+      "learning_rate": 6.124252130640423e-07,
+      "loss": 0.0604,
+      "step": 6978
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8291769579982922,
+      "learning_rate": 6.123245812515843e-07,
+      "loss": 0.2673,
+      "step": 6979
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4631818470052978,
+      "learning_rate": 6.122239446471338e-07,
+      "loss": 0.1917,
+      "step": 6980
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0010906980464314,
+      "learning_rate": 6.121233032549842e-07,
+      "loss": 0.0943,
+      "step": 6981
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.113879147230086,
+      "learning_rate": 6.120226570794291e-07,
+      "loss": 0.2011,
+      "step": 6982
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1995621808630295,
+      "learning_rate": 6.119220061247622e-07,
+      "loss": 0.0945,
+      "step": 6983
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8629359079521771,
+      "learning_rate": 6.118213503952778e-07,
+      "loss": 0.1741,
+      "step": 6984
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7448026705895965,
+      "learning_rate": 6.117206898952696e-07,
+      "loss": 0.3602,
+      "step": 6985
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.197204476150808,
+      "learning_rate": 6.116200246290322e-07,
+      "loss": 0.3732,
+      "step": 6986
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7836219455196711,
+      "learning_rate": 6.115193546008601e-07,
+      "loss": 0.2002,
+      "step": 6987
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5205671280907647,
+      "learning_rate": 6.114186798150482e-07,
+      "loss": 0.2355,
+      "step": 6988
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7519950190590827,
+      "learning_rate": 6.113180002758915e-07,
+      "loss": 0.2139,
+      "step": 6989
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3892107135968623,
+      "learning_rate": 6.112173159876851e-07,
+      "loss": 0.0879,
+      "step": 6990
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.31456452934603335,
+      "learning_rate": 6.111166269547243e-07,
+      "loss": 0.0036,
+      "step": 6991
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.795319291289239,
+      "learning_rate": 6.110159331813049e-07,
+      "loss": 0.1771,
+      "step": 6992
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.76803851220041,
+      "learning_rate": 6.109152346717228e-07,
+      "loss": 0.2949,
+      "step": 6993
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.126107545010468,
+      "learning_rate": 6.108145314302736e-07,
+      "loss": 0.3448,
+      "step": 6994
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7010751213890889,
+      "learning_rate": 6.107138234612539e-07,
+      "loss": 0.342,
+      "step": 6995
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1487274465706263,
+      "learning_rate": 6.106131107689598e-07,
+      "loss": 0.1234,
+      "step": 6996
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0047599593340455,
+      "learning_rate": 6.105123933576881e-07,
+      "loss": 0.2634,
+      "step": 6997
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6502696156537707,
+      "learning_rate": 6.104116712317355e-07,
+      "loss": 0.468,
+      "step": 6998
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.27978610515785746,
+      "learning_rate": 6.10310944395399e-07,
+      "loss": 0.1076,
+      "step": 6999
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5266859976788861,
+      "learning_rate": 6.102102128529759e-07,
+      "loss": 0.0514,
+      "step": 7000
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.49816105419925955,
+      "learning_rate": 6.101094766087637e-07,
+      "loss": 0.1085,
+      "step": 7001
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8781912021438211,
+      "learning_rate": 6.100087356670596e-07,
+      "loss": 0.302,
+      "step": 7002
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.143626452489102,
+      "learning_rate": 6.09907990032162e-07,
+      "loss": 0.2635,
+      "step": 7003
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9082295169453426,
+      "learning_rate": 6.098072397083684e-07,
+      "loss": 0.1063,
+      "step": 7004
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3790899514979397,
+      "learning_rate": 6.097064846999773e-07,
+      "loss": 0.1614,
+      "step": 7005
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5109056251055746,
+      "learning_rate": 6.096057250112869e-07,
+      "loss": 0.1268,
+      "step": 7006
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3445931844183288,
+      "learning_rate": 6.095049606465962e-07,
+      "loss": 0.1876,
+      "step": 7007
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6897933713277216,
+      "learning_rate": 6.094041916102035e-07,
+      "loss": 0.0207,
+      "step": 7008
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.701035327618761,
+      "learning_rate": 6.093034179064081e-07,
+      "loss": 0.1681,
+      "step": 7009
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5862399618767468,
+      "learning_rate": 6.092026395395091e-07,
+      "loss": 0.282,
+      "step": 7010
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9519031633352236,
+      "learning_rate": 6.091018565138061e-07,
+      "loss": 0.0982,
+      "step": 7011
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.167690752353747,
+      "learning_rate": 6.090010688335987e-07,
+      "loss": 0.1385,
+      "step": 7012
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6973553591500395,
+      "learning_rate": 6.089002765031864e-07,
+      "loss": 0.22,
+      "step": 7013
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.19367020577117855,
+      "learning_rate": 6.087994795268695e-07,
+      "loss": 0.0833,
+      "step": 7014
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5990229607046382,
+      "learning_rate": 6.08698677908948e-07,
+      "loss": 0.2395,
+      "step": 7015
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8620603415141842,
+      "learning_rate": 6.085978716537223e-07,
+      "loss": 0.3999,
+      "step": 7016
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.049386554922625,
+      "learning_rate": 6.084970607654931e-07,
+      "loss": 0.1278,
+      "step": 7017
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6789723515233934,
+      "learning_rate": 6.083962452485614e-07,
+      "loss": 0.0992,
+      "step": 7018
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6013876531051031,
+      "learning_rate": 6.082954251072278e-07,
+      "loss": 0.068,
+      "step": 7019
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7665096500380996,
+      "learning_rate": 6.081946003457936e-07,
+      "loss": 0.1732,
+      "step": 7020
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6786080414216922,
+      "learning_rate": 6.080937709685604e-07,
+      "loss": 0.1515,
+      "step": 7021
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6232705392964524,
+      "learning_rate": 6.079929369798297e-07,
+      "loss": 0.3126,
+      "step": 7022
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9174738349237397,
+      "learning_rate": 6.078920983839031e-07,
+      "loss": 0.0919,
+      "step": 7023
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3351450552038322,
+      "learning_rate": 6.077912551850828e-07,
+      "loss": 0.1825,
+      "step": 7024
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.23047285273482546,
+      "learning_rate": 6.076904073876706e-07,
+      "loss": 0.1486,
+      "step": 7025
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.8596867267407,
+      "learning_rate": 6.075895549959693e-07,
+      "loss": 0.0656,
+      "step": 7026
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5862264008202591,
+      "learning_rate": 6.074886980142813e-07,
+      "loss": 0.2388,
+      "step": 7027
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6502582586265484,
+      "learning_rate": 6.073878364469094e-07,
+      "loss": 0.3046,
+      "step": 7028
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7072089060085425,
+      "learning_rate": 6.072869702981565e-07,
+      "loss": 0.016,
+      "step": 7029
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9694642254886912,
+      "learning_rate": 6.071860995723257e-07,
+      "loss": 0.6971,
+      "step": 7030
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6990240558170092,
+      "learning_rate": 6.070852242737206e-07,
+      "loss": 0.16,
+      "step": 7031
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.936223825646583,
+      "learning_rate": 6.069843444066444e-07,
+      "loss": 0.295,
+      "step": 7032
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.40226374749511673,
+      "learning_rate": 6.06883459975401e-07,
+      "loss": 0.1209,
+      "step": 7033
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.935690253055944,
+      "learning_rate": 6.067825709842945e-07,
+      "loss": 0.1124,
+      "step": 7034
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2207454136067766,
+      "learning_rate": 6.066816774376287e-07,
+      "loss": 0.1075,
+      "step": 7035
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5618052578226107,
+      "learning_rate": 6.06580779339708e-07,
+      "loss": 0.0529,
+      "step": 7036
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7092968831146726,
+      "learning_rate": 6.064798766948371e-07,
+      "loss": 0.2881,
+      "step": 7037
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.862719814670639,
+      "learning_rate": 6.063789695073208e-07,
+      "loss": 0.1436,
+      "step": 7038
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6966503669595614,
+      "learning_rate": 6.062780577814636e-07,
+      "loss": 0.2753,
+      "step": 7039
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.15349847192339625,
+      "learning_rate": 6.061771415215708e-07,
+      "loss": 0.0889,
+      "step": 7040
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8643961482115704,
+      "learning_rate": 6.060762207319479e-07,
+      "loss": 0.1374,
+      "step": 7041
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2582579644147097,
+      "learning_rate": 6.059752954168999e-07,
+      "loss": 0.0937,
+      "step": 7042
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8432267169130003,
+      "learning_rate": 6.058743655807331e-07,
+      "loss": 0.2547,
+      "step": 7043
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2976526130842025,
+      "learning_rate": 6.057734312277526e-07,
+      "loss": 0.0898,
+      "step": 7044
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.102839488835296,
+      "learning_rate": 6.056724923622651e-07,
+      "loss": 0.1107,
+      "step": 7045
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.550269931757731,
+      "learning_rate": 6.055715489885768e-07,
+      "loss": 0.0329,
+      "step": 7046
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4707632456794082,
+      "learning_rate": 6.054706011109938e-07,
+      "loss": 0.0565,
+      "step": 7047
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.46532627867999804,
+      "learning_rate": 6.05369648733823e-07,
+      "loss": 0.0121,
+      "step": 7048
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.047816250916057,
+      "learning_rate": 6.052686918613712e-07,
+      "loss": 0.4253,
+      "step": 7049
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0216356209529873,
+      "learning_rate": 6.051677304979453e-07,
+      "loss": 0.1661,
+      "step": 7050
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7515596358866237,
+      "learning_rate": 6.050667646478527e-07,
+      "loss": 0.3816,
+      "step": 7051
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5555317867772822,
+      "learning_rate": 6.049657943154006e-07,
+      "loss": 0.2741,
+      "step": 7052
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.45035019037585133,
+      "learning_rate": 6.048648195048968e-07,
+      "loss": 0.0255,
+      "step": 7053
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3586966690743851,
+      "learning_rate": 6.047638402206489e-07,
+      "loss": 0.1486,
+      "step": 7054
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7127962571969929,
+      "learning_rate": 6.046628564669651e-07,
+      "loss": 0.0519,
+      "step": 7055
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1781048738549795,
+      "learning_rate": 6.045618682481535e-07,
+      "loss": 0.3443,
+      "step": 7056
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4974924732471555,
+      "learning_rate": 6.044608755685222e-07,
+      "loss": 0.1785,
+      "step": 7057
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9627407303530922,
+      "learning_rate": 6.043598784323802e-07,
+      "loss": 0.1179,
+      "step": 7058
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8824047630303993,
+      "learning_rate": 6.042588768440357e-07,
+      "loss": 0.4079,
+      "step": 7059
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7897017631778379,
+      "learning_rate": 6.041578708077981e-07,
+      "loss": 0.2406,
+      "step": 7060
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4281677492674576,
+      "learning_rate": 6.040568603279763e-07,
+      "loss": 0.1699,
+      "step": 7061
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3898733033420859,
+      "learning_rate": 6.039558454088795e-07,
+      "loss": 0.1781,
+      "step": 7062
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5578960512714476,
+      "learning_rate": 6.038548260548176e-07,
+      "loss": 0.2769,
+      "step": 7063
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7492669585363871,
+      "learning_rate": 6.037538022700999e-07,
+      "loss": 0.1844,
+      "step": 7064
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6116049102028214,
+      "learning_rate": 6.036527740590365e-07,
+      "loss": 0.2459,
+      "step": 7065
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2831282294351607,
+      "learning_rate": 6.035517414259377e-07,
+      "loss": 0.1085,
+      "step": 7066
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0692156727939737,
+      "learning_rate": 6.03450704375113e-07,
+      "loss": 0.4134,
+      "step": 7067
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9839753343140111,
+      "learning_rate": 6.033496629108736e-07,
+      "loss": 0.2069,
+      "step": 7068
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9041922192505167,
+      "learning_rate": 6.032486170375296e-07,
+      "loss": 0.0116,
+      "step": 7069
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4681185734971004,
+      "learning_rate": 6.031475667593919e-07,
+      "loss": 0.2982,
+      "step": 7070
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7860744215186456,
+      "learning_rate": 6.030465120807719e-07,
+      "loss": 0.0097,
+      "step": 7071
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9383937659706048,
+      "learning_rate": 6.029454530059806e-07,
+      "loss": 0.1902,
+      "step": 7072
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.718398069911204,
+      "learning_rate": 6.028443895393291e-07,
+      "loss": 0.3911,
+      "step": 7073
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.06919293034300165,
+      "learning_rate": 6.027433216851294e-07,
+      "loss": 0.0005,
+      "step": 7074
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0320483799463922,
+      "learning_rate": 6.026422494476929e-07,
+      "loss": 0.1796,
+      "step": 7075
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9449293296619714,
+      "learning_rate": 6.025411728313317e-07,
+      "loss": 0.1266,
+      "step": 7076
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2993277087210431,
+      "learning_rate": 6.02440091840358e-07,
+      "loss": 0.1716,
+      "step": 7077
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.945913960885128,
+      "learning_rate": 6.023390064790841e-07,
+      "loss": 0.298,
+      "step": 7078
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8939332327962871,
+      "learning_rate": 6.022379167518225e-07,
+      "loss": 0.0611,
+      "step": 7079
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9663232671040488,
+      "learning_rate": 6.021368226628857e-07,
+      "loss": 0.4879,
+      "step": 7080
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3242663789682058,
+      "learning_rate": 6.020357242165868e-07,
+      "loss": 0.1704,
+      "step": 7081
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9455796030817295,
+      "learning_rate": 6.019346214172388e-07,
+      "loss": 0.1807,
+      "step": 7082
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4402759090689443,
+      "learning_rate": 6.018335142691548e-07,
+      "loss": 0.0593,
+      "step": 7083
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2060208908726957,
+      "learning_rate": 6.017324027766486e-07,
+      "loss": 0.2904,
+      "step": 7084
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7129259771847446,
+      "learning_rate": 6.016312869440334e-07,
+      "loss": 0.4456,
+      "step": 7085
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7854997394334815,
+      "learning_rate": 6.015301667756233e-07,
+      "loss": 0.1428,
+      "step": 7086
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.48994875507244606,
+      "learning_rate": 6.014290422757322e-07,
+      "loss": 0.2895,
+      "step": 7087
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7294957042734611,
+      "learning_rate": 6.013279134486742e-07,
+      "loss": 0.0426,
+      "step": 7088
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7434832182245137,
+      "learning_rate": 6.012267802987636e-07,
+      "loss": 0.1383,
+      "step": 7089
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4636452503714737,
+      "learning_rate": 6.011256428303152e-07,
+      "loss": 0.1793,
+      "step": 7090
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7988272816439137,
+      "learning_rate": 6.010245010476436e-07,
+      "loss": 0.3583,
+      "step": 7091
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9552552379750733,
+      "learning_rate": 6.009233549550636e-07,
+      "loss": 0.3605,
+      "step": 7092
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.43371975585381,
+      "learning_rate": 6.008222045568907e-07,
+      "loss": 0.1882,
+      "step": 7093
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3571109562706452,
+      "learning_rate": 6.007210498574395e-07,
+      "loss": 0.2746,
+      "step": 7094
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4418232209930374,
+      "learning_rate": 6.00619890861026e-07,
+      "loss": 0.1772,
+      "step": 7095
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.088758666672563,
+      "learning_rate": 6.005187275719657e-07,
+      "loss": 0.0755,
+      "step": 7096
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5687223310030136,
+      "learning_rate": 6.004175599945743e-07,
+      "loss": 0.1623,
+      "step": 7097
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.01592613710664,
+      "learning_rate": 6.003163881331681e-07,
+      "loss": 0.2278,
+      "step": 7098
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3370256316854185,
+      "learning_rate": 6.002152119920629e-07,
+      "loss": 0.1968,
+      "step": 7099
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4808284341035867,
+      "learning_rate": 6.001140315755755e-07,
+      "loss": 0.2689,
+      "step": 7100
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.011765000959112,
+      "learning_rate": 6.000128468880222e-07,
+      "loss": 0.3323,
+      "step": 7101
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8513305845378742,
+      "learning_rate": 5.999116579337198e-07,
+      "loss": 0.2085,
+      "step": 7102
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6186146638521367,
+      "learning_rate": 5.998104647169852e-07,
+      "loss": 0.0298,
+      "step": 7103
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7770787457397188,
+      "learning_rate": 5.997092672421356e-07,
+      "loss": 0.2867,
+      "step": 7104
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6759549912096738,
+      "learning_rate": 5.996080655134881e-07,
+      "loss": 0.3032,
+      "step": 7105
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.375055395363622,
+      "learning_rate": 5.995068595353604e-07,
+      "loss": 0.0204,
+      "step": 7106
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8979471716423942,
+      "learning_rate": 5.994056493120699e-07,
+      "loss": 0.2618,
+      "step": 7107
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5738377713029338,
+      "learning_rate": 5.993044348479347e-07,
+      "loss": 0.1037,
+      "step": 7108
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5764327562948564,
+      "learning_rate": 5.992032161472726e-07,
+      "loss": 0.1666,
+      "step": 7109
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5461270918707353,
+      "learning_rate": 5.99101993214402e-07,
+      "loss": 0.3417,
+      "step": 7110
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4657147448614762,
+      "learning_rate": 5.99000766053641e-07,
+      "loss": 0.1103,
+      "step": 7111
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8157136799194386,
+      "learning_rate": 5.988995346693084e-07,
+      "loss": 0.1337,
+      "step": 7112
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.46844818672747407,
+      "learning_rate": 5.987982990657228e-07,
+      "loss": 0.1807,
+      "step": 7113
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8863593073604794,
+      "learning_rate": 5.986970592472033e-07,
+      "loss": 0.3098,
+      "step": 7114
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5580882911711244,
+      "learning_rate": 5.985958152180686e-07,
+      "loss": 0.2203,
+      "step": 7115
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9407780401334193,
+      "learning_rate": 5.984945669826382e-07,
+      "loss": 0.2465,
+      "step": 7116
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.699661212881519,
+      "learning_rate": 5.983933145452318e-07,
+      "loss": 0.2772,
+      "step": 7117
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.24115919400100697,
+      "learning_rate": 5.982920579101687e-07,
+      "loss": 0.0113,
+      "step": 7118
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5354065445651078,
+      "learning_rate": 5.981907970817688e-07,
+      "loss": 0.1612,
+      "step": 7119
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6437441543164536,
+      "learning_rate": 5.980895320643521e-07,
+      "loss": 0.2031,
+      "step": 7120
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7059838005150271,
+      "learning_rate": 5.979882628622389e-07,
+      "loss": 0.4494,
+      "step": 7121
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.31758025093563497,
+      "learning_rate": 5.978869894797493e-07,
+      "loss": 0.1973,
+      "step": 7122
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6120633224019312,
+      "learning_rate": 5.97785711921204e-07,
+      "loss": 0.3558,
+      "step": 7123
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.312367372031094,
+      "learning_rate": 5.976844301909236e-07,
+      "loss": 0.2304,
+      "step": 7124
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3313741831082385,
+      "learning_rate": 5.975831442932291e-07,
+      "loss": 0.1006,
+      "step": 7125
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5129312842561959,
+      "learning_rate": 5.974818542324414e-07,
+      "loss": 0.22,
+      "step": 7126
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.32401724292769385,
+      "learning_rate": 5.97380560012882e-07,
+      "loss": 0.2122,
+      "step": 7127
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4625502269175985,
+      "learning_rate": 5.972792616388721e-07,
+      "loss": 0.055,
+      "step": 7128
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9942322616535594,
+      "learning_rate": 5.971779591147332e-07,
+      "loss": 0.1874,
+      "step": 7129
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4563264253733514,
+      "learning_rate": 5.970766524447875e-07,
+      "loss": 0.087,
+      "step": 7130
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2157699901808448,
+      "learning_rate": 5.969753416333564e-07,
+      "loss": 0.0866,
+      "step": 7131
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8444079709273857,
+      "learning_rate": 5.968740266847623e-07,
+      "loss": 0.1228,
+      "step": 7132
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.34704186053735525,
+      "learning_rate": 5.967727076033274e-07,
+      "loss": 0.0757,
+      "step": 7133
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3237010936906624,
+      "learning_rate": 5.966713843933746e-07,
+      "loss": 0.0512,
+      "step": 7134
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.932547110155241,
+      "learning_rate": 5.965700570592261e-07,
+      "loss": 0.245,
+      "step": 7135
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7496287466544687,
+      "learning_rate": 5.964687256052046e-07,
+      "loss": 0.2027,
+      "step": 7136
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.30193107065623703,
+      "learning_rate": 5.963673900356335e-07,
+      "loss": 0.092,
+      "step": 7137
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.332263676956981,
+      "learning_rate": 5.962660503548358e-07,
+      "loss": 0.1935,
+      "step": 7138
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6921697636905982,
+      "learning_rate": 5.961647065671349e-07,
+      "loss": 0.2038,
+      "step": 7139
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7691815576656338,
+      "learning_rate": 5.960633586768542e-07,
+      "loss": 0.1293,
+      "step": 7140
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5950655910490383,
+      "learning_rate": 5.959620066883175e-07,
+      "loss": 0.163,
+      "step": 7141
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6732396003329623,
+      "learning_rate": 5.958606506058488e-07,
+      "loss": 0.1545,
+      "step": 7142
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5869733926350481,
+      "learning_rate": 5.95759290433772e-07,
+      "loss": 0.0886,
+      "step": 7143
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8168037275744525,
+      "learning_rate": 5.956579261764115e-07,
+      "loss": 0.3139,
+      "step": 7144
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.32485987063036015,
+      "learning_rate": 5.955565578380914e-07,
+      "loss": 0.1552,
+      "step": 7145
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.31954013098482065,
+      "learning_rate": 5.954551854231365e-07,
+      "loss": 0.086,
+      "step": 7146
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2954219369681322,
+      "learning_rate": 5.953538089358713e-07,
+      "loss": 0.0162,
+      "step": 7147
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.66192513564281,
+      "learning_rate": 5.952524283806214e-07,
+      "loss": 0.2002,
+      "step": 7148
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6250885992829237,
+      "learning_rate": 5.95151043761711e-07,
+      "loss": 0.2374,
+      "step": 7149
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.30212655217801865,
+      "learning_rate": 5.950496550834659e-07,
+      "loss": 0.1873,
+      "step": 7150
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.091021040170508,
+      "learning_rate": 5.949482623502116e-07,
+      "loss": 0.3885,
+      "step": 7151
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6016881690483314,
+      "learning_rate": 5.948468655662734e-07,
+      "loss": 0.311,
+      "step": 7152
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8520165271748031,
+      "learning_rate": 5.947454647359774e-07,
+      "loss": 0.4149,
+      "step": 7153
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6822457209307743,
+      "learning_rate": 5.946440598636492e-07,
+      "loss": 0.1621,
+      "step": 7154
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9286637757220354,
+      "learning_rate": 5.945426509536152e-07,
+      "loss": 0.1993,
+      "step": 7155
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4551563553112895,
+      "learning_rate": 5.944412380102017e-07,
+      "loss": 0.1555,
+      "step": 7156
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.549513819768842,
+      "learning_rate": 5.943398210377352e-07,
+      "loss": 0.317,
+      "step": 7157
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8022045638220008,
+      "learning_rate": 5.942384000405423e-07,
+      "loss": 0.2718,
+      "step": 7158
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5936247476974245,
+      "learning_rate": 5.941369750229497e-07,
+      "loss": 0.2676,
+      "step": 7159
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.43999162216149557,
+      "learning_rate": 5.940355459892844e-07,
+      "loss": 0.3624,
+      "step": 7160
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.844207234831347,
+      "learning_rate": 5.939341129438738e-07,
+      "loss": 0.054,
+      "step": 7161
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5082391569111769,
+      "learning_rate": 5.938326758910453e-07,
+      "loss": 0.2208,
+      "step": 7162
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.172219731786886,
+      "learning_rate": 5.93731234835126e-07,
+      "loss": 0.3387,
+      "step": 7163
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4864323174163239,
+      "learning_rate": 5.936297897804439e-07,
+      "loss": 0.2884,
+      "step": 7164
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4171003748458937,
+      "learning_rate": 5.935283407313268e-07,
+      "loss": 0.1463,
+      "step": 7165
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2970815946203582,
+      "learning_rate": 5.934268876921025e-07,
+      "loss": 0.3739,
+      "step": 7166
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5236303770620295,
+      "learning_rate": 5.933254306670994e-07,
+      "loss": 0.1736,
+      "step": 7167
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5700171082654275,
+      "learning_rate": 5.932239696606457e-07,
+      "loss": 0.2731,
+      "step": 7168
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.43685907066280566,
+      "learning_rate": 5.931225046770703e-07,
+      "loss": 0.2653,
+      "step": 7169
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8688550234332038,
+      "learning_rate": 5.930210357207015e-07,
+      "loss": 0.2695,
+      "step": 7170
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0139315492143588,
+      "learning_rate": 5.929195627958683e-07,
+      "loss": 0.1514,
+      "step": 7171
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.563002708663235,
+      "learning_rate": 5.928180859068999e-07,
+      "loss": 0.1547,
+      "step": 7172
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5001619767205903,
+      "learning_rate": 5.927166050581252e-07,
+      "loss": 0.033,
+      "step": 7173
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6917706009406938,
+      "learning_rate": 5.926151202538739e-07,
+      "loss": 0.1704,
+      "step": 7174
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9353287183536301,
+      "learning_rate": 5.925136314984753e-07,
+      "loss": 0.2011,
+      "step": 7175
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6112538646341721,
+      "learning_rate": 5.924121387962593e-07,
+      "loss": 0.0681,
+      "step": 7176
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.858619915778985,
+      "learning_rate": 5.923106421515556e-07,
+      "loss": 0.3036,
+      "step": 7177
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.354676634271242,
+      "learning_rate": 5.922091415686944e-07,
+      "loss": 0.0096,
+      "step": 7178
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2571455296787604,
+      "learning_rate": 5.921076370520057e-07,
+      "loss": 0.1193,
+      "step": 7179
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.736715975373125,
+      "learning_rate": 5.920061286058202e-07,
+      "loss": 0.1809,
+      "step": 7180
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.068782573052699,
+      "learning_rate": 5.919046162344683e-07,
+      "loss": 0.0239,
+      "step": 7181
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3297467235529592,
+      "learning_rate": 5.918030999422808e-07,
+      "loss": 0.0886,
+      "step": 7182
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7417717017004508,
+      "learning_rate": 5.917015797335882e-07,
+      "loss": 0.1744,
+      "step": 7183
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0933639592682791,
+      "learning_rate": 5.916000556127221e-07,
+      "loss": 0.1144,
+      "step": 7184
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.17164216430584833,
+      "learning_rate": 5.914985275840135e-07,
+      "loss": 0.0051,
+      "step": 7185
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6620853225118942,
+      "learning_rate": 5.913969956517936e-07,
+      "loss": 0.1639,
+      "step": 7186
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0864099238793041,
+      "learning_rate": 5.912954598203943e-07,
+      "loss": 0.3266,
+      "step": 7187
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6563817289774813,
+      "learning_rate": 5.91193920094147e-07,
+      "loss": 0.3287,
+      "step": 7188
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7480545118849424,
+      "learning_rate": 5.910923764773841e-07,
+      "loss": 0.0214,
+      "step": 7189
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7998208232146227,
+      "learning_rate": 5.90990828974437e-07,
+      "loss": 0.3755,
+      "step": 7190
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.85748334831464,
+      "learning_rate": 5.908892775896383e-07,
+      "loss": 0.0507,
+      "step": 7191
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1970218213760373,
+      "learning_rate": 5.907877223273202e-07,
+      "loss": 0.2678,
+      "step": 7192
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2924320625986172,
+      "learning_rate": 5.906861631918155e-07,
+      "loss": 0.2574,
+      "step": 7193
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7987424260681844,
+      "learning_rate": 5.905846001874566e-07,
+      "loss": 0.1193,
+      "step": 7194
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6319274213415333,
+      "learning_rate": 5.904830333185768e-07,
+      "loss": 0.1224,
+      "step": 7195
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3711639795844755,
+      "learning_rate": 5.903814625895088e-07,
+      "loss": 0.0397,
+      "step": 7196
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3745274574843632,
+      "learning_rate": 5.902798880045858e-07,
+      "loss": 0.0753,
+      "step": 7197
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.56204350914102,
+      "learning_rate": 5.901783095681414e-07,
+      "loss": 0.0599,
+      "step": 7198
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1506934707785388,
+      "learning_rate": 5.900767272845091e-07,
+      "loss": 0.212,
+      "step": 7199
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5970370856193812,
+      "learning_rate": 5.899751411580224e-07,
+      "loss": 0.2976,
+      "step": 7200
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9259827639952576,
+      "learning_rate": 5.898735511930155e-07,
+      "loss": 0.2,
+      "step": 7201
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.46751645919753676,
+      "learning_rate": 5.89771957393822e-07,
+      "loss": 0.2628,
+      "step": 7202
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8725741363156909,
+      "learning_rate": 5.896703597647764e-07,
+      "loss": 0.2463,
+      "step": 7203
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4982264543897335,
+      "learning_rate": 5.89568758310213e-07,
+      "loss": 0.1472,
+      "step": 7204
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1028605339442303,
+      "learning_rate": 5.894671530344664e-07,
+      "loss": 0.193,
+      "step": 7205
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5001892897512012,
+      "learning_rate": 5.893655439418711e-07,
+      "loss": 0.088,
+      "step": 7206
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.022140920096918,
+      "learning_rate": 5.892639310367622e-07,
+      "loss": 0.1458,
+      "step": 7207
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.497112180372172,
+      "learning_rate": 5.891623143234744e-07,
+      "loss": 0.189,
+      "step": 7208
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.30387165766925095,
+      "learning_rate": 5.89060693806343e-07,
+      "loss": 0.0739,
+      "step": 7209
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0157909781768688,
+      "learning_rate": 5.889590694897035e-07,
+      "loss": 0.5586,
+      "step": 7210
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1583784876385135,
+      "learning_rate": 5.888574413778913e-07,
+      "loss": 0.3272,
+      "step": 7211
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7133214714640085,
+      "learning_rate": 5.88755809475242e-07,
+      "loss": 0.1663,
+      "step": 7212
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5133567316724253,
+      "learning_rate": 5.886541737860912e-07,
+      "loss": 0.1052,
+      "step": 7213
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 17.285625438249102,
+      "learning_rate": 5.885525343147754e-07,
+      "loss": 0.2791,
+      "step": 7214
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1601010633523465,
+      "learning_rate": 5.884508910656302e-07,
+      "loss": 0.1034,
+      "step": 7215
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7205234624779004,
+      "learning_rate": 5.883492440429925e-07,
+      "loss": 0.4425,
+      "step": 7216
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9092036765999852,
+      "learning_rate": 5.882475932511984e-07,
+      "loss": 0.2444,
+      "step": 7217
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.966331431955261,
+      "learning_rate": 5.881459386945845e-07,
+      "loss": 0.1008,
+      "step": 7218
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7241359638013197,
+      "learning_rate": 5.880442803774877e-07,
+      "loss": 0.0397,
+      "step": 7219
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7865950024944246,
+      "learning_rate": 5.879426183042448e-07,
+      "loss": 0.3198,
+      "step": 7220
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2668959117213308,
+      "learning_rate": 5.878409524791929e-07,
+      "loss": 0.3804,
+      "step": 7221
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6940116579350895,
+      "learning_rate": 5.877392829066697e-07,
+      "loss": 0.1352,
+      "step": 7222
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7889425403151112,
+      "learning_rate": 5.876376095910122e-07,
+      "loss": 0.1604,
+      "step": 7223
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.13743757225120593,
+      "learning_rate": 5.87535932536558e-07,
+      "loss": 0.0081,
+      "step": 7224
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3539409891951253,
+      "learning_rate": 5.874342517476451e-07,
+      "loss": 0.1702,
+      "step": 7225
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3253394447168811,
+      "learning_rate": 5.873325672286112e-07,
+      "loss": 0.1449,
+      "step": 7226
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1508727438993049,
+      "learning_rate": 5.872308789837943e-07,
+      "loss": 0.0957,
+      "step": 7227
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.37757411481511055,
+      "learning_rate": 5.871291870175328e-07,
+      "loss": 0.1251,
+      "step": 7228
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.16487198167342196,
+      "learning_rate": 5.87027491334165e-07,
+      "loss": 0.0046,
+      "step": 7229
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6371992172879106,
+      "learning_rate": 5.869257919380297e-07,
+      "loss": 0.124,
+      "step": 7230
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9435011963184733,
+      "learning_rate": 5.868240888334652e-07,
+      "loss": 0.2358,
+      "step": 7231
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6291321543008145,
+      "learning_rate": 5.867223820248105e-07,
+      "loss": 0.2749,
+      "step": 7232
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6028154079885149,
+      "learning_rate": 5.866206715164047e-07,
+      "loss": 0.2447,
+      "step": 7233
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.243144110217354,
+      "learning_rate": 5.86518957312587e-07,
+      "loss": 0.0919,
+      "step": 7234
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7366644013030765,
+      "learning_rate": 5.864172394176965e-07,
+      "loss": 0.1285,
+      "step": 7235
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3381590499033993,
+      "learning_rate": 5.86315517836073e-07,
+      "loss": 0.1884,
+      "step": 7236
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9461233417057059,
+      "learning_rate": 5.862137925720559e-07,
+      "loss": 0.1757,
+      "step": 7237
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3710231970881075,
+      "learning_rate": 5.861120636299851e-07,
+      "loss": 0.2281,
+      "step": 7238
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6179070684772522,
+      "learning_rate": 5.860103310142005e-07,
+      "loss": 0.1681,
+      "step": 7239
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0201386018154825,
+      "learning_rate": 5.859085947290423e-07,
+      "loss": 0.3028,
+      "step": 7240
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8091902911533021,
+      "learning_rate": 5.858068547788509e-07,
+      "loss": 0.0124,
+      "step": 7241
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3980271825191142,
+      "learning_rate": 5.857051111679664e-07,
+      "loss": 0.3732,
+      "step": 7242
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8013402975837172,
+      "learning_rate": 5.856033639007297e-07,
+      "loss": 0.2799,
+      "step": 7243
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1687425807823621,
+      "learning_rate": 5.855016129814815e-07,
+      "loss": 0.0311,
+      "step": 7244
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3294547064513318,
+      "learning_rate": 5.853998584145624e-07,
+      "loss": 0.2819,
+      "step": 7245
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3755901930805092,
+      "learning_rate": 5.852981002043138e-07,
+      "loss": 0.3118,
+      "step": 7246
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.289201689858387,
+      "learning_rate": 5.851963383550766e-07,
+      "loss": 0.106,
+      "step": 7247
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5300436654021281,
+      "learning_rate": 5.850945728711925e-07,
+      "loss": 0.1291,
+      "step": 7248
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6562759244648393,
+      "learning_rate": 5.849928037570028e-07,
+      "loss": 0.1758,
+      "step": 7249
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.098462470269909,
+      "learning_rate": 5.848910310168493e-07,
+      "loss": 0.1566,
+      "step": 7250
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8302393051115002,
+      "learning_rate": 5.847892546550737e-07,
+      "loss": 0.128,
+      "step": 7251
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8359880903191275,
+      "learning_rate": 5.84687474676018e-07,
+      "loss": 0.3108,
+      "step": 7252
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4437891175800763,
+      "learning_rate": 5.845856910840245e-07,
+      "loss": 0.2676,
+      "step": 7253
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9646235215281507,
+      "learning_rate": 5.844839038834353e-07,
+      "loss": 0.0868,
+      "step": 7254
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6620742954436339,
+      "learning_rate": 5.84382113078593e-07,
+      "loss": 0.2169,
+      "step": 7255
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4860674138148537,
+      "learning_rate": 5.8428031867384e-07,
+      "loss": 0.1432,
+      "step": 7256
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5606129692529753,
+      "learning_rate": 5.841785206735192e-07,
+      "loss": 0.1234,
+      "step": 7257
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.094353071114604,
+      "learning_rate": 5.840767190819736e-07,
+      "loss": 0.2795,
+      "step": 7258
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9150744126339434,
+      "learning_rate": 5.839749139035461e-07,
+      "loss": 0.191,
+      "step": 7259
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3703948391791727,
+      "learning_rate": 5.8387310514258e-07,
+      "loss": 0.0749,
+      "step": 7260
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5517259013873832,
+      "learning_rate": 5.837712928034187e-07,
+      "loss": 0.2893,
+      "step": 7261
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0500803493405015,
+      "learning_rate": 5.836694768904054e-07,
+      "loss": 0.3375,
+      "step": 7262
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9423254492124644,
+      "learning_rate": 5.835676574078842e-07,
+      "loss": 0.075,
+      "step": 7263
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7692318263987405,
+      "learning_rate": 5.834658343601987e-07,
+      "loss": 0.4155,
+      "step": 7264
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9579505383096613,
+      "learning_rate": 5.833640077516929e-07,
+      "loss": 0.1745,
+      "step": 7265
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2862268054084534,
+      "learning_rate": 5.832621775867109e-07,
+      "loss": 0.1967,
+      "step": 7266
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6713444781147704,
+      "learning_rate": 5.831603438695971e-07,
+      "loss": 0.3673,
+      "step": 7267
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3249779682100575,
+      "learning_rate": 5.830585066046958e-07,
+      "loss": 0.4114,
+      "step": 7268
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7909573625344419,
+      "learning_rate": 5.829566657963517e-07,
+      "loss": 0.0779,
+      "step": 7269
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3061708633493678,
+      "learning_rate": 5.828548214489095e-07,
+      "loss": 0.1772,
+      "step": 7270
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2316863525602373,
+      "learning_rate": 5.82752973566714e-07,
+      "loss": 0.1857,
+      "step": 7271
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3739109090293889,
+      "learning_rate": 5.826511221541104e-07,
+      "loss": 0.0953,
+      "step": 7272
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.657124797840211,
+      "learning_rate": 5.825492672154437e-07,
+      "loss": 0.2445,
+      "step": 7273
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.568751310419939,
+      "learning_rate": 5.824474087550593e-07,
+      "loss": 0.1064,
+      "step": 7274
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5527706509889663,
+      "learning_rate": 5.823455467773026e-07,
+      "loss": 0.3305,
+      "step": 7275
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7951595167062507,
+      "learning_rate": 5.822436812865194e-07,
+      "loss": 0.0214,
+      "step": 7276
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.654831699863176,
+      "learning_rate": 5.821418122870556e-07,
+      "loss": 0.0929,
+      "step": 7277
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8192238158820631,
+      "learning_rate": 5.820399397832568e-07,
+      "loss": 0.399,
+      "step": 7278
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6710596774381883,
+      "learning_rate": 5.819380637794693e-07,
+      "loss": 0.2271,
+      "step": 7279
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0295387864219177,
+      "learning_rate": 5.818361842800392e-07,
+      "loss": 0.2058,
+      "step": 7280
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4595757345400026,
+      "learning_rate": 5.817343012893131e-07,
+      "loss": 0.0978,
+      "step": 7281
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1208460280131833,
+      "learning_rate": 5.816324148116374e-07,
+      "loss": 0.2242,
+      "step": 7282
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7370933209663031,
+      "learning_rate": 5.815305248513587e-07,
+      "loss": 0.209,
+      "step": 7283
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1254594372749078,
+      "learning_rate": 5.814286314128238e-07,
+      "loss": 0.1657,
+      "step": 7284
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5784525671144114,
+      "learning_rate": 5.8132673450038e-07,
+      "loss": 0.0118,
+      "step": 7285
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0887737723803113,
+      "learning_rate": 5.812248341183741e-07,
+      "loss": 0.1122,
+      "step": 7286
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3398571182371665,
+      "learning_rate": 5.811229302711536e-07,
+      "loss": 0.0988,
+      "step": 7287
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29411231775771174,
+      "learning_rate": 5.810210229630657e-07,
+      "loss": 0.029,
+      "step": 7288
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6714221210862924,
+      "learning_rate": 5.809191121984582e-07,
+      "loss": 0.0151,
+      "step": 7289
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7738093846839272,
+      "learning_rate": 5.808171979816786e-07,
+      "loss": 0.4841,
+      "step": 7290
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2247291459538128,
+      "learning_rate": 5.80715280317075e-07,
+      "loss": 0.1818,
+      "step": 7291
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17260494263047638,
+      "learning_rate": 5.80613359208995e-07,
+      "loss": 0.1391,
+      "step": 7292
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.147421161271533,
+      "learning_rate": 5.805114346617873e-07,
+      "loss": 0.0911,
+      "step": 7293
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.350897262005696,
+      "learning_rate": 5.804095066797999e-07,
+      "loss": 0.1607,
+      "step": 7294
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4318251929458587,
+      "learning_rate": 5.803075752673812e-07,
+      "loss": 0.025,
+      "step": 7295
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6625797293507669,
+      "learning_rate": 5.802056404288801e-07,
+      "loss": 0.009,
+      "step": 7296
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2150155533447429,
+      "learning_rate": 5.80103702168645e-07,
+      "loss": 0.348,
+      "step": 7297
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7820606372820846,
+      "learning_rate": 5.80001760491025e-07,
+      "loss": 0.3305,
+      "step": 7298
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3395156167890347,
+      "learning_rate": 5.798998154003691e-07,
+      "loss": 0.1866,
+      "step": 7299
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.583684948082188,
+      "learning_rate": 5.797978669010264e-07,
+      "loss": 0.1602,
+      "step": 7300
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7808833631802821,
+      "learning_rate": 5.796959149973463e-07,
+      "loss": 0.2628,
+      "step": 7301
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4931297365788696,
+      "learning_rate": 5.795939596936782e-07,
+      "loss": 0.0382,
+      "step": 7302
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.46770785079811705,
+      "learning_rate": 5.794920009943719e-07,
+      "loss": 0.1639,
+      "step": 7303
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5700771055707824,
+      "learning_rate": 5.793900389037769e-07,
+      "loss": 0.2143,
+      "step": 7304
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.563425993375374,
+      "learning_rate": 5.792880734262433e-07,
+      "loss": 0.2519,
+      "step": 7305
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.458020840421295,
+      "learning_rate": 5.791861045661211e-07,
+      "loss": 0.3534,
+      "step": 7306
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7477225624836562,
+      "learning_rate": 5.790841323277606e-07,
+      "loss": 0.1822,
+      "step": 7307
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1290463820403616,
+      "learning_rate": 5.789821567155119e-07,
+      "loss": 0.2205,
+      "step": 7308
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7053548395714695,
+      "learning_rate": 5.788801777337256e-07,
+      "loss": 0.1834,
+      "step": 7309
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.024969451501184,
+      "learning_rate": 5.787781953867523e-07,
+      "loss": 0.247,
+      "step": 7310
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7774324720546493,
+      "learning_rate": 5.78676209678943e-07,
+      "loss": 0.2281,
+      "step": 7311
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5249247474929776,
+      "learning_rate": 5.785742206146483e-07,
+      "loss": 0.2471,
+      "step": 7312
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.47158116919704907,
+      "learning_rate": 5.784722281982196e-07,
+      "loss": 0.1751,
+      "step": 7313
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7358734719182072,
+      "learning_rate": 5.783702324340078e-07,
+      "loss": 0.1933,
+      "step": 7314
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3645172137192145,
+      "learning_rate": 5.782682333263643e-07,
+      "loss": 0.1133,
+      "step": 7315
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.867536241972235,
+      "learning_rate": 5.781662308796406e-07,
+      "loss": 0.0963,
+      "step": 7316
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6811480977103188,
+      "learning_rate": 5.780642250981884e-07,
+      "loss": 0.4057,
+      "step": 7317
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.681771414756409,
+      "learning_rate": 5.779622159863593e-07,
+      "loss": 0.0947,
+      "step": 7318
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5311426383320308,
+      "learning_rate": 5.778602035485054e-07,
+      "loss": 0.0439,
+      "step": 7319
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0074352489639262,
+      "learning_rate": 5.777581877889787e-07,
+      "loss": 0.3747,
+      "step": 7320
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7402048501721525,
+      "learning_rate": 5.776561687121315e-07,
+      "loss": 0.3483,
+      "step": 7321
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7648789389158936,
+      "learning_rate": 5.77554146322316e-07,
+      "loss": 0.331,
+      "step": 7322
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.650351293264006,
+      "learning_rate": 5.774521206238847e-07,
+      "loss": 0.1978,
+      "step": 7323
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.773258554706342,
+      "learning_rate": 5.773500916211902e-07,
+      "loss": 0.2602,
+      "step": 7324
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.148778741358844,
+      "learning_rate": 5.772480593185853e-07,
+      "loss": 0.0301,
+      "step": 7325
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.766184658234635,
+      "learning_rate": 5.771460237204229e-07,
+      "loss": 0.2557,
+      "step": 7326
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.39052958858279657,
+      "learning_rate": 5.770439848310562e-07,
+      "loss": 0.1391,
+      "step": 7327
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5991400704799148,
+      "learning_rate": 5.769419426548381e-07,
+      "loss": 0.1637,
+      "step": 7328
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5392510406057364,
+      "learning_rate": 5.76839897196122e-07,
+      "loss": 0.1154,
+      "step": 7329
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5869935013530925,
+      "learning_rate": 5.767378484592616e-07,
+      "loss": 0.2897,
+      "step": 7330
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.40047283983111004,
+      "learning_rate": 5.766357964486102e-07,
+      "loss": 0.2059,
+      "step": 7331
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2511239699126518,
+      "learning_rate": 5.765337411685216e-07,
+      "loss": 0.1408,
+      "step": 7332
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6251798292657806,
+      "learning_rate": 5.764316826233498e-07,
+      "loss": 0.2621,
+      "step": 7333
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5509022837755441,
+      "learning_rate": 5.763296208174488e-07,
+      "loss": 0.2851,
+      "step": 7334
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8798314347754198,
+      "learning_rate": 5.762275557551726e-07,
+      "loss": 0.3619,
+      "step": 7335
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1918609967717406,
+      "learning_rate": 5.761254874408759e-07,
+      "loss": 0.0782,
+      "step": 7336
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9119931217971832,
+      "learning_rate": 5.760234158789126e-07,
+      "loss": 0.4166,
+      "step": 7337
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8541241720669797,
+      "learning_rate": 5.759213410736376e-07,
+      "loss": 0.0546,
+      "step": 7338
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6403395795517418,
+      "learning_rate": 5.758192630294058e-07,
+      "loss": 0.2085,
+      "step": 7339
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6724951175395882,
+      "learning_rate": 5.757171817505716e-07,
+      "loss": 0.2122,
+      "step": 7340
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7958559112389705,
+      "learning_rate": 5.756150972414903e-07,
+      "loss": 0.0467,
+      "step": 7341
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7387496108044898,
+      "learning_rate": 5.75513009506517e-07,
+      "loss": 0.1968,
+      "step": 7342
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9963018733725529,
+      "learning_rate": 5.754109185500069e-07,
+      "loss": 0.3962,
+      "step": 7343
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5125945984411493,
+      "learning_rate": 5.753088243763153e-07,
+      "loss": 0.061,
+      "step": 7344
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.129892852925207,
+      "learning_rate": 5.752067269897979e-07,
+      "loss": 0.0721,
+      "step": 7345
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7931316184757881,
+      "learning_rate": 5.751046263948104e-07,
+      "loss": 0.0084,
+      "step": 7346
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9499235591276186,
+      "learning_rate": 5.750025225957085e-07,
+      "loss": 0.1203,
+      "step": 7347
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.563436624583075,
+      "learning_rate": 5.749004155968482e-07,
+      "loss": 0.1158,
+      "step": 7348
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8241276446340091,
+      "learning_rate": 5.747983054025856e-07,
+      "loss": 0.2044,
+      "step": 7349
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.101218515866801,
+      "learning_rate": 5.746961920172771e-07,
+      "loss": 0.1827,
+      "step": 7350
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3685416761720732,
+      "learning_rate": 5.745940754452787e-07,
+      "loss": 0.0704,
+      "step": 7351
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.328030911984748,
+      "learning_rate": 5.744919556909472e-07,
+      "loss": 0.2337,
+      "step": 7352
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5497695070730573,
+      "learning_rate": 5.74389832758639e-07,
+      "loss": 0.1309,
+      "step": 7353
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7591838272308836,
+      "learning_rate": 5.742877066527112e-07,
+      "loss": 0.0285,
+      "step": 7354
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8302343064697865,
+      "learning_rate": 5.741855773775204e-07,
+      "loss": 0.2295,
+      "step": 7355
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5090856157851922,
+      "learning_rate": 5.740834449374237e-07,
+      "loss": 0.2969,
+      "step": 7356
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2033450691961722,
+      "learning_rate": 5.739813093367783e-07,
+      "loss": 0.0484,
+      "step": 7357
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.584509287221558,
+      "learning_rate": 5.738791705799415e-07,
+      "loss": 0.1798,
+      "step": 7358
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4247587619689443,
+      "learning_rate": 5.737770286712708e-07,
+      "loss": 0.285,
+      "step": 7359
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.6729317035023294,
+      "learning_rate": 5.736748836151237e-07,
+      "loss": 0.1208,
+      "step": 7360
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6018869856424918,
+      "learning_rate": 5.73572735415858e-07,
+      "loss": 0.1735,
+      "step": 7361
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0309512263073584,
+      "learning_rate": 5.734705840778315e-07,
+      "loss": 0.5402,
+      "step": 7362
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3507469001297476,
+      "learning_rate": 5.733684296054022e-07,
+      "loss": 0.1199,
+      "step": 7363
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.592307700239587,
+      "learning_rate": 5.732662720029282e-07,
+      "loss": 0.1362,
+      "step": 7364
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3931602978053745,
+      "learning_rate": 5.731641112747679e-07,
+      "loss": 0.1187,
+      "step": 7365
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1124017828310615,
+      "learning_rate": 5.730619474252792e-07,
+      "loss": 0.1946,
+      "step": 7366
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5105519902196064,
+      "learning_rate": 5.729597804588212e-07,
+      "loss": 0.2302,
+      "step": 7367
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2018239662291923,
+      "learning_rate": 5.728576103797524e-07,
+      "loss": 0.4881,
+      "step": 7368
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1376805720728558,
+      "learning_rate": 5.727554371924313e-07,
+      "loss": 0.3844,
+      "step": 7369
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6324332393685804,
+      "learning_rate": 5.72653260901217e-07,
+      "loss": 0.4303,
+      "step": 7370
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.65881572345634,
+      "learning_rate": 5.725510815104685e-07,
+      "loss": 0.0819,
+      "step": 7371
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7395975743305332,
+      "learning_rate": 5.724488990245451e-07,
+      "loss": 0.1954,
+      "step": 7372
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9277443571557993,
+      "learning_rate": 5.723467134478059e-07,
+      "loss": 0.2163,
+      "step": 7373
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9069014862888083,
+      "learning_rate": 5.722445247846106e-07,
+      "loss": 0.2253,
+      "step": 7374
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.11844499542924179,
+      "learning_rate": 5.721423330393187e-07,
+      "loss": 0.0032,
+      "step": 7375
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8113056982316724,
+      "learning_rate": 5.720401382162898e-07,
+      "loss": 0.1503,
+      "step": 7376
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9212818922283537,
+      "learning_rate": 5.719379403198837e-07,
+      "loss": 0.1709,
+      "step": 7377
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9404735582256118,
+      "learning_rate": 5.718357393544605e-07,
+      "loss": 0.1065,
+      "step": 7378
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4801437221205867,
+      "learning_rate": 5.717335353243802e-07,
+      "loss": 0.1478,
+      "step": 7379
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4510494955548074,
+      "learning_rate": 5.716313282340032e-07,
+      "loss": 0.159,
+      "step": 7380
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5956475286412855,
+      "learning_rate": 5.715291180876896e-07,
+      "loss": 0.1442,
+      "step": 7381
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.172564143546517,
+      "learning_rate": 5.714269048898002e-07,
+      "loss": 0.0202,
+      "step": 7382
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.574979745638005,
+      "learning_rate": 5.713246886446953e-07,
+      "loss": 0.0843,
+      "step": 7383
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9970643512039582,
+      "learning_rate": 5.712224693567358e-07,
+      "loss": 0.1385,
+      "step": 7384
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6911268598909803,
+      "learning_rate": 5.711202470302827e-07,
+      "loss": 0.2269,
+      "step": 7385
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.80939766757664,
+      "learning_rate": 5.710180216696968e-07,
+      "loss": 0.3549,
+      "step": 7386
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.610968088004343,
+      "learning_rate": 5.709157932793394e-07,
+      "loss": 0.0185,
+      "step": 7387
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7329459952435532,
+      "learning_rate": 5.708135618635717e-07,
+      "loss": 0.1785,
+      "step": 7388
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8214124310681195,
+      "learning_rate": 5.70711327426755e-07,
+      "loss": 0.2332,
+      "step": 7389
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.653476522592435,
+      "learning_rate": 5.706090899732508e-07,
+      "loss": 0.1851,
+      "step": 7390
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8452045293615673,
+      "learning_rate": 5.705068495074211e-07,
+      "loss": 0.1723,
+      "step": 7391
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8629686321621239,
+      "learning_rate": 5.704046060336275e-07,
+      "loss": 0.2495,
+      "step": 7392
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.380436473286159,
+      "learning_rate": 5.703023595562318e-07,
+      "loss": 0.2289,
+      "step": 7393
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4993895908190598,
+      "learning_rate": 5.702001100795961e-07,
+      "loss": 0.2069,
+      "step": 7394
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.507355922001305,
+      "learning_rate": 5.700978576080826e-07,
+      "loss": 0.2583,
+      "step": 7395
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.514544843370766,
+      "learning_rate": 5.699956021460537e-07,
+      "loss": 0.0457,
+      "step": 7396
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0435372857076322,
+      "learning_rate": 5.698933436978715e-07,
+      "loss": 0.4678,
+      "step": 7397
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2240897383019091,
+      "learning_rate": 5.697910822678988e-07,
+      "loss": 0.2138,
+      "step": 7398
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.75916899659083,
+      "learning_rate": 5.696888178604982e-07,
+      "loss": 0.2188,
+      "step": 7399
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7958821348665706,
+      "learning_rate": 5.695865504800327e-07,
+      "loss": 0.2863,
+      "step": 7400
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9350666028729014,
+      "learning_rate": 5.694842801308651e-07,
+      "loss": 0.1918,
+      "step": 7401
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0548114986016022,
+      "learning_rate": 5.693820068173583e-07,
+      "loss": 0.2893,
+      "step": 7402
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5524496773427505,
+      "learning_rate": 5.692797305438756e-07,
+      "loss": 0.3367,
+      "step": 7403
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6890498424681148,
+      "learning_rate": 5.691774513147802e-07,
+      "loss": 0.2766,
+      "step": 7404
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2570451277405057,
+      "learning_rate": 5.690751691344359e-07,
+      "loss": 0.3281,
+      "step": 7405
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7342886843729234,
+      "learning_rate": 5.689728840072059e-07,
+      "loss": 0.1861,
+      "step": 7406
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2393823786284757,
+      "learning_rate": 5.688705959374542e-07,
+      "loss": 0.1758,
+      "step": 7407
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6368756048863157,
+      "learning_rate": 5.687683049295441e-07,
+      "loss": 0.1287,
+      "step": 7408
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3698230996427294,
+      "learning_rate": 5.686660109878401e-07,
+      "loss": 0.1459,
+      "step": 7409
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0428440778245631,
+      "learning_rate": 5.68563714116706e-07,
+      "loss": 0.1315,
+      "step": 7410
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4462005025358093,
+      "learning_rate": 5.68461414320506e-07,
+      "loss": 0.2594,
+      "step": 7411
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4810305048733831,
+      "learning_rate": 5.683591116036045e-07,
+      "loss": 0.2054,
+      "step": 7412
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.091110397138657,
+      "learning_rate": 5.682568059703659e-07,
+      "loss": 0.1844,
+      "step": 7413
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.111523344208612,
+      "learning_rate": 5.681544974251547e-07,
+      "loss": 0.1186,
+      "step": 7414
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5531265967104229,
+      "learning_rate": 5.680521859723355e-07,
+      "loss": 0.201,
+      "step": 7415
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.36396224093760593,
+      "learning_rate": 5.679498716162733e-07,
+      "loss": 0.1059,
+      "step": 7416
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.232194012764126,
+      "learning_rate": 5.67847554361333e-07,
+      "loss": 0.2394,
+      "step": 7417
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.33948903903857913,
+      "learning_rate": 5.677452342118797e-07,
+      "loss": 0.0658,
+      "step": 7418
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9176724911364541,
+      "learning_rate": 5.676429111722784e-07,
+      "loss": 0.0953,
+      "step": 7419
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0441066460826038,
+      "learning_rate": 5.675405852468948e-07,
+      "loss": 0.1832,
+      "step": 7420
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7920029983208265,
+      "learning_rate": 5.674382564400938e-07,
+      "loss": 0.2583,
+      "step": 7421
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.048901779663038,
+      "learning_rate": 5.673359247562412e-07,
+      "loss": 0.257,
+      "step": 7422
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.615553286073353,
+      "learning_rate": 5.67233590199703e-07,
+      "loss": 0.0901,
+      "step": 7423
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7062176481074847,
+      "learning_rate": 5.671312527748444e-07,
+      "loss": 0.0728,
+      "step": 7424
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.232755665082155,
+      "learning_rate": 5.670289124860317e-07,
+      "loss": 0.2293,
+      "step": 7425
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.405574665183514,
+      "learning_rate": 5.669265693376309e-07,
+      "loss": 0.0083,
+      "step": 7426
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1098301197267593,
+      "learning_rate": 5.66824223334008e-07,
+      "loss": 0.3068,
+      "step": 7427
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9236546297958498,
+      "learning_rate": 5.667218744795293e-07,
+      "loss": 0.2577,
+      "step": 7428
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8847137481375129,
+      "learning_rate": 5.666195227785615e-07,
+      "loss": 0.1334,
+      "step": 7429
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5823681045348512,
+      "learning_rate": 5.665171682354709e-07,
+      "loss": 0.1977,
+      "step": 7430
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3705110138490335,
+      "learning_rate": 5.664148108546242e-07,
+      "loss": 0.0945,
+      "step": 7431
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1590817549719374,
+      "learning_rate": 5.663124506403881e-07,
+      "loss": 0.3621,
+      "step": 7432
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.905325163363612,
+      "learning_rate": 5.662100875971297e-07,
+      "loss": 0.1864,
+      "step": 7433
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0797349867376271,
+      "learning_rate": 5.661077217292155e-07,
+      "loss": 0.2406,
+      "step": 7434
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.195393183468775,
+      "learning_rate": 5.660053530410132e-07,
+      "loss": 0.2057,
+      "step": 7435
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.40924590451085185,
+      "learning_rate": 5.6590298153689e-07,
+      "loss": 0.2426,
+      "step": 7436
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6132333443663023,
+      "learning_rate": 5.658006072212132e-07,
+      "loss": 0.1041,
+      "step": 7437
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.378726814524991,
+      "learning_rate": 5.656982300983499e-07,
+      "loss": 0.2694,
+      "step": 7438
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.546295418977469,
+      "learning_rate": 5.655958501726682e-07,
+      "loss": 0.2877,
+      "step": 7439
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9890241239705159,
+      "learning_rate": 5.654934674485356e-07,
+      "loss": 0.1202,
+      "step": 7440
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5518453900124864,
+      "learning_rate": 5.653910819303202e-07,
+      "loss": 0.1582,
+      "step": 7441
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.31820669561369536,
+      "learning_rate": 5.652886936223896e-07,
+      "loss": 0.1387,
+      "step": 7442
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.194544907467919,
+      "learning_rate": 5.65186302529112e-07,
+      "loss": 0.5461,
+      "step": 7443
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3864319810046312,
+      "learning_rate": 5.650839086548559e-07,
+      "loss": 0.2458,
+      "step": 7444
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2756607149434038,
+      "learning_rate": 5.649815120039894e-07,
+      "loss": 0.0146,
+      "step": 7445
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.374367894987499,
+      "learning_rate": 5.648791125808808e-07,
+      "loss": 0.1254,
+      "step": 7446
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6856197673962292,
+      "learning_rate": 5.647767103898989e-07,
+      "loss": 0.093,
+      "step": 7447
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9346600309424645,
+      "learning_rate": 5.646743054354123e-07,
+      "loss": 0.2221,
+      "step": 7448
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.312934572514989,
+      "learning_rate": 5.6457189772179e-07,
+      "loss": 0.1119,
+      "step": 7449
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8698171801457183,
+      "learning_rate": 5.644694872534007e-07,
+      "loss": 0.3173,
+      "step": 7450
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5585476449319329,
+      "learning_rate": 5.643670740346134e-07,
+      "loss": 0.3569,
+      "step": 7451
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1521584245809629,
+      "learning_rate": 5.642646580697973e-07,
+      "loss": 0.0535,
+      "step": 7452
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.824701361709204,
+      "learning_rate": 5.641622393633217e-07,
+      "loss": 0.251,
+      "step": 7453
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4205844762570796,
+      "learning_rate": 5.640598179195561e-07,
+      "loss": 0.1754,
+      "step": 7454
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8573693617300268,
+      "learning_rate": 5.639573937428698e-07,
+      "loss": 0.1083,
+      "step": 7455
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1005043004659185,
+      "learning_rate": 5.638549668376325e-07,
+      "loss": 0.1672,
+      "step": 7456
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6114767191442775,
+      "learning_rate": 5.637525372082139e-07,
+      "loss": 0.1807,
+      "step": 7457
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8702120397028751,
+      "learning_rate": 5.63650104858984e-07,
+      "loss": 0.1615,
+      "step": 7458
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.750906079323879,
+      "learning_rate": 5.635476697943127e-07,
+      "loss": 0.1844,
+      "step": 7459
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4672155006839803,
+      "learning_rate": 5.6344523201857e-07,
+      "loss": 0.0849,
+      "step": 7460
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.31078634128353105,
+      "learning_rate": 5.633427915361261e-07,
+      "loss": 0.115,
+      "step": 7461
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.36883160702436,
+      "learning_rate": 5.632403483513514e-07,
+      "loss": 0.0598,
+      "step": 7462
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7199565022113346,
+      "learning_rate": 5.631379024686163e-07,
+      "loss": 0.0862,
+      "step": 7463
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9325619587341416,
+      "learning_rate": 5.630354538922914e-07,
+      "loss": 0.3025,
+      "step": 7464
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7220012421406382,
+      "learning_rate": 5.629330026267474e-07,
+      "loss": 0.1827,
+      "step": 7465
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4894598119503782,
+      "learning_rate": 5.628305486763551e-07,
+      "loss": 0.1976,
+      "step": 7466
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.740669590038806,
+      "learning_rate": 5.627280920454851e-07,
+      "loss": 0.2961,
+      "step": 7467
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1176548194685672,
+      "learning_rate": 5.626256327385086e-07,
+      "loss": 0.128,
+      "step": 7468
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.31588475411744266,
+      "learning_rate": 5.625231707597966e-07,
+      "loss": 0.0828,
+      "step": 7469
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7131728843382178,
+      "learning_rate": 5.624207061137205e-07,
+      "loss": 0.158,
+      "step": 7470
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.46312935719627,
+      "learning_rate": 5.623182388046516e-07,
+      "loss": 0.1385,
+      "step": 7471
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1777465574904347,
+      "learning_rate": 5.622157688369615e-07,
+      "loss": 0.2871,
+      "step": 7472
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.688964848925245,
+      "learning_rate": 5.621132962150216e-07,
+      "loss": 0.0781,
+      "step": 7473
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5968462726565935,
+      "learning_rate": 5.620108209432036e-07,
+      "loss": 0.1112,
+      "step": 7474
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3971323826497248,
+      "learning_rate": 5.619083430258793e-07,
+      "loss": 0.1836,
+      "step": 7475
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0694918118281551,
+      "learning_rate": 5.618058624674207e-07,
+      "loss": 0.1519,
+      "step": 7476
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0335790561850045,
+      "learning_rate": 5.617033792721997e-07,
+      "loss": 0.2359,
+      "step": 7477
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4831809868341667,
+      "learning_rate": 5.616008934445883e-07,
+      "loss": 0.2011,
+      "step": 7478
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4898245992729707,
+      "learning_rate": 5.614984049889593e-07,
+      "loss": 0.1118,
+      "step": 7479
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.1437450886365275,
+      "learning_rate": 5.613959139096845e-07,
+      "loss": 0.3033,
+      "step": 7480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7385254767518705,
+      "learning_rate": 5.612934202111367e-07,
+      "loss": 0.2655,
+      "step": 7481
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3418980351521246,
+      "learning_rate": 5.611909238976884e-07,
+      "loss": 0.2096,
+      "step": 7482
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2870817067320607,
+      "learning_rate": 5.610884249737121e-07,
+      "loss": 0.1744,
+      "step": 7483
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0469729269311743,
+      "learning_rate": 5.60985923443581e-07,
+      "loss": 0.1054,
+      "step": 7484
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2029980528324711,
+      "learning_rate": 5.608834193116677e-07,
+      "loss": 0.1731,
+      "step": 7485
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7293884920226757,
+      "learning_rate": 5.607809125823453e-07,
+      "loss": 0.0062,
+      "step": 7486
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 20.011926147180226,
+      "learning_rate": 5.606784032599869e-07,
+      "loss": 0.1207,
+      "step": 7487
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6523455597712159,
+      "learning_rate": 5.60575891348966e-07,
+      "loss": 0.2639,
+      "step": 7488
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5988232100390637,
+      "learning_rate": 5.604733768536559e-07,
+      "loss": 0.2087,
+      "step": 7489
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1281720086035099,
+      "learning_rate": 5.603708597784298e-07,
+      "loss": 0.0091,
+      "step": 7490
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5289742693658647,
+      "learning_rate": 5.602683401276614e-07,
+      "loss": 0.1171,
+      "step": 7491
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1895943799066078,
+      "learning_rate": 5.601658179057247e-07,
+      "loss": 0.1158,
+      "step": 7492
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8388780051991538,
+      "learning_rate": 5.600632931169932e-07,
+      "loss": 0.3189,
+      "step": 7493
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.686891110765236,
+      "learning_rate": 5.599607657658408e-07,
+      "loss": 0.0356,
+      "step": 7494
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6788510211169609,
+      "learning_rate": 5.598582358566415e-07,
+      "loss": 0.2257,
+      "step": 7495
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.364153192213852,
+      "learning_rate": 5.597557033937697e-07,
+      "loss": 0.2356,
+      "step": 7496
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.37181014590039935,
+      "learning_rate": 5.596531683815992e-07,
+      "loss": 0.1165,
+      "step": 7497
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7648144265397897,
+      "learning_rate": 5.595506308245049e-07,
+      "loss": 0.3077,
+      "step": 7498
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4327140745052023,
+      "learning_rate": 5.594480907268609e-07,
+      "loss": 0.2935,
+      "step": 7499
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2451480230515064,
+      "learning_rate": 5.593455480930418e-07,
+      "loss": 0.3365,
+      "step": 7500
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.915083767056461,
+      "learning_rate": 5.592430029274224e-07,
+      "loss": 0.0809,
+      "step": 7501
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9222138475209727,
+      "learning_rate": 5.591404552343774e-07,
+      "loss": 0.3594,
+      "step": 7502
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.32992894311582377,
+      "learning_rate": 5.590379050182817e-07,
+      "loss": 0.0914,
+      "step": 7503
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.206815532507948,
+      "learning_rate": 5.589353522835102e-07,
+      "loss": 0.112,
+      "step": 7504
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8363646515703453,
+      "learning_rate": 5.58832797034438e-07,
+      "loss": 0.2132,
+      "step": 7505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8144334905459782,
+      "learning_rate": 5.587302392754407e-07,
+      "loss": 0.3436,
+      "step": 7506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8848522820728025,
+      "learning_rate": 5.586276790108931e-07,
+      "loss": 0.1702,
+      "step": 7507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.25257579505733574,
+      "learning_rate": 5.585251162451709e-07,
+      "loss": 0.0853,
+      "step": 7508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3327463292421773,
+      "learning_rate": 5.584225509826497e-07,
+      "loss": 0.0967,
+      "step": 7509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3346992988258828,
+      "learning_rate": 5.583199832277049e-07,
+      "loss": 0.1898,
+      "step": 7510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4112276192280022,
+      "learning_rate": 5.582174129847125e-07,
+      "loss": 0.1245,
+      "step": 7511
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8455868299535237,
+      "learning_rate": 5.581148402580481e-07,
+      "loss": 0.0945,
+      "step": 7512
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7961139849131209,
+      "learning_rate": 5.580122650520879e-07,
+      "loss": 0.2118,
+      "step": 7513
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9728331830756659,
+      "learning_rate": 5.579096873712077e-07,
+      "loss": 0.2882,
+      "step": 7514
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.145700284257323,
+      "learning_rate": 5.578071072197839e-07,
+      "loss": 0.1685,
+      "step": 7515
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4719600960107473,
+      "learning_rate": 5.577045246021928e-07,
+      "loss": 0.1886,
+      "step": 7516
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9118244024819626,
+      "learning_rate": 5.576019395228106e-07,
+      "loss": 0.0788,
+      "step": 7517
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9239190246386999,
+      "learning_rate": 5.574993519860138e-07,
+      "loss": 0.3041,
+      "step": 7518
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6822214387634311,
+      "learning_rate": 5.573967619961791e-07,
+      "loss": 0.1335,
+      "step": 7519
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0513784452022978,
+      "learning_rate": 5.572941695576834e-07,
+      "loss": 0.0111,
+      "step": 7520
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7666833761008804,
+      "learning_rate": 5.57191574674903e-07,
+      "loss": 0.1138,
+      "step": 7521
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6404449871594167,
+      "learning_rate": 5.570889773522149e-07,
+      "loss": 0.1827,
+      "step": 7522
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8274110168994253,
+      "learning_rate": 5.569863775939965e-07,
+      "loss": 0.2472,
+      "step": 7523
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0695101580292428,
+      "learning_rate": 5.568837754046246e-07,
+      "loss": 0.2303,
+      "step": 7524
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1625372819514135,
+      "learning_rate": 5.567811707884765e-07,
+      "loss": 0.2299,
+      "step": 7525
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.239994514135691,
+      "learning_rate": 5.566785637499296e-07,
+      "loss": 0.3032,
+      "step": 7526
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.845062834876523,
+      "learning_rate": 5.565759542933611e-07,
+      "loss": 0.2958,
+      "step": 7527
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.90402531981841,
+      "learning_rate": 5.564733424231487e-07,
+      "loss": 0.4111,
+      "step": 7528
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1355457669657243,
+      "learning_rate": 5.5637072814367e-07,
+      "loss": 0.379,
+      "step": 7529
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1256920054797275,
+      "learning_rate": 5.562681114593028e-07,
+      "loss": 0.1953,
+      "step": 7530
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9255408113191933,
+      "learning_rate": 5.561654923744248e-07,
+      "loss": 0.306,
+      "step": 7531
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6651820663004722,
+      "learning_rate": 5.560628708934138e-07,
+      "loss": 0.2565,
+      "step": 7532
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2996596651136072,
+      "learning_rate": 5.559602470206483e-07,
+      "loss": 0.4174,
+      "step": 7533
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.48131214773171593,
+      "learning_rate": 5.55857620760506e-07,
+      "loss": 0.2003,
+      "step": 7534
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0320814499631323,
+      "learning_rate": 5.557549921173655e-07,
+      "loss": 0.5093,
+      "step": 7535
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0990309741661553,
+      "learning_rate": 5.556523610956047e-07,
+      "loss": 0.2306,
+      "step": 7536
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8260995074990696,
+      "learning_rate": 5.555497276996024e-07,
+      "loss": 0.1701,
+      "step": 7537
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1976736039695113,
+      "learning_rate": 5.554470919337372e-07,
+      "loss": 0.2351,
+      "step": 7538
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.40345372153506504,
+      "learning_rate": 5.553444538023873e-07,
+      "loss": 0.0877,
+      "step": 7539
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5977917822238236,
+      "learning_rate": 5.55241813309932e-07,
+      "loss": 0.0766,
+      "step": 7540
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7745418848440369,
+      "learning_rate": 5.551391704607497e-07,
+      "loss": 0.2889,
+      "step": 7541
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.42312695308428366,
+      "learning_rate": 5.550365252592196e-07,
+      "loss": 0.1046,
+      "step": 7542
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6641934247142979,
+      "learning_rate": 5.549338777097208e-07,
+      "loss": 0.3636,
+      "step": 7543
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.952245733703652,
+      "learning_rate": 5.548312278166322e-07,
+      "loss": 0.2157,
+      "step": 7544
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7019529118628163,
+      "learning_rate": 5.547285755843334e-07,
+      "loss": 0.2253,
+      "step": 7545
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9408301507018305,
+      "learning_rate": 5.546259210172034e-07,
+      "loss": 0.2198,
+      "step": 7546
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9697559515545209,
+      "learning_rate": 5.54523264119622e-07,
+      "loss": 0.1452,
+      "step": 7547
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.314768569281151,
+      "learning_rate": 5.544206048959682e-07,
+      "loss": 0.2431,
+      "step": 7548
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7991822471163085,
+      "learning_rate": 5.543179433506222e-07,
+      "loss": 0.0965,
+      "step": 7549
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7061729152233741,
+      "learning_rate": 5.542152794879636e-07,
+      "loss": 0.2957,
+      "step": 7550
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5756950985548269,
+      "learning_rate": 5.54112613312372e-07,
+      "loss": 0.256,
+      "step": 7551
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2867756764185683,
+      "learning_rate": 5.540099448282276e-07,
+      "loss": 0.0989,
+      "step": 7552
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6086105040269305,
+      "learning_rate": 5.539072740399104e-07,
+      "loss": 0.112,
+      "step": 7553
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.138626460754883,
+      "learning_rate": 5.538046009518006e-07,
+      "loss": 0.059,
+      "step": 7554
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6693955455920618,
+      "learning_rate": 5.537019255682783e-07,
+      "loss": 0.3693,
+      "step": 7555
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.216398338530876,
+      "learning_rate": 5.535992478937239e-07,
+      "loss": 0.2422,
+      "step": 7556
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6205476139449395,
+      "learning_rate": 5.534965679325179e-07,
+      "loss": 0.1238,
+      "step": 7557
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2505535639024559,
+      "learning_rate": 5.533938856890407e-07,
+      "loss": 0.0778,
+      "step": 7558
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7440686785079905,
+      "learning_rate": 5.532912011676729e-07,
+      "loss": 0.0899,
+      "step": 7559
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.782080365904402,
+      "learning_rate": 5.531885143727955e-07,
+      "loss": 0.09,
+      "step": 7560
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9205299052777522,
+      "learning_rate": 5.530858253087891e-07,
+      "loss": 0.1384,
+      "step": 7561
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0070649464553347,
+      "learning_rate": 5.529831339800348e-07,
+      "loss": 0.2301,
+      "step": 7562
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.642698626302945,
+      "learning_rate": 5.528804403909133e-07,
+      "loss": 0.1444,
+      "step": 7563
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5385271444594335,
+      "learning_rate": 5.52777744545806e-07,
+      "loss": 0.1689,
+      "step": 7564
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6910624278217642,
+      "learning_rate": 5.52675046449094e-07,
+      "loss": 0.1171,
+      "step": 7565
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5188507620744114,
+      "learning_rate": 5.525723461051587e-07,
+      "loss": 0.118,
+      "step": 7566
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7215030641675393,
+      "learning_rate": 5.524696435183812e-07,
+      "loss": 0.1521,
+      "step": 7567
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27550035434511455,
+      "learning_rate": 5.523669386931433e-07,
+      "loss": 0.0337,
+      "step": 7568
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.45155692176302054,
+      "learning_rate": 5.522642316338268e-07,
+      "loss": 0.0892,
+      "step": 7569
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5821565862132517,
+      "learning_rate": 5.521615223448129e-07,
+      "loss": 0.2217,
+      "step": 7570
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.732416575891742,
+      "learning_rate": 5.520588108304836e-07,
+      "loss": 0.2106,
+      "step": 7571
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.594161838262372,
+      "learning_rate": 5.519560970952207e-07,
+      "loss": 0.0865,
+      "step": 7572
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4948336791965397,
+      "learning_rate": 5.518533811434064e-07,
+      "loss": 0.2018,
+      "step": 7573
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0229297133671198,
+      "learning_rate": 5.517506629794226e-07,
+      "loss": 0.2161,
+      "step": 7574
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6807350704565484,
+      "learning_rate": 5.516479426076515e-07,
+      "loss": 0.1555,
+      "step": 7575
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.0254761901459135,
+      "learning_rate": 5.515452200324753e-07,
+      "loss": 0.1612,
+      "step": 7576
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.57679777406315,
+      "learning_rate": 5.514424952582765e-07,
+      "loss": 0.2426,
+      "step": 7577
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8858005241840776,
+      "learning_rate": 5.513397682894373e-07,
+      "loss": 0.2008,
+      "step": 7578
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6883412880363127,
+      "learning_rate": 5.512370391303404e-07,
+      "loss": 0.2956,
+      "step": 7579
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.642081543611526,
+      "learning_rate": 5.511343077853684e-07,
+      "loss": 0.1615,
+      "step": 7580
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7327974006760869,
+      "learning_rate": 5.510315742589042e-07,
+      "loss": 0.3597,
+      "step": 7581
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5961600903985111,
+      "learning_rate": 5.509288385553302e-07,
+      "loss": 0.1655,
+      "step": 7582
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5370388498415422,
+      "learning_rate": 5.508261006790298e-07,
+      "loss": 0.3869,
+      "step": 7583
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8474899756471022,
+      "learning_rate": 5.507233606343857e-07,
+      "loss": 0.3243,
+      "step": 7584
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2173954578916704,
+      "learning_rate": 5.506206184257809e-07,
+      "loss": 0.3129,
+      "step": 7585
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5548149607465687,
+      "learning_rate": 5.505178740575989e-07,
+      "loss": 0.2358,
+      "step": 7586
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2944959997760943,
+      "learning_rate": 5.504151275342228e-07,
+      "loss": 0.1349,
+      "step": 7587
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 24.841566155442354,
+      "learning_rate": 5.503123788600361e-07,
+      "loss": 0.0623,
+      "step": 7588
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8311751932502293,
+      "learning_rate": 5.502096280394222e-07,
+      "loss": 0.1832,
+      "step": 7589
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7656308631754906,
+      "learning_rate": 5.501068750767646e-07,
+      "loss": 0.2056,
+      "step": 7590
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8605056054119362,
+      "learning_rate": 5.500041199764469e-07,
+      "loss": 0.1156,
+      "step": 7591
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5339431055225383,
+      "learning_rate": 5.49901362742853e-07,
+      "loss": 0.0966,
+      "step": 7592
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7953356023979216,
+      "learning_rate": 5.497986033803664e-07,
+      "loss": 0.2662,
+      "step": 7593
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5622668629224475,
+      "learning_rate": 5.496958418933715e-07,
+      "loss": 0.0963,
+      "step": 7594
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6164005777732087,
+      "learning_rate": 5.495930782862521e-07,
+      "loss": 0.3443,
+      "step": 7595
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9408372696437483,
+      "learning_rate": 5.494903125633923e-07,
+      "loss": 0.2415,
+      "step": 7596
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6880557385671987,
+      "learning_rate": 5.493875447291762e-07,
+      "loss": 0.1342,
+      "step": 7597
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.44870256035805056,
+      "learning_rate": 5.492847747879882e-07,
+      "loss": 0.1061,
+      "step": 7598
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8199473668297109,
+      "learning_rate": 5.491820027442126e-07,
+      "loss": 0.3366,
+      "step": 7599
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7325136084105575,
+      "learning_rate": 5.490792286022339e-07,
+      "loss": 0.2507,
+      "step": 7600
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.728868102794531,
+      "learning_rate": 5.489764523664366e-07,
+      "loss": 0.2105,
+      "step": 7601
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3552811345763844,
+      "learning_rate": 5.488736740412056e-07,
+      "loss": 0.3901,
+      "step": 7602
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9693907483553831,
+      "learning_rate": 5.487708936309252e-07,
+      "loss": 0.2329,
+      "step": 7603
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6543900507727194,
+      "learning_rate": 5.486681111399804e-07,
+      "loss": 0.3631,
+      "step": 7604
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3797306778892442,
+      "learning_rate": 5.485653265727563e-07,
+      "loss": 0.0862,
+      "step": 7605
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.347410830998166,
+      "learning_rate": 5.484625399336378e-07,
+      "loss": 0.1362,
+      "step": 7606
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1322417870599837,
+      "learning_rate": 5.483597512270097e-07,
+      "loss": 0.3191,
+      "step": 7607
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3830461419814357,
+      "learning_rate": 5.482569604572576e-07,
+      "loss": 0.1132,
+      "step": 7608
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4995163021122291,
+      "learning_rate": 5.481541676287664e-07,
+      "loss": 0.3372,
+      "step": 7609
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6449794252728354,
+      "learning_rate": 5.480513727459218e-07,
+      "loss": 0.2552,
+      "step": 7610
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3200613387108409,
+      "learning_rate": 5.479485758131089e-07,
+      "loss": 0.1286,
+      "step": 7611
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4558450421203875,
+      "learning_rate": 5.478457768347132e-07,
+      "loss": 0.1366,
+      "step": 7612
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.501176775713478,
+      "learning_rate": 5.477429758151208e-07,
+      "loss": 0.2014,
+      "step": 7613
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.168999490477346,
+      "learning_rate": 5.476401727587168e-07,
+      "loss": 0.2244,
+      "step": 7614
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5360430642494545,
+      "learning_rate": 5.475373676698874e-07,
+      "loss": 0.2527,
+      "step": 7615
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3431825819123024,
+      "learning_rate": 5.474345605530185e-07,
+      "loss": 0.2389,
+      "step": 7616
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.755240154250283,
+      "learning_rate": 5.473317514124957e-07,
+      "loss": 0.2358,
+      "step": 7617
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5671909924975638,
+      "learning_rate": 5.472289402527053e-07,
+      "loss": 0.2044,
+      "step": 7618
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0461198665266376,
+      "learning_rate": 5.471261270780333e-07,
+      "loss": 0.3341,
+      "step": 7619
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7318666519389283,
+      "learning_rate": 5.470233118928659e-07,
+      "loss": 0.098,
+      "step": 7620
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.196377638093798,
+      "learning_rate": 5.469204947015897e-07,
+      "loss": 0.1907,
+      "step": 7621
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.052346199280715,
+      "learning_rate": 5.468176755085907e-07,
+      "loss": 0.3102,
+      "step": 7622
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4156714251904758,
+      "learning_rate": 5.467148543182556e-07,
+      "loss": 0.0306,
+      "step": 7623
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2150847494755244,
+      "learning_rate": 5.466120311349709e-07,
+      "loss": 0.0505,
+      "step": 7624
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09483021303767877,
+      "learning_rate": 5.465092059631234e-07,
+      "loss": 0.0652,
+      "step": 7625
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5774180036422483,
+      "learning_rate": 5.464063788070995e-07,
+      "loss": 0.0294,
+      "step": 7626
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1717246689561696,
+      "learning_rate": 5.463035496712862e-07,
+      "loss": 0.2189,
+      "step": 7627
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.15920007005961984,
+      "learning_rate": 5.462007185600705e-07,
+      "loss": 0.0866,
+      "step": 7628
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.562923039175392,
+      "learning_rate": 5.460978854778392e-07,
+      "loss": 0.1522,
+      "step": 7629
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.784044452622361,
+      "learning_rate": 5.459950504289794e-07,
+      "loss": 0.1205,
+      "step": 7630
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5308654101621021,
+      "learning_rate": 5.458922134178784e-07,
+      "loss": 0.0208,
+      "step": 7631
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.900879184419944,
+      "learning_rate": 5.457893744489233e-07,
+      "loss": 0.1425,
+      "step": 7632
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.22269806232356187,
+      "learning_rate": 5.456865335265013e-07,
+      "loss": 0.007,
+      "step": 7633
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9320836456837528,
+      "learning_rate": 5.45583690655e-07,
+      "loss": 0.2754,
+      "step": 7634
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8461662886483006,
+      "learning_rate": 5.454808458388069e-07,
+      "loss": 0.1706,
+      "step": 7635
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7396967270574977,
+      "learning_rate": 5.453779990823094e-07,
+      "loss": 0.1853,
+      "step": 7636
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1345758146153235,
+      "learning_rate": 5.452751503898953e-07,
+      "loss": 0.2223,
+      "step": 7637
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.09067743686668761,
+      "learning_rate": 5.451722997659522e-07,
+      "loss": 0.0006,
+      "step": 7638
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4105353757140416,
+      "learning_rate": 5.450694472148679e-07,
+      "loss": 0.3065,
+      "step": 7639
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2832544663749155,
+      "learning_rate": 5.449665927410305e-07,
+      "loss": 0.0066,
+      "step": 7640
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6266049999684713,
+      "learning_rate": 5.44863736348828e-07,
+      "loss": 0.1949,
+      "step": 7641
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7392414920433288,
+      "learning_rate": 5.447608780426481e-07,
+      "loss": 0.0945,
+      "step": 7642
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4139969306125162,
+      "learning_rate": 5.446580178268794e-07,
+      "loss": 0.1347,
+      "step": 7643
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.381408684843561,
+      "learning_rate": 5.445551557059097e-07,
+      "loss": 0.1924,
+      "step": 7644
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5822479549167303,
+      "learning_rate": 5.444522916841275e-07,
+      "loss": 0.4079,
+      "step": 7645
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.772179732324928,
+      "learning_rate": 5.443494257659211e-07,
+      "loss": 0.1675,
+      "step": 7646
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6034245235777258,
+      "learning_rate": 5.442465579556792e-07,
+      "loss": 0.1451,
+      "step": 7647
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0490176186344715,
+      "learning_rate": 5.441436882577901e-07,
+      "loss": 0.4069,
+      "step": 7648
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.994696639605889,
+      "learning_rate": 5.440408166766426e-07,
+      "loss": 0.2387,
+      "step": 7649
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6424124901515734,
+      "learning_rate": 5.439379432166254e-07,
+      "loss": 0.3026,
+      "step": 7650
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.852007642684793,
+      "learning_rate": 5.438350678821269e-07,
+      "loss": 0.2627,
+      "step": 7651
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.395700268164772,
+      "learning_rate": 5.437321906775366e-07,
+      "loss": 0.2675,
+      "step": 7652
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.27957112006074175,
+      "learning_rate": 5.43629311607243e-07,
+      "loss": 0.1985,
+      "step": 7653
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7109060020565691,
+      "learning_rate": 5.435264306756353e-07,
+      "loss": 0.2381,
+      "step": 7654
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7992833436780332,
+      "learning_rate": 5.434235478871025e-07,
+      "loss": 0.267,
+      "step": 7655
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8420379115632857,
+      "learning_rate": 5.433206632460339e-07,
+      "loss": 0.1712,
+      "step": 7656
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0141722606768333,
+      "learning_rate": 5.432177767568188e-07,
+      "loss": 0.2647,
+      "step": 7657
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9163221936286211,
+      "learning_rate": 5.431148884238463e-07,
+      "loss": 0.2378,
+      "step": 7658
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.11536793358381,
+      "learning_rate": 5.430119982515061e-07,
+      "loss": 0.3102,
+      "step": 7659
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6877843730171554,
+      "learning_rate": 5.429091062441877e-07,
+      "loss": 0.1391,
+      "step": 7660
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8357158283203828,
+      "learning_rate": 5.428062124062803e-07,
+      "loss": 0.1612,
+      "step": 7661
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3087892038315234,
+      "learning_rate": 5.427033167421739e-07,
+      "loss": 0.1355,
+      "step": 7662
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6497803644720554,
+      "learning_rate": 5.426004192562583e-07,
+      "loss": 0.1091,
+      "step": 7663
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1518661309368448,
+      "learning_rate": 5.424975199529231e-07,
+      "loss": 0.0588,
+      "step": 7664
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7817363220829404,
+      "learning_rate": 5.423946188365583e-07,
+      "loss": 0.2561,
+      "step": 7665
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.0568310816484825,
+      "learning_rate": 5.422917159115538e-07,
+      "loss": 0.0106,
+      "step": 7666
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0509192402800769,
+      "learning_rate": 5.421888111822996e-07,
+      "loss": 0.1834,
+      "step": 7667
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.638077809428933,
+      "learning_rate": 5.42085904653186e-07,
+      "loss": 0.0882,
+      "step": 7668
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1955041141489864,
+      "learning_rate": 5.419829963286032e-07,
+      "loss": 0.0549,
+      "step": 7669
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5011330821533714,
+      "learning_rate": 5.418800862129411e-07,
+      "loss": 0.129,
+      "step": 7670
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6610469479758729,
+      "learning_rate": 5.417771743105907e-07,
+      "loss": 0.2157,
+      "step": 7671
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.32319978083462303,
+      "learning_rate": 5.416742606259418e-07,
+      "loss": 0.0035,
+      "step": 7672
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.25474548508449196,
+      "learning_rate": 5.415713451633852e-07,
+      "loss": 0.1364,
+      "step": 7673
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5000377958750638,
+      "learning_rate": 5.414684279273115e-07,
+      "loss": 0.1953,
+      "step": 7674
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9522464291738733,
+      "learning_rate": 5.413655089221113e-07,
+      "loss": 0.3496,
+      "step": 7675
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3830960507033039,
+      "learning_rate": 5.412625881521753e-07,
+      "loss": 0.1597,
+      "step": 7676
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.8450301876350155,
+      "learning_rate": 5.411596656218945e-07,
+      "loss": 0.1851,
+      "step": 7677
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6179371119718056,
+      "learning_rate": 5.410567413356593e-07,
+      "loss": 0.1354,
+      "step": 7678
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8340167567328702,
+      "learning_rate": 5.409538152978612e-07,
+      "loss": 0.2141,
+      "step": 7679
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8976651429360185,
+      "learning_rate": 5.40850887512891e-07,
+      "loss": 0.2676,
+      "step": 7680
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1350004580546227,
+      "learning_rate": 5.407479579851398e-07,
+      "loss": 0.3251,
+      "step": 7681
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8125019697000947,
+      "learning_rate": 5.406450267189989e-07,
+      "loss": 0.1211,
+      "step": 7682
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.87036871741752,
+      "learning_rate": 5.405420937188591e-07,
+      "loss": 0.0286,
+      "step": 7683
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2302845580684012,
+      "learning_rate": 5.404391589891125e-07,
+      "loss": 0.3608,
+      "step": 7684
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.849230930705265,
+      "learning_rate": 5.403362225341499e-07,
+      "loss": 0.205,
+      "step": 7685
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8302906844553808,
+      "learning_rate": 5.40233284358363e-07,
+      "loss": 0.221,
+      "step": 7686
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.663861848975396,
+      "learning_rate": 5.401303444661433e-07,
+      "loss": 0.2397,
+      "step": 7687
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9918275103532312,
+      "learning_rate": 5.400274028618824e-07,
+      "loss": 0.1558,
+      "step": 7688
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7139400232361313,
+      "learning_rate": 5.39924459549972e-07,
+      "loss": 0.3042,
+      "step": 7689
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6811553655827545,
+      "learning_rate": 5.398215145348039e-07,
+      "loss": 0.2626,
+      "step": 7690
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6472167918658924,
+      "learning_rate": 5.3971856782077e-07,
+      "loss": 0.2076,
+      "step": 7691
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7166800549540868,
+      "learning_rate": 5.396156194122621e-07,
+      "loss": 0.2499,
+      "step": 7692
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7760320156479528,
+      "learning_rate": 5.395126693136723e-07,
+      "loss": 0.4403,
+      "step": 7693
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.44126949815312816,
+      "learning_rate": 5.394097175293926e-07,
+      "loss": 0.2611,
+      "step": 7694
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2981542533191037,
+      "learning_rate": 5.393067640638151e-07,
+      "loss": 0.1411,
+      "step": 7695
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3773437959150394,
+      "learning_rate": 5.39203808921332e-07,
+      "loss": 0.2844,
+      "step": 7696
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9578332391244648,
+      "learning_rate": 5.391008521063356e-07,
+      "loss": 0.3256,
+      "step": 7697
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6184362630267441,
+      "learning_rate": 5.389978936232185e-07,
+      "loss": 0.0678,
+      "step": 7698
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.015152660420145,
+      "learning_rate": 5.388949334763724e-07,
+      "loss": 0.3061,
+      "step": 7699
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.14821951159348457,
+      "learning_rate": 5.387919716701905e-07,
+      "loss": 0.077,
+      "step": 7700
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8196056775202845,
+      "learning_rate": 5.386890082090652e-07,
+      "loss": 0.0755,
+      "step": 7701
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.27443719106242326,
+      "learning_rate": 5.385860430973889e-07,
+      "loss": 0.168,
+      "step": 7702
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.39796304383492087,
+      "learning_rate": 5.384830763395544e-07,
+      "loss": 0.2335,
+      "step": 7703
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2585751688268829,
+      "learning_rate": 5.383801079399546e-07,
+      "loss": 0.1074,
+      "step": 7704
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2769851886238829,
+      "learning_rate": 5.382771379029822e-07,
+      "loss": 0.1108,
+      "step": 7705
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.638441502383699,
+      "learning_rate": 5.381741662330302e-07,
+      "loss": 0.1724,
+      "step": 7706
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.681914166537708,
+      "learning_rate": 5.380711929344914e-07,
+      "loss": 0.2414,
+      "step": 7707
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.42036962401488187,
+      "learning_rate": 5.37968218011759e-07,
+      "loss": 0.1828,
+      "step": 7708
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6681332202550524,
+      "learning_rate": 5.378652414692262e-07,
+      "loss": 0.1142,
+      "step": 7709
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6668070210396156,
+      "learning_rate": 5.37762263311286e-07,
+      "loss": 0.2297,
+      "step": 7710
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17273657349610713,
+      "learning_rate": 5.376592835423319e-07,
+      "loss": 0.0043,
+      "step": 7711
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6394614784215694,
+      "learning_rate": 5.37556302166757e-07,
+      "loss": 0.2194,
+      "step": 7712
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9990973771703578,
+      "learning_rate": 5.374533191889546e-07,
+      "loss": 0.3003,
+      "step": 7713
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.852651630229182,
+      "learning_rate": 5.373503346133183e-07,
+      "loss": 0.0856,
+      "step": 7714
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7313290835938698,
+      "learning_rate": 5.372473484442418e-07,
+      "loss": 0.0975,
+      "step": 7715
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7619842201842432,
+      "learning_rate": 5.371443606861186e-07,
+      "loss": 0.1442,
+      "step": 7716
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.48601837270091935,
+      "learning_rate": 5.370413713433419e-07,
+      "loss": 0.2027,
+      "step": 7717
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9943991060640079,
+      "learning_rate": 5.369383804203062e-07,
+      "loss": 0.1661,
+      "step": 7718
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7764327557555127,
+      "learning_rate": 5.368353879214048e-07,
+      "loss": 0.1986,
+      "step": 7719
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9571235265037613,
+      "learning_rate": 5.367323938510319e-07,
+      "loss": 0.0151,
+      "step": 7720
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1331415179899828,
+      "learning_rate": 5.36629398213581e-07,
+      "loss": 0.0049,
+      "step": 7721
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9564062075025476,
+      "learning_rate": 5.365264010134465e-07,
+      "loss": 0.2393,
+      "step": 7722
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9988487704387783,
+      "learning_rate": 5.364234022550222e-07,
+      "loss": 0.0996,
+      "step": 7723
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6578332170961195,
+      "learning_rate": 5.363204019427023e-07,
+      "loss": 0.4422,
+      "step": 7724
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0935508995397545,
+      "learning_rate": 5.362174000808812e-07,
+      "loss": 0.3207,
+      "step": 7725
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6625098167845284,
+      "learning_rate": 5.361143966739528e-07,
+      "loss": 0.2386,
+      "step": 7726
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6384002449405118,
+      "learning_rate": 5.360113917263119e-07,
+      "loss": 0.2626,
+      "step": 7727
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.86185255036157,
+      "learning_rate": 5.359083852423525e-07,
+      "loss": 0.5264,
+      "step": 7728
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5290037514723575,
+      "learning_rate": 5.358053772264691e-07,
+      "loss": 0.1684,
+      "step": 7729
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.22971659840906652,
+      "learning_rate": 5.357023676830565e-07,
+      "loss": 0.0068,
+      "step": 7730
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0084694477299414,
+      "learning_rate": 5.35599356616509e-07,
+      "loss": 0.3455,
+      "step": 7731
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5719038292544176,
+      "learning_rate": 5.354963440312215e-07,
+      "loss": 0.1823,
+      "step": 7732
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8107698385400323,
+      "learning_rate": 5.353933299315885e-07,
+      "loss": 0.0623,
+      "step": 7733
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5501289802182807,
+      "learning_rate": 5.35290314322005e-07,
+      "loss": 0.3329,
+      "step": 7734
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3193147033692225,
+      "learning_rate": 5.351872972068656e-07,
+      "loss": 0.214,
+      "step": 7735
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8932772123066981,
+      "learning_rate": 5.350842785905654e-07,
+      "loss": 0.5223,
+      "step": 7736
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3815366076253965,
+      "learning_rate": 5.349812584774994e-07,
+      "loss": 0.1413,
+      "step": 7737
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0877090864882917,
+      "learning_rate": 5.348782368720625e-07,
+      "loss": 0.3399,
+      "step": 7738
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.649411908566595,
+      "learning_rate": 5.347752137786501e-07,
+      "loss": 0.0373,
+      "step": 7739
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.168303644676484,
+      "learning_rate": 5.346721892016571e-07,
+      "loss": 0.2095,
+      "step": 7740
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6210928324996892,
+      "learning_rate": 5.345691631454788e-07,
+      "loss": 0.1635,
+      "step": 7741
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4875449275210424,
+      "learning_rate": 5.344661356145104e-07,
+      "loss": 0.1373,
+      "step": 7742
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.35111158500776496,
+      "learning_rate": 5.343631066131476e-07,
+      "loss": 0.2429,
+      "step": 7743
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.130496377326105,
+      "learning_rate": 5.342600761457853e-07,
+      "loss": 0.286,
+      "step": 7744
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6032419223302801,
+      "learning_rate": 5.341570442168194e-07,
+      "loss": 0.0359,
+      "step": 7745
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0336253550312409,
+      "learning_rate": 5.340540108306454e-07,
+      "loss": 0.1477,
+      "step": 7746
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4896560824713781,
+      "learning_rate": 5.339509759916589e-07,
+      "loss": 0.1968,
+      "step": 7747
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.624759526829821,
+      "learning_rate": 5.338479397042553e-07,
+      "loss": 0.3045,
+      "step": 7748
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2795796516751618,
+      "learning_rate": 5.337449019728306e-07,
+      "loss": 0.1251,
+      "step": 7749
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8266157280587519,
+      "learning_rate": 5.336418628017807e-07,
+      "loss": 0.3139,
+      "step": 7750
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6785908836524304,
+      "learning_rate": 5.335388221955012e-07,
+      "loss": 0.2336,
+      "step": 7751
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3237713765769747,
+      "learning_rate": 5.334357801583881e-07,
+      "loss": 0.2462,
+      "step": 7752
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5192653336984875,
+      "learning_rate": 5.333327366948374e-07,
+      "loss": 0.1865,
+      "step": 7753
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.24939429534706634,
+      "learning_rate": 5.332296918092453e-07,
+      "loss": 0.2587,
+      "step": 7754
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.38511437198997533,
+      "learning_rate": 5.331266455060077e-07,
+      "loss": 0.011,
+      "step": 7755
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2836501620836957,
+      "learning_rate": 5.330235977895205e-07,
+      "loss": 0.1769,
+      "step": 7756
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.995630535414285,
+      "learning_rate": 5.329205486641806e-07,
+      "loss": 0.385,
+      "step": 7757
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7148093968339438,
+      "learning_rate": 5.328174981343838e-07,
+      "loss": 0.3027,
+      "step": 7758
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8158328911905972,
+      "learning_rate": 5.327144462045266e-07,
+      "loss": 0.3766,
+      "step": 7759
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.58806648258431,
+      "learning_rate": 5.326113928790053e-07,
+      "loss": 0.2653,
+      "step": 7760
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9748352938096041,
+      "learning_rate": 5.325083381622164e-07,
+      "loss": 0.178,
+      "step": 7761
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9685051298089195,
+      "learning_rate": 5.324052820585563e-07,
+      "loss": 0.2109,
+      "step": 7762
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.344605254927458,
+      "learning_rate": 5.323022245724219e-07,
+      "loss": 0.1343,
+      "step": 7763
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.313227171798812,
+      "learning_rate": 5.321991657082096e-07,
+      "loss": 0.0295,
+      "step": 7764
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5575780887903956,
+      "learning_rate": 5.320961054703163e-07,
+      "loss": 0.305,
+      "step": 7765
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.36333360036625995,
+      "learning_rate": 5.319930438631386e-07,
+      "loss": 0.002,
+      "step": 7766
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4898434219005641,
+      "learning_rate": 5.318899808910733e-07,
+      "loss": 0.1105,
+      "step": 7767
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.24221747318964093,
+      "learning_rate": 5.317869165585173e-07,
+      "loss": 0.0364,
+      "step": 7768
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3506909444437065,
+      "learning_rate": 5.316838508698676e-07,
+      "loss": 0.2376,
+      "step": 7769
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7355945954046756,
+      "learning_rate": 5.315807838295208e-07,
+      "loss": 0.2195,
+      "step": 7770
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6566585438574053,
+      "learning_rate": 5.314777154418746e-07,
+      "loss": 0.1648,
+      "step": 7771
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.41189169152852995,
+      "learning_rate": 5.313746457113257e-07,
+      "loss": 0.1249,
+      "step": 7772
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6918205197923989,
+      "learning_rate": 5.312715746422713e-07,
+      "loss": 0.3339,
+      "step": 7773
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2473419047145127,
+      "learning_rate": 5.311685022391088e-07,
+      "loss": 0.1643,
+      "step": 7774
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5809377724992406,
+      "learning_rate": 5.310654285062352e-07,
+      "loss": 0.2165,
+      "step": 7775
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.015666218305165,
+      "learning_rate": 5.30962353448048e-07,
+      "loss": 0.1307,
+      "step": 7776
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.38949374229009076,
+      "learning_rate": 5.308592770689447e-07,
+      "loss": 0.1978,
+      "step": 7777
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.157184007758185,
+      "learning_rate": 5.307561993733225e-07,
+      "loss": 0.2324,
+      "step": 7778
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.22596655073587732,
+      "learning_rate": 5.306531203655789e-07,
+      "loss": 0.0781,
+      "step": 7779
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7766956549889035,
+      "learning_rate": 5.305500400501116e-07,
+      "loss": 0.2524,
+      "step": 7780
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.533611785353364,
+      "learning_rate": 5.304469584313184e-07,
+      "loss": 0.1424,
+      "step": 7781
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6913264970543666,
+      "learning_rate": 5.303438755135966e-07,
+      "loss": 0.4107,
+      "step": 7782
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9672104309768748,
+      "learning_rate": 5.30240791301344e-07,
+      "loss": 0.1589,
+      "step": 7783
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6988466547244304,
+      "learning_rate": 5.301377057989585e-07,
+      "loss": 0.1725,
+      "step": 7784
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2910088842559605,
+      "learning_rate": 5.30034619010838e-07,
+      "loss": 0.1654,
+      "step": 7785
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.355324140429315,
+      "learning_rate": 5.299315309413801e-07,
+      "loss": 0.3068,
+      "step": 7786
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8952307282853894,
+      "learning_rate": 5.298284415949831e-07,
+      "loss": 0.2814,
+      "step": 7787
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9455958195092917,
+      "learning_rate": 5.297253509760447e-07,
+      "loss": 0.1514,
+      "step": 7788
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0228700995109932,
+      "learning_rate": 5.29622259088963e-07,
+      "loss": 0.0032,
+      "step": 7789
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.914851414661287,
+      "learning_rate": 5.295191659381361e-07,
+      "loss": 0.1251,
+      "step": 7790
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1990458833847757,
+      "learning_rate": 5.294160715279625e-07,
+      "loss": 0.3261,
+      "step": 7791
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.189237769919022,
+      "learning_rate": 5.293129758628401e-07,
+      "loss": 0.1638,
+      "step": 7792
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4307245719668564,
+      "learning_rate": 5.292098789471672e-07,
+      "loss": 0.1495,
+      "step": 7793
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9181822888819925,
+      "learning_rate": 5.291067807853421e-07,
+      "loss": 0.0756,
+      "step": 7794
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0261945761045828,
+      "learning_rate": 5.290036813817633e-07,
+      "loss": 0.1123,
+      "step": 7795
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8237465208574568,
+      "learning_rate": 5.289005807408291e-07,
+      "loss": 0.1983,
+      "step": 7796
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9719494841336705,
+      "learning_rate": 5.287974788669379e-07,
+      "loss": 0.5802,
+      "step": 7797
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7977473228851141,
+      "learning_rate": 5.286943757644885e-07,
+      "loss": 0.3637,
+      "step": 7798
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5200201574343315,
+      "learning_rate": 5.285912714378794e-07,
+      "loss": 0.1093,
+      "step": 7799
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.46932306713004374,
+      "learning_rate": 5.28488165891509e-07,
+      "loss": 0.0479,
+      "step": 7800
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.44605212307224,
+      "learning_rate": 5.283850591297764e-07,
+      "loss": 0.2126,
+      "step": 7801
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.7413993988323,
+      "learning_rate": 5.2828195115708e-07,
+      "loss": 0.3058,
+      "step": 7802
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4379561362648015,
+      "learning_rate": 5.281788419778187e-07,
+      "loss": 0.0603,
+      "step": 7803
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8758741891476285,
+      "learning_rate": 5.280757315963914e-07,
+      "loss": 0.1908,
+      "step": 7804
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4412617542815155,
+      "learning_rate": 5.279726200171968e-07,
+      "loss": 0.2986,
+      "step": 7805
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5032658183060224,
+      "learning_rate": 5.278695072446342e-07,
+      "loss": 0.0798,
+      "step": 7806
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4376415470066588,
+      "learning_rate": 5.277663932831022e-07,
+      "loss": 0.0424,
+      "step": 7807
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9769449621105338,
+      "learning_rate": 5.276632781370003e-07,
+      "loss": 0.1485,
+      "step": 7808
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7515751597096842,
+      "learning_rate": 5.275601618107272e-07,
+      "loss": 0.431,
+      "step": 7809
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7794421637018186,
+      "learning_rate": 5.274570443086822e-07,
+      "loss": 0.2394,
+      "step": 7810
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5447644869777392,
+      "learning_rate": 5.273539256352645e-07,
+      "loss": 0.168,
+      "step": 7811
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8876176330913443,
+      "learning_rate": 5.272508057948734e-07,
+      "loss": 0.1879,
+      "step": 7812
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9865574689471549,
+      "learning_rate": 5.27147684791908e-07,
+      "loss": 0.2503,
+      "step": 7813
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9871320823600839,
+      "learning_rate": 5.270445626307679e-07,
+      "loss": 0.3219,
+      "step": 7814
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7105275225394994,
+      "learning_rate": 5.269414393158522e-07,
+      "loss": 0.1235,
+      "step": 7815
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1802947614800248,
+      "learning_rate": 5.268383148515607e-07,
+      "loss": 0.0591,
+      "step": 7816
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.3279453807796,
+      "learning_rate": 5.267351892422928e-07,
+      "loss": 0.1419,
+      "step": 7817
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4359935887534894,
+      "learning_rate": 5.266320624924479e-07,
+      "loss": 0.4444,
+      "step": 7818
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.335113781847231,
+      "learning_rate": 5.265289346064258e-07,
+      "loss": 0.2569,
+      "step": 7819
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.056877467538794,
+      "learning_rate": 5.264258055886258e-07,
+      "loss": 0.2682,
+      "step": 7820
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6868373893259,
+      "learning_rate": 5.26322675443448e-07,
+      "loss": 0.1618,
+      "step": 7821
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5173004207683524,
+      "learning_rate": 5.26219544175292e-07,
+      "loss": 0.2116,
+      "step": 7822
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5744844435517102,
+      "learning_rate": 5.261164117885572e-07,
+      "loss": 0.2363,
+      "step": 7823
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0058124276324367,
+      "learning_rate": 5.26013278287644e-07,
+      "loss": 0.2604,
+      "step": 7824
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5659951737869565,
+      "learning_rate": 5.259101436769522e-07,
+      "loss": 0.1254,
+      "step": 7825
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8246655049506346,
+      "learning_rate": 5.258070079608814e-07,
+      "loss": 0.0466,
+      "step": 7826
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.354953605724272,
+      "learning_rate": 5.257038711438318e-07,
+      "loss": 0.1545,
+      "step": 7827
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.035775037723114,
+      "learning_rate": 5.256007332302033e-07,
+      "loss": 0.2671,
+      "step": 7828
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.089958319818489,
+      "learning_rate": 5.254975942243962e-07,
+      "loss": 0.3936,
+      "step": 7829
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.762537594903154,
+      "learning_rate": 5.253944541308105e-07,
+      "loss": 0.1401,
+      "step": 7830
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.746543211746795,
+      "learning_rate": 5.252913129538462e-07,
+      "loss": 0.2737,
+      "step": 7831
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.28309018431784894,
+      "learning_rate": 5.251881706979036e-07,
+      "loss": 0.12,
+      "step": 7832
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8489990124044013,
+      "learning_rate": 5.25085027367383e-07,
+      "loss": 0.2752,
+      "step": 7833
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3177622837747427,
+      "learning_rate": 5.249818829666849e-07,
+      "loss": 0.1412,
+      "step": 7834
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8600800060063669,
+      "learning_rate": 5.248787375002093e-07,
+      "loss": 0.3269,
+      "step": 7835
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9871049459997794,
+      "learning_rate": 5.247755909723569e-07,
+      "loss": 0.0822,
+      "step": 7836
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.824574174471444,
+      "learning_rate": 5.24672443387528e-07,
+      "loss": 0.2035,
+      "step": 7837
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7726990955671356,
+      "learning_rate": 5.245692947501229e-07,
+      "loss": 0.0721,
+      "step": 7838
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7629832334784106,
+      "learning_rate": 5.244661450645424e-07,
+      "loss": 0.0428,
+      "step": 7839
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.025522337254423,
+      "learning_rate": 5.243629943351868e-07,
+      "loss": 0.4119,
+      "step": 7840
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0978245388548087,
+      "learning_rate": 5.242598425664569e-07,
+      "loss": 0.3479,
+      "step": 7841
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4144828682749771,
+      "learning_rate": 5.241566897627535e-07,
+      "loss": 0.35,
+      "step": 7842
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.348071760526536,
+      "learning_rate": 5.240535359284771e-07,
+      "loss": 0.0055,
+      "step": 7843
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.121454223171519,
+      "learning_rate": 5.239503810680285e-07,
+      "loss": 0.1295,
+      "step": 7844
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.88673366259795,
+      "learning_rate": 5.238472251858085e-07,
+      "loss": 0.02,
+      "step": 7845
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.562742651472133,
+      "learning_rate": 5.23744068286218e-07,
+      "loss": 0.2023,
+      "step": 7846
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.845274488769913,
+      "learning_rate": 5.236409103736578e-07,
+      "loss": 0.0736,
+      "step": 7847
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6935526470572801,
+      "learning_rate": 5.235377514525287e-07,
+      "loss": 0.2186,
+      "step": 7848
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.491441170439203,
+      "learning_rate": 5.234345915272319e-07,
+      "loss": 0.1227,
+      "step": 7849
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0239583355372224,
+      "learning_rate": 5.233314306021683e-07,
+      "loss": 0.1545,
+      "step": 7850
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8937449633580066,
+      "learning_rate": 5.232282686817391e-07,
+      "loss": 0.2691,
+      "step": 7851
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.48963595647131725,
+      "learning_rate": 5.23125105770345e-07,
+      "loss": 0.0804,
+      "step": 7852
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.800896218318243,
+      "learning_rate": 5.230219418723877e-07,
+      "loss": 0.1848,
+      "step": 7853
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6642178462979964,
+      "learning_rate": 5.229187769922678e-07,
+      "loss": 0.3277,
+      "step": 7854
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6979156812551884,
+      "learning_rate": 5.228156111343869e-07,
+      "loss": 0.0602,
+      "step": 7855
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7067320927297459,
+      "learning_rate": 5.227124443031463e-07,
+      "loss": 0.1629,
+      "step": 7856
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7787555692149855,
+      "learning_rate": 5.226092765029471e-07,
+      "loss": 0.1465,
+      "step": 7857
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9791138546302344,
+      "learning_rate": 5.225061077381906e-07,
+      "loss": 0.1425,
+      "step": 7858
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.396157475943847,
+      "learning_rate": 5.224029380132784e-07,
+      "loss": 0.0152,
+      "step": 7859
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5747056921221125,
+      "learning_rate": 5.222997673326117e-07,
+      "loss": 0.0786,
+      "step": 7860
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4217302165875072,
+      "learning_rate": 5.221965957005923e-07,
+      "loss": 0.0959,
+      "step": 7861
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.778564964888106,
+      "learning_rate": 5.220934231216213e-07,
+      "loss": 0.1542,
+      "step": 7862
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.471767682061118,
+      "learning_rate": 5.219902496001007e-07,
+      "loss": 0.4055,
+      "step": 7863
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5294869636570287,
+      "learning_rate": 5.218870751404318e-07,
+      "loss": 0.0591,
+      "step": 7864
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7789689219152222,
+      "learning_rate": 5.217838997470161e-07,
+      "loss": 0.2538,
+      "step": 7865
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.266033380340139,
+      "learning_rate": 5.216807234242556e-07,
+      "loss": 0.1506,
+      "step": 7866
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6533613723205102,
+      "learning_rate": 5.215775461765518e-07,
+      "loss": 0.1102,
+      "step": 7867
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7032095492239564,
+      "learning_rate": 5.214743680083063e-07,
+      "loss": 0.325,
+      "step": 7868
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2152376585789004,
+      "learning_rate": 5.213711889239213e-07,
+      "loss": 0.2158,
+      "step": 7869
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.031521999729483,
+      "learning_rate": 5.212680089277985e-07,
+      "loss": 0.023,
+      "step": 7870
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4208887784432154,
+      "learning_rate": 5.211648280243395e-07,
+      "loss": 0.0118,
+      "step": 7871
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.47270387623683613,
+      "learning_rate": 5.210616462179464e-07,
+      "loss": 0.2021,
+      "step": 7872
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.21269858601264,
+      "learning_rate": 5.209584635130213e-07,
+      "loss": 0.2684,
+      "step": 7873
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7791515887512775,
+      "learning_rate": 5.20855279913966e-07,
+      "loss": 0.2264,
+      "step": 7874
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.987061945698786,
+      "learning_rate": 5.207520954251824e-07,
+      "loss": 0.2461,
+      "step": 7875
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1747753057153916,
+      "learning_rate": 5.206489100510728e-07,
+      "loss": 0.2769,
+      "step": 7876
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6665906890843033,
+      "learning_rate": 5.205457237960391e-07,
+      "loss": 0.2738,
+      "step": 7877
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.369531591363996,
+      "learning_rate": 5.204425366644835e-07,
+      "loss": 0.1376,
+      "step": 7878
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8279046066898146,
+      "learning_rate": 5.203393486608083e-07,
+      "loss": 0.1184,
+      "step": 7879
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3569836166745123,
+      "learning_rate": 5.202361597894156e-07,
+      "loss": 0.0574,
+      "step": 7880
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7107200054599689,
+      "learning_rate": 5.201329700547076e-07,
+      "loss": 0.2351,
+      "step": 7881
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.19613903012530348,
+      "learning_rate": 5.200297794610866e-07,
+      "loss": 0.0029,
+      "step": 7882
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3353815282918122,
+      "learning_rate": 5.199265880129549e-07,
+      "loss": 0.2413,
+      "step": 7883
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.24590316881461446,
+      "learning_rate": 5.19823395714715e-07,
+      "loss": 0.0928,
+      "step": 7884
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6912026268662488,
+      "learning_rate": 5.197202025707692e-07,
+      "loss": 0.0113,
+      "step": 7885
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3384877353144549,
+      "learning_rate": 5.196170085855197e-07,
+      "loss": 0.2684,
+      "step": 7886
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6400111663334485,
+      "learning_rate": 5.195138137633695e-07,
+      "loss": 0.2284,
+      "step": 7887
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7398213844163112,
+      "learning_rate": 5.194106181087205e-07,
+      "loss": 0.2143,
+      "step": 7888
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2830114388100147,
+      "learning_rate": 5.193074216259756e-07,
+      "loss": 0.2561,
+      "step": 7889
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9472489104447508,
+      "learning_rate": 5.192042243195374e-07,
+      "loss": 0.2109,
+      "step": 7890
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.367648599858341,
+      "learning_rate": 5.191010261938084e-07,
+      "loss": 0.4713,
+      "step": 7891
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.155433735597053,
+      "learning_rate": 5.18997827253191e-07,
+      "loss": 0.1569,
+      "step": 7892
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.406164477273476,
+      "learning_rate": 5.188946275020883e-07,
+      "loss": 0.0912,
+      "step": 7893
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9541769863246171,
+      "learning_rate": 5.187914269449027e-07,
+      "loss": 0.1282,
+      "step": 7894
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.12016431612114,
+      "learning_rate": 5.186882255860371e-07,
+      "loss": 0.2475,
+      "step": 7895
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.077277458584877,
+      "learning_rate": 5.185850234298942e-07,
+      "loss": 0.0833,
+      "step": 7896
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1287624437146637,
+      "learning_rate": 5.184818204808768e-07,
+      "loss": 0.4477,
+      "step": 7897
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.067688187277838,
+      "learning_rate": 5.183786167433879e-07,
+      "loss": 0.1426,
+      "step": 7898
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7476661301573576,
+      "learning_rate": 5.182754122218301e-07,
+      "loss": 0.0813,
+      "step": 7899
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2548690992889109,
+      "learning_rate": 5.181722069206067e-07,
+      "loss": 0.1678,
+      "step": 7900
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4858376999577905,
+      "learning_rate": 5.180690008441202e-07,
+      "loss": 0.0102,
+      "step": 7901
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0979010552464186,
+      "learning_rate": 5.179657939967739e-07,
+      "loss": 0.2151,
+      "step": 7902
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7467724528418,
+      "learning_rate": 5.178625863829708e-07,
+      "loss": 0.1008,
+      "step": 7903
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0258098759250416,
+      "learning_rate": 5.177593780071138e-07,
+      "loss": 0.216,
+      "step": 7904
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8528197600824754,
+      "learning_rate": 5.176561688736059e-07,
+      "loss": 0.2002,
+      "step": 7905
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4726365861437474,
+      "learning_rate": 5.175529589868505e-07,
+      "loss": 0.1345,
+      "step": 7906
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.778146508546821,
+      "learning_rate": 5.174497483512505e-07,
+      "loss": 0.1732,
+      "step": 7907
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.175122001822574,
+      "learning_rate": 5.173465369712092e-07,
+      "loss": 0.0348,
+      "step": 7908
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3902211936326443,
+      "learning_rate": 5.172433248511298e-07,
+      "loss": 0.2418,
+      "step": 7909
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6413738610015453,
+      "learning_rate": 5.171401119954155e-07,
+      "loss": 0.2419,
+      "step": 7910
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9434512738195908,
+      "learning_rate": 5.170368984084695e-07,
+      "loss": 0.0773,
+      "step": 7911
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.49654956072595885,
+      "learning_rate": 5.169336840946951e-07,
+      "loss": 0.2155,
+      "step": 7912
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6266402889062209,
+      "learning_rate": 5.168304690584957e-07,
+      "loss": 0.2152,
+      "step": 7913
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0088775659474432,
+      "learning_rate": 5.167272533042747e-07,
+      "loss": 0.1774,
+      "step": 7914
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.413291250801521,
+      "learning_rate": 5.166240368364355e-07,
+      "loss": 0.3786,
+      "step": 7915
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4757878077723017,
+      "learning_rate": 5.165208196593815e-07,
+      "loss": 0.1337,
+      "step": 7916
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5831083215292089,
+      "learning_rate": 5.16417601777516e-07,
+      "loss": 0.2393,
+      "step": 7917
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2954726822787284,
+      "learning_rate": 5.163143831952428e-07,
+      "loss": 0.015,
+      "step": 7918
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0815712859292015,
+      "learning_rate": 5.16211163916965e-07,
+      "loss": 0.2246,
+      "step": 7919
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8915610110045376,
+      "learning_rate": 5.161079439470865e-07,
+      "loss": 0.3047,
+      "step": 7920
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.616067334512679,
+      "learning_rate": 5.160047232900105e-07,
+      "loss": 0.0679,
+      "step": 7921
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6201283893735887,
+      "learning_rate": 5.159015019501412e-07,
+      "loss": 0.1053,
+      "step": 7922
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5590432125548226,
+      "learning_rate": 5.157982799318816e-07,
+      "loss": 0.0049,
+      "step": 7923
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5662754624081587,
+      "learning_rate": 5.156950572396357e-07,
+      "loss": 0.002,
+      "step": 7924
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3489688627740372,
+      "learning_rate": 5.15591833877807e-07,
+      "loss": 0.1944,
+      "step": 7925
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9277712734648788,
+      "learning_rate": 5.154886098507994e-07,
+      "loss": 0.2823,
+      "step": 7926
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3126595205279847,
+      "learning_rate": 5.153853851630167e-07,
+      "loss": 0.2276,
+      "step": 7927
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5437867303477449,
+      "learning_rate": 5.152821598188624e-07,
+      "loss": 0.2645,
+      "step": 7928
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.298265152104033,
+      "learning_rate": 5.151789338227404e-07,
+      "loss": 0.2065,
+      "step": 7929
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3523082298627538,
+      "learning_rate": 5.150757071790546e-07,
+      "loss": 0.2012,
+      "step": 7930
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6762209296542762,
+      "learning_rate": 5.14972479892209e-07,
+      "loss": 0.1619,
+      "step": 7931
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8012859942811726,
+      "learning_rate": 5.148692519666071e-07,
+      "loss": 0.1933,
+      "step": 7932
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3174116071937148,
+      "learning_rate": 5.147660234066532e-07,
+      "loss": 0.1909,
+      "step": 7933
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.46627943344389033,
+      "learning_rate": 5.146627942167509e-07,
+      "loss": 0.0972,
+      "step": 7934
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.42565532044733345,
+      "learning_rate": 5.145595644013044e-07,
+      "loss": 0.1496,
+      "step": 7935
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.854961639675671,
+      "learning_rate": 5.144563339647177e-07,
+      "loss": 0.1301,
+      "step": 7936
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.828730714431461,
+      "learning_rate": 5.143531029113946e-07,
+      "loss": 0.3431,
+      "step": 7937
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8837225539057707,
+      "learning_rate": 5.142498712457392e-07,
+      "loss": 0.1427,
+      "step": 7938
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3271662302507762,
+      "learning_rate": 5.141466389721557e-07,
+      "loss": 0.1194,
+      "step": 7939
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.29422800625131323,
+      "learning_rate": 5.140434060950482e-07,
+      "loss": 0.1945,
+      "step": 7940
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8082749986629059,
+      "learning_rate": 5.139401726188208e-07,
+      "loss": 0.2746,
+      "step": 7941
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.66322914498958,
+      "learning_rate": 5.138369385478774e-07,
+      "loss": 0.3334,
+      "step": 7942
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5024296011165815,
+      "learning_rate": 5.137337038866227e-07,
+      "loss": 0.0649,
+      "step": 7943
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5067471540265036,
+      "learning_rate": 5.136304686394604e-07,
+      "loss": 0.0759,
+      "step": 7944
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6354339781473719,
+      "learning_rate": 5.135272328107949e-07,
+      "loss": 0.1112,
+      "step": 7945
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.129014749509955,
+      "learning_rate": 5.134239964050307e-07,
+      "loss": 0.1003,
+      "step": 7946
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5008406881251112,
+      "learning_rate": 5.133207594265715e-07,
+      "loss": 0.1838,
+      "step": 7947
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1048411268653404,
+      "learning_rate": 5.132175218798221e-07,
+      "loss": 0.1741,
+      "step": 7948
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0134837594097874,
+      "learning_rate": 5.131142837691865e-07,
+      "loss": 0.1174,
+      "step": 7949
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.01198196378554,
+      "learning_rate": 5.130110450990693e-07,
+      "loss": 0.1294,
+      "step": 7950
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.56648443745084,
+      "learning_rate": 5.129078058738747e-07,
+      "loss": 0.0392,
+      "step": 7951
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2297248998717523,
+      "learning_rate": 5.128045660980072e-07,
+      "loss": 0.0891,
+      "step": 7952
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4581243116909357,
+      "learning_rate": 5.127013257758712e-07,
+      "loss": 0.1297,
+      "step": 7953
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9345380450389321,
+      "learning_rate": 5.125980849118712e-07,
+      "loss": 0.189,
+      "step": 7954
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7370901183156554,
+      "learning_rate": 5.124948435104114e-07,
+      "loss": 0.1185,
+      "step": 7955
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7193480222061377,
+      "learning_rate": 5.123916015758964e-07,
+      "loss": 0.2652,
+      "step": 7956
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6466032299607144,
+      "learning_rate": 5.122883591127309e-07,
+      "loss": 0.1024,
+      "step": 7957
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7860832056788873,
+      "learning_rate": 5.121851161253192e-07,
+      "loss": 0.226,
+      "step": 7958
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6013005123668494,
+      "learning_rate": 5.120818726180661e-07,
+      "loss": 0.0711,
+      "step": 7959
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.064753175786098,
+      "learning_rate": 5.11978628595376e-07,
+      "loss": 0.1304,
+      "step": 7960
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9750951139813014,
+      "learning_rate": 5.118753840616535e-07,
+      "loss": 0.0555,
+      "step": 7961
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9487612432620487,
+      "learning_rate": 5.117721390213033e-07,
+      "loss": 0.4431,
+      "step": 7962
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3780561210469224,
+      "learning_rate": 5.116688934787299e-07,
+      "loss": 0.227,
+      "step": 7963
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9622208866925224,
+      "learning_rate": 5.11565647438338e-07,
+      "loss": 0.2718,
+      "step": 7964
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3846857939877564,
+      "learning_rate": 5.114624009045324e-07,
+      "loss": 0.4056,
+      "step": 7965
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6896302473879041,
+      "learning_rate": 5.113591538817176e-07,
+      "loss": 0.0749,
+      "step": 7966
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.1611748449305255,
+      "learning_rate": 5.112559063742986e-07,
+      "loss": 0.3602,
+      "step": 7967
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9255598049382037,
+      "learning_rate": 5.111526583866799e-07,
+      "loss": 0.2475,
+      "step": 7968
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5252074153445964,
+      "learning_rate": 5.110494099232665e-07,
+      "loss": 0.1219,
+      "step": 7969
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.55610246662506,
+      "learning_rate": 5.109461609884631e-07,
+      "loss": 0.193,
+      "step": 7970
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7640356996064016,
+      "learning_rate": 5.108429115866744e-07,
+      "loss": 0.2268,
+      "step": 7971
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4296766191799689,
+      "learning_rate": 5.107396617223052e-07,
+      "loss": 0.2183,
+      "step": 7972
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.612663793624046,
+      "learning_rate": 5.106364113997607e-07,
+      "loss": 0.3109,
+      "step": 7973
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2406635949629878,
+      "learning_rate": 5.105331606234452e-07,
+      "loss": 0.2663,
+      "step": 7974
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3077453661861215,
+      "learning_rate": 5.10429909397764e-07,
+      "loss": 0.0636,
+      "step": 7975
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7591910911749146,
+      "learning_rate": 5.103266577271219e-07,
+      "loss": 0.0717,
+      "step": 7976
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3426273054482245,
+      "learning_rate": 5.102234056159239e-07,
+      "loss": 0.3577,
+      "step": 7977
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4971718407893129,
+      "learning_rate": 5.101201530685748e-07,
+      "loss": 0.127,
+      "step": 7978
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0861819445275214,
+      "learning_rate": 5.100169000894796e-07,
+      "loss": 0.1992,
+      "step": 7979
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9030532265442714,
+      "learning_rate": 5.099136466830434e-07,
+      "loss": 0.226,
+      "step": 7980
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1638092337308699,
+      "learning_rate": 5.09810392853671e-07,
+      "loss": 0.2711,
+      "step": 7981
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5815643900116362,
+      "learning_rate": 5.097071386057676e-07,
+      "loss": 0.2693,
+      "step": 7982
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3478150074781302,
+      "learning_rate": 5.096038839437381e-07,
+      "loss": 0.2425,
+      "step": 7983
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.34556668354772546,
+      "learning_rate": 5.095006288719875e-07,
+      "loss": 0.1279,
+      "step": 7984
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.696040851455979,
+      "learning_rate": 5.093973733949212e-07,
+      "loss": 0.2189,
+      "step": 7985
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8684467332166635,
+      "learning_rate": 5.09294117516944e-07,
+      "loss": 0.4641,
+      "step": 7986
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.106608551218204,
+      "learning_rate": 5.09190861242461e-07,
+      "loss": 0.283,
+      "step": 7987
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6784539215400707,
+      "learning_rate": 5.090876045758774e-07,
+      "loss": 0.3893,
+      "step": 7988
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.562374800966257,
+      "learning_rate": 5.089843475215983e-07,
+      "loss": 0.1535,
+      "step": 7989
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6739943047222257,
+      "learning_rate": 5.08881090084029e-07,
+      "loss": 0.2647,
+      "step": 7990
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4446705559212927,
+      "learning_rate": 5.087778322675744e-07,
+      "loss": 0.3003,
+      "step": 7991
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1347843340063282,
+      "learning_rate": 5.086745740766398e-07,
+      "loss": 0.3486,
+      "step": 7992
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6966441014613145,
+      "learning_rate": 5.085713155156305e-07,
+      "loss": 0.0802,
+      "step": 7993
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6872586725807186,
+      "learning_rate": 5.084680565889517e-07,
+      "loss": 0.209,
+      "step": 7994
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4954523590391869,
+      "learning_rate": 5.083647973010085e-07,
+      "loss": 0.0734,
+      "step": 7995
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.630303025252287,
+      "learning_rate": 5.082615376562063e-07,
+      "loss": 0.1912,
+      "step": 7996
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5332288454112831,
+      "learning_rate": 5.081582776589502e-07,
+      "loss": 0.2678,
+      "step": 7997
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9638121437248867,
+      "learning_rate": 5.080550173136456e-07,
+      "loss": 0.1524,
+      "step": 7998
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5817877016507166,
+      "learning_rate": 5.079517566246979e-07,
+      "loss": 0.2753,
+      "step": 7999
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7366627544026684,
+      "learning_rate": 5.078484955965121e-07,
+      "loss": 0.2207,
+      "step": 8000
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6617489013541173,
+      "learning_rate": 5.077452342334938e-07,
+      "loss": 0.1174,
+      "step": 8001
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.537758947180974,
+      "learning_rate": 5.076419725400482e-07,
+      "loss": 0.1951,
+      "step": 8002
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.894337567628483,
+      "learning_rate": 5.075387105205809e-07,
+      "loss": 0.0404,
+      "step": 8003
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.408447184632947,
+      "learning_rate": 5.074354481794968e-07,
+      "loss": 0.0025,
+      "step": 8004
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2473253288384847,
+      "learning_rate": 5.073321855212016e-07,
+      "loss": 0.0523,
+      "step": 8005
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.38759903538036455,
+      "learning_rate": 5.072289225501007e-07,
+      "loss": 0.0103,
+      "step": 8006
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7517533132809295,
+      "learning_rate": 5.071256592705993e-07,
+      "loss": 0.082,
+      "step": 8007
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.21340932328299736,
+      "learning_rate": 5.07022395687103e-07,
+      "loss": 0.0664,
+      "step": 8008
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6888549592613122,
+      "learning_rate": 5.069191318040171e-07,
+      "loss": 0.1864,
+      "step": 8009
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6952146654491679,
+      "learning_rate": 5.068158676257471e-07,
+      "loss": 0.2177,
+      "step": 8010
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.21022522516618827,
+      "learning_rate": 5.067126031566987e-07,
+      "loss": 0.009,
+      "step": 8011
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3295249984265424,
+      "learning_rate": 5.06609338401277e-07,
+      "loss": 0.2187,
+      "step": 8012
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3381397979980587,
+      "learning_rate": 5.065060733638877e-07,
+      "loss": 0.2721,
+      "step": 8013
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7951994549328867,
+      "learning_rate": 5.064028080489363e-07,
+      "loss": 0.0808,
+      "step": 8014
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.208335598045493,
+      "learning_rate": 5.062995424608283e-07,
+      "loss": 0.3566,
+      "step": 8015
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9711718417727587,
+      "learning_rate": 5.061962766039691e-07,
+      "loss": 0.1127,
+      "step": 8016
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3681723380290377,
+      "learning_rate": 5.060930104827641e-07,
+      "loss": 0.0907,
+      "step": 8017
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7362546339027762,
+      "learning_rate": 5.05989744101619e-07,
+      "loss": 0.1672,
+      "step": 8018
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.690255265327017,
+      "learning_rate": 5.058864774649395e-07,
+      "loss": 0.1658,
+      "step": 8019
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.873125487517271,
+      "learning_rate": 5.05783210577131e-07,
+      "loss": 0.2203,
+      "step": 8020
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5923813617712705,
+      "learning_rate": 5.056799434425992e-07,
+      "loss": 0.2328,
+      "step": 8021
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8319052137680274,
+      "learning_rate": 5.055766760657496e-07,
+      "loss": 0.1973,
+      "step": 8022
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2575692636570914,
+      "learning_rate": 5.054734084509877e-07,
+      "loss": 0.1985,
+      "step": 8023
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0313095356288016,
+      "learning_rate": 5.053701406027192e-07,
+      "loss": 0.1867,
+      "step": 8024
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.44176030278131667,
+      "learning_rate": 5.052668725253498e-07,
+      "loss": 0.109,
+      "step": 8025
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6980213760967009,
+      "learning_rate": 5.051636042232849e-07,
+      "loss": 0.1877,
+      "step": 8026
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5090893499613691,
+      "learning_rate": 5.050603357009304e-07,
+      "loss": 0.1255,
+      "step": 8027
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.744190893066737,
+      "learning_rate": 5.049570669626917e-07,
+      "loss": 0.1226,
+      "step": 8028
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.417775319956826,
+      "learning_rate": 5.048537980129747e-07,
+      "loss": 0.0854,
+      "step": 8029
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.750635235017525,
+      "learning_rate": 5.047505288561847e-07,
+      "loss": 0.2062,
+      "step": 8030
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.683348780517243,
+      "learning_rate": 5.046472594967278e-07,
+      "loss": 0.1323,
+      "step": 8031
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.108180759399494,
+      "learning_rate": 5.045439899390094e-07,
+      "loss": 0.0988,
+      "step": 8032
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4179128445023675,
+      "learning_rate": 5.044407201874353e-07,
+      "loss": 0.0593,
+      "step": 8033
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.550250873218855,
+      "learning_rate": 5.04337450246411e-07,
+      "loss": 0.1507,
+      "step": 8034
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.48309367073510967,
+      "learning_rate": 5.042341801203424e-07,
+      "loss": 0.0078,
+      "step": 8035
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8426078244778038,
+      "learning_rate": 5.041309098136351e-07,
+      "loss": 0.1387,
+      "step": 8036
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.48805343814886276,
+      "learning_rate": 5.04027639330695e-07,
+      "loss": 0.2846,
+      "step": 8037
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 14.76702358007144,
+      "learning_rate": 5.039243686759277e-07,
+      "loss": 0.1401,
+      "step": 8038
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3855713237176372,
+      "learning_rate": 5.038210978537388e-07,
+      "loss": 0.0997,
+      "step": 8039
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4823045750901206,
+      "learning_rate": 5.037178268685344e-07,
+      "loss": 0.1423,
+      "step": 8040
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6589906452343282,
+      "learning_rate": 5.036145557247199e-07,
+      "loss": 0.0878,
+      "step": 8041
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1415394052155747,
+      "learning_rate": 5.035112844267014e-07,
+      "loss": 0.0197,
+      "step": 8042
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6135851235172674,
+      "learning_rate": 5.034080129788843e-07,
+      "loss": 0.1854,
+      "step": 8043
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6905787597383316,
+      "learning_rate": 5.033047413856745e-07,
+      "loss": 0.2306,
+      "step": 8044
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7618880486185982,
+      "learning_rate": 5.032014696514776e-07,
+      "loss": 0.2863,
+      "step": 8045
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1490263618140528,
+      "learning_rate": 5.030981977806998e-07,
+      "loss": 0.2538,
+      "step": 8046
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.41906248444879696,
+      "learning_rate": 5.029949257777466e-07,
+      "loss": 0.1505,
+      "step": 8047
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6151812296385359,
+      "learning_rate": 5.028916536470239e-07,
+      "loss": 0.1525,
+      "step": 8048
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.321947098669094,
+      "learning_rate": 5.027883813929373e-07,
+      "loss": 0.255,
+      "step": 8049
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.0067335337780323,
+      "learning_rate": 5.026851090198929e-07,
+      "loss": 0.5148,
+      "step": 8050
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2197256165289168,
+      "learning_rate": 5.025818365322964e-07,
+      "loss": 0.2783,
+      "step": 8051
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4318952023099847,
+      "learning_rate": 5.024785639345534e-07,
+      "loss": 0.233,
+      "step": 8052
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.467107979145924,
+      "learning_rate": 5.023752912310699e-07,
+      "loss": 0.2264,
+      "step": 8053
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5960508151237881,
+      "learning_rate": 5.022720184262517e-07,
+      "loss": 0.0043,
+      "step": 8054
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7491023884257654,
+      "learning_rate": 5.021687455245046e-07,
+      "loss": 0.1087,
+      "step": 8055
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8026981893353763,
+      "learning_rate": 5.020654725302347e-07,
+      "loss": 0.1684,
+      "step": 8056
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6393626698686294,
+      "learning_rate": 5.019621994478473e-07,
+      "loss": 0.2286,
+      "step": 8057
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5835826189171106,
+      "learning_rate": 5.018589262817488e-07,
+      "loss": 0.2519,
+      "step": 8058
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.71906783880137,
+      "learning_rate": 5.017556530363445e-07,
+      "loss": 0.1609,
+      "step": 8059
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6192050745601048,
+      "learning_rate": 5.016523797160406e-07,
+      "loss": 0.0912,
+      "step": 8060
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5135148904147495,
+      "learning_rate": 5.015491063252429e-07,
+      "loss": 0.1984,
+      "step": 8061
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7676244659370985,
+      "learning_rate": 5.014458328683572e-07,
+      "loss": 0.0042,
+      "step": 8062
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.676370431133513,
+      "learning_rate": 5.013425593497892e-07,
+      "loss": 0.3381,
+      "step": 8063
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5266436808843096,
+      "learning_rate": 5.012392857739452e-07,
+      "loss": 0.1857,
+      "step": 8064
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9299334948981965,
+      "learning_rate": 5.011360121452306e-07,
+      "loss": 0.5212,
+      "step": 8065
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4985207563224874,
+      "learning_rate": 5.010327384680515e-07,
+      "loss": 0.0652,
+      "step": 8066
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2844133747117827,
+      "learning_rate": 5.009294647468136e-07,
+      "loss": 0.109,
+      "step": 8067
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.378840373785456,
+      "learning_rate": 5.00826190985923e-07,
+      "loss": 0.091,
+      "step": 8068
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2423092356654564,
+      "learning_rate": 5.007229171897854e-07,
+      "loss": 0.262,
+      "step": 8069
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6809100805891322,
+      "learning_rate": 5.006196433628068e-07,
+      "loss": 0.133,
+      "step": 8070
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.807786554075676,
+      "learning_rate": 5.005163695093927e-07,
+      "loss": 0.2683,
+      "step": 8071
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6403892779241127,
+      "learning_rate": 5.004130956339494e-07,
+      "loss": 0.0859,
+      "step": 8072
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7590827412452761,
+      "learning_rate": 5.003098217408826e-07,
+      "loss": 0.0805,
+      "step": 8073
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.672439477342041,
+      "learning_rate": 5.002065478345982e-07,
+      "loss": 0.2191,
+      "step": 8074
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9034156078321618,
+      "learning_rate": 5.00103273919502e-07,
+      "loss": 0.0136,
+      "step": 8075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9520891091385693,
+      "learning_rate": 5e-07,
+      "loss": 0.1591,
+      "step": 8076
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0735962217333608,
+      "learning_rate": 4.998967260804982e-07,
+      "loss": 0.363,
+      "step": 8077
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0739018029784582,
+      "learning_rate": 4.997934521654018e-07,
+      "loss": 0.2883,
+      "step": 8078
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7644867479130036,
+      "learning_rate": 4.996901782591174e-07,
+      "loss": 0.3885,
+      "step": 8079
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7211748433681862,
+      "learning_rate": 4.995869043660506e-07,
+      "loss": 0.079,
+      "step": 8080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3702103745019236,
+      "learning_rate": 4.994836304906073e-07,
+      "loss": 0.0758,
+      "step": 8081
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0644432870665215,
+      "learning_rate": 4.993803566371933e-07,
+      "loss": 0.0639,
+      "step": 8082
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8060213093348012,
+      "learning_rate": 4.992770828102147e-07,
+      "loss": 0.3468,
+      "step": 8083
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5310872523741449,
+      "learning_rate": 4.991738090140769e-07,
+      "loss": 0.3825,
+      "step": 8084
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6983589634138904,
+      "learning_rate": 4.990705352531864e-07,
+      "loss": 0.1557,
+      "step": 8085
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5453860264192577,
+      "learning_rate": 4.989672615319485e-07,
+      "loss": 0.2753,
+      "step": 8086
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.588703415308047,
+      "learning_rate": 4.988639878547694e-07,
+      "loss": 0.1344,
+      "step": 8087
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8166596433992229,
+      "learning_rate": 4.987607142260548e-07,
+      "loss": 0.2103,
+      "step": 8088
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1569791097214266,
+      "learning_rate": 4.986574406502107e-07,
+      "loss": 0.1128,
+      "step": 8089
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.8820478279134125,
+      "learning_rate": 4.98554167131643e-07,
+      "loss": 0.2281,
+      "step": 8090
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.1672236181499045,
+      "learning_rate": 4.984508936747572e-07,
+      "loss": 0.1638,
+      "step": 8091
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7543266365961565,
+      "learning_rate": 4.983476202839594e-07,
+      "loss": 0.1106,
+      "step": 8092
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8841857722847184,
+      "learning_rate": 4.982443469636555e-07,
+      "loss": 0.1923,
+      "step": 8093
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.9459934202711238,
+      "learning_rate": 4.981410737182514e-07,
+      "loss": 0.0412,
+      "step": 8094
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3946510977912714,
+      "learning_rate": 4.980378005521527e-07,
+      "loss": 0.1793,
+      "step": 8095
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1675751686893905,
+      "learning_rate": 4.979345274697654e-07,
+      "loss": 0.0289,
+      "step": 8096
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.47801455840366996,
+      "learning_rate": 4.978312544754953e-07,
+      "loss": 0.3539,
+      "step": 8097
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5212935629372658,
+      "learning_rate": 4.977279815737482e-07,
+      "loss": 0.0023,
+      "step": 8098
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8551706395574106,
+      "learning_rate": 4.976247087689301e-07,
+      "loss": 0.0983,
+      "step": 8099
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0236467719199167,
+      "learning_rate": 4.975214360654467e-07,
+      "loss": 0.2156,
+      "step": 8100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.770254602796096,
+      "learning_rate": 4.974181634677036e-07,
+      "loss": 0.1489,
+      "step": 8101
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3393677786302405,
+      "learning_rate": 4.973148909801071e-07,
+      "loss": 0.1005,
+      "step": 8102
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2674565230397858,
+      "learning_rate": 4.972116186070625e-07,
+      "loss": 0.2204,
+      "step": 8103
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5216410523777196,
+      "learning_rate": 4.971083463529762e-07,
+      "loss": 0.2565,
+      "step": 8104
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5289063925044168,
+      "learning_rate": 4.970050742222536e-07,
+      "loss": 0.2912,
+      "step": 8105
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3860747416830018,
+      "learning_rate": 4.969018022193003e-07,
+      "loss": 0.1317,
+      "step": 8106
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4942623652535483,
+      "learning_rate": 4.967985303485224e-07,
+      "loss": 0.1161,
+      "step": 8107
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8004291099596728,
+      "learning_rate": 4.966952586143256e-07,
+      "loss": 0.2606,
+      "step": 8108
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7968568147246406,
+      "learning_rate": 4.965919870211159e-07,
+      "loss": 0.2674,
+      "step": 8109
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6414534403017824,
+      "learning_rate": 4.964887155732987e-07,
+      "loss": 0.1584,
+      "step": 8110
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9856321369312853,
+      "learning_rate": 4.9638544427528e-07,
+      "loss": 0.2173,
+      "step": 8111
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5929900285271503,
+      "learning_rate": 4.962821731314655e-07,
+      "loss": 0.3186,
+      "step": 8112
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1246579842009279,
+      "learning_rate": 4.961789021462611e-07,
+      "loss": 0.1268,
+      "step": 8113
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6573588297530798,
+      "learning_rate": 4.960756313240723e-07,
+      "loss": 0.1884,
+      "step": 8114
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0487922556717102,
+      "learning_rate": 4.959723606693051e-07,
+      "loss": 0.1636,
+      "step": 8115
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0779077508305832,
+      "learning_rate": 4.958690901863648e-07,
+      "loss": 0.1861,
+      "step": 8116
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.667903773938721,
+      "learning_rate": 4.957658198796577e-07,
+      "loss": 0.2523,
+      "step": 8117
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.47649954407524425,
+      "learning_rate": 4.956625497535892e-07,
+      "loss": 0.3787,
+      "step": 8118
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8589198922433372,
+      "learning_rate": 4.955592798125648e-07,
+      "loss": 0.4041,
+      "step": 8119
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6860354350736951,
+      "learning_rate": 4.954560100609908e-07,
+      "loss": 0.2068,
+      "step": 8120
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.332317558630018,
+      "learning_rate": 4.953527405032723e-07,
+      "loss": 0.3692,
+      "step": 8121
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.933375547824541,
+      "learning_rate": 4.952494711438154e-07,
+      "loss": 0.2179,
+      "step": 8122
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0123585012208458,
+      "learning_rate": 4.951462019870254e-07,
+      "loss": 0.1651,
+      "step": 8123
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6589280995115026,
+      "learning_rate": 4.950429330373082e-07,
+      "loss": 0.2217,
+      "step": 8124
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8875536056595668,
+      "learning_rate": 4.949396642990697e-07,
+      "loss": 0.1683,
+      "step": 8125
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0292126746650092,
+      "learning_rate": 4.94836395776715e-07,
+      "loss": 0.3058,
+      "step": 8126
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6532360720076447,
+      "learning_rate": 4.947331274746502e-07,
+      "loss": 0.0899,
+      "step": 8127
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.44401196696336753,
+      "learning_rate": 4.946298593972808e-07,
+      "loss": 0.0918,
+      "step": 8128
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8374962262196353,
+      "learning_rate": 4.945265915490121e-07,
+      "loss": 0.3585,
+      "step": 8129
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.34241429195748,
+      "learning_rate": 4.944233239342504e-07,
+      "loss": 0.0184,
+      "step": 8130
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6473484301622712,
+      "learning_rate": 4.943200565574007e-07,
+      "loss": 0.1428,
+      "step": 8131
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 12.825841666555736,
+      "learning_rate": 4.942167894228689e-07,
+      "loss": 0.1661,
+      "step": 8132
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4350198260722425,
+      "learning_rate": 4.941135225350605e-07,
+      "loss": 0.304,
+      "step": 8133
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.22369393361009066,
+      "learning_rate": 4.94010255898381e-07,
+      "loss": 0.0747,
+      "step": 8134
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.16485481347532424,
+      "learning_rate": 4.93906989517236e-07,
+      "loss": 0.089,
+      "step": 8135
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6101082522220975,
+      "learning_rate": 4.938037233960311e-07,
+      "loss": 0.1237,
+      "step": 8136
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3702393562442652,
+      "learning_rate": 4.937004575391719e-07,
+      "loss": 0.1818,
+      "step": 8137
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.872984812179066,
+      "learning_rate": 4.935971919510636e-07,
+      "loss": 0.0108,
+      "step": 8138
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.944398182559204,
+      "learning_rate": 4.934939266361123e-07,
+      "loss": 0.0112,
+      "step": 8139
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0587341436362527,
+      "learning_rate": 4.933906615987229e-07,
+      "loss": 0.2331,
+      "step": 8140
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0097210622684705,
+      "learning_rate": 4.932873968433014e-07,
+      "loss": 0.0713,
+      "step": 8141
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9011634437214178,
+      "learning_rate": 4.931841323742528e-07,
+      "loss": 0.0172,
+      "step": 8142
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0085254744974845,
+      "learning_rate": 4.930808681959829e-07,
+      "loss": 0.209,
+      "step": 8143
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6101017080712451,
+      "learning_rate": 4.92977604312897e-07,
+      "loss": 0.0775,
+      "step": 8144
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0447089952474162,
+      "learning_rate": 4.928743407294008e-07,
+      "loss": 0.1055,
+      "step": 8145
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.546086501221723,
+      "learning_rate": 4.927710774498995e-07,
+      "loss": 0.2003,
+      "step": 8146
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4433914650160102,
+      "learning_rate": 4.926678144787985e-07,
+      "loss": 0.1689,
+      "step": 8147
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7219197014496946,
+      "learning_rate": 4.925645518205033e-07,
+      "loss": 0.2977,
+      "step": 8148
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9241352289020095,
+      "learning_rate": 4.924612894794192e-07,
+      "loss": 0.2784,
+      "step": 8149
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6593509286864717,
+      "learning_rate": 4.923580274599518e-07,
+      "loss": 0.2903,
+      "step": 8150
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9204912285373574,
+      "learning_rate": 4.922547657665061e-07,
+      "loss": 0.1956,
+      "step": 8151
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.264522924498254,
+      "learning_rate": 4.921515044034878e-07,
+      "loss": 0.3374,
+      "step": 8152
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5032700619783951,
+      "learning_rate": 4.920482433753021e-07,
+      "loss": 0.2798,
+      "step": 8153
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.956876684470016,
+      "learning_rate": 4.919449826863544e-07,
+      "loss": 0.2016,
+      "step": 8154
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.7909706573922373,
+      "learning_rate": 4.918417223410497e-07,
+      "loss": 0.213,
+      "step": 8155
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7344354376992372,
+      "learning_rate": 4.917384623437937e-07,
+      "loss": 0.2235,
+      "step": 8156
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4713305981333957,
+      "learning_rate": 4.916352026989914e-07,
+      "loss": 0.0422,
+      "step": 8157
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5609733558667588,
+      "learning_rate": 4.915319434110484e-07,
+      "loss": 0.1526,
+      "step": 8158
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.303771899028892,
+      "learning_rate": 4.914286844843695e-07,
+      "loss": 0.158,
+      "step": 8159
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3765189718166573,
+      "learning_rate": 4.913254259233602e-07,
+      "loss": 0.3024,
+      "step": 8160
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.322163771166202,
+      "learning_rate": 4.912221677324257e-07,
+      "loss": 0.1188,
+      "step": 8161
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.397630194609182,
+      "learning_rate": 4.911189099159711e-07,
+      "loss": 0.0038,
+      "step": 8162
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7918383108872507,
+      "learning_rate": 4.910156524784017e-07,
+      "loss": 0.1372,
+      "step": 8163
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5311296124532012,
+      "learning_rate": 4.909123954241225e-07,
+      "loss": 0.1052,
+      "step": 8164
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.707926056032602,
+      "learning_rate": 4.908091387575391e-07,
+      "loss": 0.175,
+      "step": 8165
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5344325190448418,
+      "learning_rate": 4.90705882483056e-07,
+      "loss": 0.1296,
+      "step": 8166
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.700488779754395,
+      "learning_rate": 4.906026266050788e-07,
+      "loss": 0.4026,
+      "step": 8167
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.32097578211959354,
+      "learning_rate": 4.904993711280124e-07,
+      "loss": 0.1006,
+      "step": 8168
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8326151176037822,
+      "learning_rate": 4.903961160562618e-07,
+      "loss": 0.0026,
+      "step": 8169
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3794740402403776,
+      "learning_rate": 4.902928613942325e-07,
+      "loss": 0.1105,
+      "step": 8170
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.33940217049646326,
+      "learning_rate": 4.901896071463289e-07,
+      "loss": 0.2032,
+      "step": 8171
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3758703619299881,
+      "learning_rate": 4.900863533169566e-07,
+      "loss": 0.3885,
+      "step": 8172
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.611254928262428,
+      "learning_rate": 4.899830999105204e-07,
+      "loss": 0.1637,
+      "step": 8173
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.332136382986466,
+      "learning_rate": 4.898798469314253e-07,
+      "loss": 0.1619,
+      "step": 8174
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6545439993081186,
+      "learning_rate": 4.89776594384076e-07,
+      "loss": 0.1125,
+      "step": 8175
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6373618442771709,
+      "learning_rate": 4.896733422728781e-07,
+      "loss": 0.2989,
+      "step": 8176
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0231394374024527,
+      "learning_rate": 4.895700906022359e-07,
+      "loss": 0.315,
+      "step": 8177
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6125522670776106,
+      "learning_rate": 4.894668393765548e-07,
+      "loss": 0.2293,
+      "step": 8178
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6645183435609646,
+      "learning_rate": 4.893635886002394e-07,
+      "loss": 0.2517,
+      "step": 8179
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8889959015194648,
+      "learning_rate": 4.892603382776947e-07,
+      "loss": 0.165,
+      "step": 8180
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.960100897863591,
+      "learning_rate": 4.891570884133255e-07,
+      "loss": 0.1212,
+      "step": 8181
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4169013265792246,
+      "learning_rate": 4.89053839011537e-07,
+      "loss": 0.1354,
+      "step": 8182
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 15.974883256278313,
+      "learning_rate": 4.889505900767334e-07,
+      "loss": 0.4403,
+      "step": 8183
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4604167259060572,
+      "learning_rate": 4.8884734161332e-07,
+      "loss": 0.2273,
+      "step": 8184
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.206047113363974,
+      "learning_rate": 4.887440936257013e-07,
+      "loss": 0.2439,
+      "step": 8185
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4879592225809976,
+      "learning_rate": 4.886408461182824e-07,
+      "loss": 0.1288,
+      "step": 8186
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8666582105075445,
+      "learning_rate": 4.885375990954677e-07,
+      "loss": 0.0935,
+      "step": 8187
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5234728481060329,
+      "learning_rate": 4.88434352561662e-07,
+      "loss": 0.1426,
+      "step": 8188
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5205276731923292,
+      "learning_rate": 4.883311065212703e-07,
+      "loss": 0.1332,
+      "step": 8189
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6284872415583754,
+      "learning_rate": 4.882278609786968e-07,
+      "loss": 0.1104,
+      "step": 8190
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5106344655459321,
+      "learning_rate": 4.881246159383466e-07,
+      "loss": 0.0892,
+      "step": 8191
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7396302828353256,
+      "learning_rate": 4.88021371404624e-07,
+      "loss": 0.2086,
+      "step": 8192
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6956739818468076,
+      "learning_rate": 4.87918127381934e-07,
+      "loss": 0.3581,
+      "step": 8193
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0877921362422047,
+      "learning_rate": 4.878148838746806e-07,
+      "loss": 0.0218,
+      "step": 8194
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.673432447629786,
+      "learning_rate": 4.87711640887269e-07,
+      "loss": 0.1537,
+      "step": 8195
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.948180460074467,
+      "learning_rate": 4.876083984241035e-07,
+      "loss": 0.0797,
+      "step": 8196
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9047276678021269,
+      "learning_rate": 4.875051564895886e-07,
+      "loss": 0.2154,
+      "step": 8197
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.46054497364789626,
+      "learning_rate": 4.874019150881287e-07,
+      "loss": 0.0631,
+      "step": 8198
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9628242709982492,
+      "learning_rate": 4.872986742241288e-07,
+      "loss": 0.4154,
+      "step": 8199
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8853121666096614,
+      "learning_rate": 4.871954339019928e-07,
+      "loss": 0.0893,
+      "step": 8200
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8700554086125147,
+      "learning_rate": 4.870921941261252e-07,
+      "loss": 0.007,
+      "step": 8201
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.710502323168251,
+      "learning_rate": 4.869889549009308e-07,
+      "loss": 0.3938,
+      "step": 8202
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3829422967772256,
+      "learning_rate": 4.868857162308135e-07,
+      "loss": 0.1249,
+      "step": 8203
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.791836083047208,
+      "learning_rate": 4.86782478120178e-07,
+      "loss": 0.293,
+      "step": 8204
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.055559361123662,
+      "learning_rate": 4.866792405734286e-07,
+      "loss": 0.1588,
+      "step": 8205
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.12669878074895569,
+      "learning_rate": 4.865760035949695e-07,
+      "loss": 0.0149,
+      "step": 8206
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0831513869789322,
+      "learning_rate": 4.864727671892049e-07,
+      "loss": 0.1228,
+      "step": 8207
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.674063684024899,
+      "learning_rate": 4.863695313605397e-07,
+      "loss": 0.1532,
+      "step": 8208
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.784761189667426,
+      "learning_rate": 4.862662961133773e-07,
+      "loss": 0.2814,
+      "step": 8209
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3034063368760405,
+      "learning_rate": 4.861630614521225e-07,
+      "loss": 0.272,
+      "step": 8210
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6401069113083235,
+      "learning_rate": 4.860598273811792e-07,
+      "loss": 0.1154,
+      "step": 8211
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6507259723895588,
+      "learning_rate": 4.859565939049519e-07,
+      "loss": 0.3146,
+      "step": 8212
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7603477601240486,
+      "learning_rate": 4.858533610278443e-07,
+      "loss": 0.2414,
+      "step": 8213
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6837864369543493,
+      "learning_rate": 4.857501287542608e-07,
+      "loss": 0.1865,
+      "step": 8214
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6469380514526999,
+      "learning_rate": 4.856468970886056e-07,
+      "loss": 0.0477,
+      "step": 8215
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6632697507099637,
+      "learning_rate": 4.855436660352824e-07,
+      "loss": 0.1616,
+      "step": 8216
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1212084672513034,
+      "learning_rate": 4.854404355986958e-07,
+      "loss": 0.0772,
+      "step": 8217
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8867303347652737,
+      "learning_rate": 4.853372057832491e-07,
+      "loss": 0.1253,
+      "step": 8218
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.42147078461829746,
+      "learning_rate": 4.852339765933469e-07,
+      "loss": 0.2253,
+      "step": 8219
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0304215517377768,
+      "learning_rate": 4.851307480333929e-07,
+      "loss": 0.2961,
+      "step": 8220
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0333451791890018,
+      "learning_rate": 4.850275201077911e-07,
+      "loss": 0.232,
+      "step": 8221
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4349674322535713,
+      "learning_rate": 4.849242928209453e-07,
+      "loss": 0.2093,
+      "step": 8222
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2486247488705318,
+      "learning_rate": 4.848210661772595e-07,
+      "loss": 0.2181,
+      "step": 8223
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5097889743442834,
+      "learning_rate": 4.847178401811375e-07,
+      "loss": 0.357,
+      "step": 8224
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3235845555502364,
+      "learning_rate": 4.846146148369834e-07,
+      "loss": 0.075,
+      "step": 8225
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9198231270911172,
+      "learning_rate": 4.845113901492004e-07,
+      "loss": 0.1831,
+      "step": 8226
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8586642988102762,
+      "learning_rate": 4.844081661221929e-07,
+      "loss": 0.0167,
+      "step": 8227
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6726872701153576,
+      "learning_rate": 4.843049427603644e-07,
+      "loss": 0.211,
+      "step": 8228
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8395887871025307,
+      "learning_rate": 4.842017200681184e-07,
+      "loss": 0.0067,
+      "step": 8229
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2945067975585641,
+      "learning_rate": 4.840984980498589e-07,
+      "loss": 0.0833,
+      "step": 8230
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9712970287543637,
+      "learning_rate": 4.839952767099894e-07,
+      "loss": 0.1664,
+      "step": 8231
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2145687671104668,
+      "learning_rate": 4.838920560529136e-07,
+      "loss": 0.3168,
+      "step": 8232
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9919768412091605,
+      "learning_rate": 4.837888360830349e-07,
+      "loss": 0.1024,
+      "step": 8233
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1152289836489446,
+      "learning_rate": 4.836856168047574e-07,
+      "loss": 0.2503,
+      "step": 8234
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.43487629793650373,
+      "learning_rate": 4.835823982224839e-07,
+      "loss": 0.1746,
+      "step": 8235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7776933543405603,
+      "learning_rate": 4.834791803406186e-07,
+      "loss": 0.4996,
+      "step": 8236
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7138515491082422,
+      "learning_rate": 4.833759631635644e-07,
+      "loss": 0.1812,
+      "step": 8237
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7899423270112802,
+      "learning_rate": 4.832727466957254e-07,
+      "loss": 0.2266,
+      "step": 8238
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9439042794776742,
+      "learning_rate": 4.831695309415042e-07,
+      "loss": 0.1149,
+      "step": 8239
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5698927914817434,
+      "learning_rate": 4.83066315905305e-07,
+      "loss": 0.354,
+      "step": 8240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0048873877797317,
+      "learning_rate": 4.829631015915306e-07,
+      "loss": 0.035,
+      "step": 8241
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0012648658704233,
+      "learning_rate": 4.828598880045846e-07,
+      "loss": 0.134,
+      "step": 8242
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.793594013577755,
+      "learning_rate": 4.827566751488704e-07,
+      "loss": 0.1977,
+      "step": 8243
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.171597253648042,
+      "learning_rate": 4.826534630287908e-07,
+      "loss": 0.1717,
+      "step": 8244
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2448125237617151,
+      "learning_rate": 4.825502516487496e-07,
+      "loss": 0.2479,
+      "step": 8245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.377931561316245,
+      "learning_rate": 4.824470410131495e-07,
+      "loss": 0.1566,
+      "step": 8246
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.982739061442296,
+      "learning_rate": 4.823438311263942e-07,
+      "loss": 0.2143,
+      "step": 8247
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.45563477761264554,
+      "learning_rate": 4.822406219928863e-07,
+      "loss": 0.2273,
+      "step": 8248
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.088298522228207,
+      "learning_rate": 4.821374136170293e-07,
+      "loss": 0.1257,
+      "step": 8249
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.375275171763511,
+      "learning_rate": 4.820342060032261e-07,
+      "loss": 0.0603,
+      "step": 8250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.949710767535515,
+      "learning_rate": 4.819309991558798e-07,
+      "loss": 0.359,
+      "step": 8251
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.590046422935963,
+      "learning_rate": 4.818277930793933e-07,
+      "loss": 0.4137,
+      "step": 8252
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.485660221582689,
+      "learning_rate": 4.817245877781698e-07,
+      "loss": 0.1023,
+      "step": 8253
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2500216822562424,
+      "learning_rate": 4.816213832566121e-07,
+      "loss": 0.0481,
+      "step": 8254
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.501707110922901,
+      "learning_rate": 4.815181795191233e-07,
+      "loss": 0.1328,
+      "step": 8255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6027320587153924,
+      "learning_rate": 4.814149765701059e-07,
+      "loss": 0.1465,
+      "step": 8256
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0408515438402133,
+      "learning_rate": 4.81311774413963e-07,
+      "loss": 0.1136,
+      "step": 8257
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1432097162698354,
+      "learning_rate": 4.812085730550973e-07,
+      "loss": 0.1676,
+      "step": 8258
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0042730486274967,
+      "learning_rate": 4.811053724979116e-07,
+      "loss": 0.1835,
+      "step": 8259
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7901281894729675,
+      "learning_rate": 4.81002172746809e-07,
+      "loss": 0.1528,
+      "step": 8260
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.40423764933552675,
+      "learning_rate": 4.808989738061916e-07,
+      "loss": 0.0702,
+      "step": 8261
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.34243011998723405,
+      "learning_rate": 4.807957756804627e-07,
+      "loss": 0.053,
+      "step": 8262
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4038619697821567,
+      "learning_rate": 4.806925783740242e-07,
+      "loss": 0.1311,
+      "step": 8263
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6184916906991575,
+      "learning_rate": 4.805893818912794e-07,
+      "loss": 0.1294,
+      "step": 8264
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.62261932255158,
+      "learning_rate": 4.804861862366305e-07,
+      "loss": 0.1913,
+      "step": 8265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6886609271115808,
+      "learning_rate": 4.803829914144802e-07,
+      "loss": 0.4007,
+      "step": 8266
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7613069744678602,
+      "learning_rate": 4.802797974292308e-07,
+      "loss": 0.1288,
+      "step": 8267
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3144447465818885,
+      "learning_rate": 4.80176604285285e-07,
+      "loss": 0.2315,
+      "step": 8268
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.47126022938693496,
+      "learning_rate": 4.800734119870451e-07,
+      "loss": 0.1008,
+      "step": 8269
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4993473224307254,
+      "learning_rate": 4.799702205389134e-07,
+      "loss": 0.0805,
+      "step": 8270
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.826096688904793,
+      "learning_rate": 4.798670299452926e-07,
+      "loss": 0.2461,
+      "step": 8271
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1566726732922138,
+      "learning_rate": 4.797638402105844e-07,
+      "loss": 0.0678,
+      "step": 8272
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 17.1333157525917,
+      "learning_rate": 4.796606513391917e-07,
+      "loss": 0.1987,
+      "step": 8273
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9522791582660433,
+      "learning_rate": 4.795574633355164e-07,
+      "loss": 0.0395,
+      "step": 8274
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8561925291307796,
+      "learning_rate": 4.794542762039608e-07,
+      "loss": 0.2843,
+      "step": 8275
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.071059948220585,
+      "learning_rate": 4.793510899489272e-07,
+      "loss": 0.0931,
+      "step": 8276
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.473368344647713,
+      "learning_rate": 4.792479045748177e-07,
+      "loss": 0.2107,
+      "step": 8277
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9678500358403068,
+      "learning_rate": 4.79144720086034e-07,
+      "loss": 0.2189,
+      "step": 8278
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.70270008822198,
+      "learning_rate": 4.790415364869787e-07,
+      "loss": 0.2395,
+      "step": 8279
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7649746568862957,
+      "learning_rate": 4.789383537820534e-07,
+      "loss": 0.1648,
+      "step": 8280
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9893349122689812,
+      "learning_rate": 4.788351719756605e-07,
+      "loss": 0.2831,
+      "step": 8281
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6704188958555899,
+      "learning_rate": 4.787319910722015e-07,
+      "loss": 0.1827,
+      "step": 8282
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7719067889338662,
+      "learning_rate": 4.786288110760787e-07,
+      "loss": 0.0498,
+      "step": 8283
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9135942110973932,
+      "learning_rate": 4.785256319916936e-07,
+      "loss": 0.1649,
+      "step": 8284
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5359750663694142,
+      "learning_rate": 4.784224538234482e-07,
+      "loss": 0.1537,
+      "step": 8285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.112424574766902,
+      "learning_rate": 4.783192765757446e-07,
+      "loss": 0.2091,
+      "step": 8286
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5248860382975025,
+      "learning_rate": 4.782161002529839e-07,
+      "loss": 0.3016,
+      "step": 8287
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3642197574348445,
+      "learning_rate": 4.781129248595684e-07,
+      "loss": 0.3299,
+      "step": 8288
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9654919145150089,
+      "learning_rate": 4.780097503998993e-07,
+      "loss": 0.3039,
+      "step": 8289
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7129230089890654,
+      "learning_rate": 4.779065768783786e-07,
+      "loss": 0.2663,
+      "step": 8290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.492018950625692,
+      "learning_rate": 4.778034042994077e-07,
+      "loss": 0.0489,
+      "step": 8291
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.016346917646898,
+      "learning_rate": 4.777002326673883e-07,
+      "loss": 0.2143,
+      "step": 8292
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7027076661134657,
+      "learning_rate": 4.775970619867215e-07,
+      "loss": 0.1086,
+      "step": 8293
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7552783355285612,
+      "learning_rate": 4.774938922618094e-07,
+      "loss": 0.1274,
+      "step": 8294
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 17.187041279410437,
+      "learning_rate": 4.773907234970528e-07,
+      "loss": 0.3003,
+      "step": 8295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.561537872363602,
+      "learning_rate": 4.772875556968537e-07,
+      "loss": 0.2272,
+      "step": 8296
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8076328473818086,
+      "learning_rate": 4.771843888656131e-07,
+      "loss": 0.2778,
+      "step": 8297
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.30795960532586475,
+      "learning_rate": 4.770812230077322e-07,
+      "loss": 0.0026,
+      "step": 8298
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3285228663365966,
+      "learning_rate": 4.769780581276125e-07,
+      "loss": 0.1642,
+      "step": 8299
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.649480294227551,
+      "learning_rate": 4.768748942296549e-07,
+      "loss": 0.1039,
+      "step": 8300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5681854671972938,
+      "learning_rate": 4.7677173131826096e-07,
+      "loss": 0.1768,
+      "step": 8301
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8524605023286805,
+      "learning_rate": 4.7666856939783167e-07,
+      "loss": 0.163,
+      "step": 8302
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9369709721737559,
+      "learning_rate": 4.765654084727681e-07,
+      "loss": 0.1572,
+      "step": 8303
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.7409964722526485,
+      "learning_rate": 4.764622485474712e-07,
+      "loss": 0.3333,
+      "step": 8304
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7505081269417608,
+      "learning_rate": 4.763590896263423e-07,
+      "loss": 0.2993,
+      "step": 8305
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7082784780927405,
+      "learning_rate": 4.762559317137819e-07,
+      "loss": 0.4248,
+      "step": 8306
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7729870754292155,
+      "learning_rate": 4.7615277481419146e-07,
+      "loss": 0.1719,
+      "step": 8307
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8689005251881637,
+      "learning_rate": 4.7604961893197137e-07,
+      "loss": 0.3522,
+      "step": 8308
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.343606459450366,
+      "learning_rate": 4.759464640715229e-07,
+      "loss": 0.114,
+      "step": 8309
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.34035926840061,
+      "learning_rate": 4.758433102372465e-07,
+      "loss": 0.2122,
+      "step": 8310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7962579966207999,
+      "learning_rate": 4.7574015743354305e-07,
+      "loss": 0.1081,
+      "step": 8311
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7334214270279522,
+      "learning_rate": 4.756370056648133e-07,
+      "loss": 0.0126,
+      "step": 8312
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4473313375010857,
+      "learning_rate": 4.7553385493545766e-07,
+      "loss": 0.2959,
+      "step": 8313
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6052795437915548,
+      "learning_rate": 4.754307052498773e-07,
+      "loss": 0.2558,
+      "step": 8314
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.52268626158944,
+      "learning_rate": 4.753275566124721e-07,
+      "loss": 0.2504,
+      "step": 8315
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6730348211603852,
+      "learning_rate": 4.7522440902764323e-07,
+      "loss": 0.0809,
+      "step": 8316
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8276657829322512,
+      "learning_rate": 4.751212624997906e-07,
+      "loss": 0.3571,
+      "step": 8317
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0171656654102184,
+      "learning_rate": 4.7501811703331516e-07,
+      "loss": 0.3255,
+      "step": 8318
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3961201876532447,
+      "learning_rate": 4.7491497263261687e-07,
+      "loss": 0.314,
+      "step": 8319
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6164111459767151,
+      "learning_rate": 4.7481182930209633e-07,
+      "loss": 0.3303,
+      "step": 8320
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.327110020906333,
+      "learning_rate": 4.7470868704615387e-07,
+      "loss": 0.0943,
+      "step": 8321
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.27534488706768706,
+      "learning_rate": 4.746055458691896e-07,
+      "loss": 0.0504,
+      "step": 8322
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7571062987207461,
+      "learning_rate": 4.745024057756037e-07,
+      "loss": 0.1668,
+      "step": 8323
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7103832911043411,
+      "learning_rate": 4.743992667697966e-07,
+      "loss": 0.2673,
+      "step": 8324
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.40543707517658106,
+      "learning_rate": 4.742961288561684e-07,
+      "loss": 0.067,
+      "step": 8325
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.155291203497054,
+      "learning_rate": 4.7419299203911867e-07,
+      "loss": 0.1394,
+      "step": 8326
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.633389690731605,
+      "learning_rate": 4.74089856323048e-07,
+      "loss": 0.3341,
+      "step": 8327
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7601287600239229,
+      "learning_rate": 4.739867217123559e-07,
+      "loss": 0.1004,
+      "step": 8328
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.659000570859291,
+      "learning_rate": 4.738835882114427e-07,
+      "loss": 0.0744,
+      "step": 8329
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.727445611828362,
+      "learning_rate": 4.73780455824708e-07,
+      "loss": 0.2002,
+      "step": 8330
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7044838640912626,
+      "learning_rate": 4.73677324556552e-07,
+      "loss": 0.0962,
+      "step": 8331
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6210387091077625,
+      "learning_rate": 4.7357419441137404e-07,
+      "loss": 0.2306,
+      "step": 8332
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1456649589567296,
+      "learning_rate": 4.734710653935743e-07,
+      "loss": 0.0845,
+      "step": 8333
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8166196626153188,
+      "learning_rate": 4.73367937507552e-07,
+      "loss": 0.3285,
+      "step": 8334
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6904261910706524,
+      "learning_rate": 4.732648107577072e-07,
+      "loss": 0.1853,
+      "step": 8335
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4770212767922006,
+      "learning_rate": 4.7316168514843914e-07,
+      "loss": 0.1259,
+      "step": 8336
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6304307296852425,
+      "learning_rate": 4.7305856068414776e-07,
+      "loss": 0.0056,
+      "step": 8337
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.526324203251784,
+      "learning_rate": 4.729554373692323e-07,
+      "loss": 0.1671,
+      "step": 8338
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1042738785690747,
+      "learning_rate": 4.72852315208092e-07,
+      "loss": 0.124,
+      "step": 8339
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6621693609820087,
+      "learning_rate": 4.727491942051268e-07,
+      "loss": 0.1009,
+      "step": 8340
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6882980695951783,
+      "learning_rate": 4.7264607436473555e-07,
+      "loss": 0.2576,
+      "step": 8341
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 15.985279478692943,
+      "learning_rate": 4.725429556913179e-07,
+      "loss": 0.173,
+      "step": 8342
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9648885535488356,
+      "learning_rate": 4.724398381892728e-07,
+      "loss": 0.0932,
+      "step": 8343
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.839036168880391,
+      "learning_rate": 4.723367218629998e-07,
+      "loss": 0.0475,
+      "step": 8344
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0665152967812122,
+      "learning_rate": 4.722336067168977e-07,
+      "loss": 0.1767,
+      "step": 8345
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8784662023791162,
+      "learning_rate": 4.7213049275536584e-07,
+      "loss": 0.337,
+      "step": 8346
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.988119990288558,
+      "learning_rate": 4.720273799828031e-07,
+      "loss": 0.0906,
+      "step": 8347
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.33164467170179124,
+      "learning_rate": 4.7192426840360864e-07,
+      "loss": 0.2812,
+      "step": 8348
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5021048271459208,
+      "learning_rate": 4.718211580221812e-07,
+      "loss": 0.1779,
+      "step": 8349
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1088301088217718,
+      "learning_rate": 4.7171804884292e-07,
+      "loss": 0.2481,
+      "step": 8350
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5711181883692436,
+      "learning_rate": 4.7161494087022354e-07,
+      "loss": 0.2516,
+      "step": 8351
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.866843021786571,
+      "learning_rate": 4.7151183410849094e-07,
+      "loss": 0.3975,
+      "step": 8352
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.32527701729108605,
+      "learning_rate": 4.7140872856212077e-07,
+      "loss": 0.1093,
+      "step": 8353
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6621450184134362,
+      "learning_rate": 4.713056242355115e-07,
+      "loss": 0.0823,
+      "step": 8354
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3107572932525853,
+      "learning_rate": 4.7120252113306207e-07,
+      "loss": 0.1065,
+      "step": 8355
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.05815418833054292,
+      "learning_rate": 4.710994192591709e-07,
+      "loss": 0.0006,
+      "step": 8356
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3691203708536689,
+      "learning_rate": 4.709963186182368e-07,
+      "loss": 0.0926,
+      "step": 8357
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0321675826769945,
+      "learning_rate": 4.708932192146579e-07,
+      "loss": 0.3407,
+      "step": 8358
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2388699622096775,
+      "learning_rate": 4.707901210528329e-07,
+      "loss": 0.3973,
+      "step": 8359
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6445735646207422,
+      "learning_rate": 4.706870241371598e-07,
+      "loss": 0.1492,
+      "step": 8360
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6405927916498171,
+      "learning_rate": 4.705839284720375e-07,
+      "loss": 0.546,
+      "step": 8361
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.217280316421169,
+      "learning_rate": 4.704808340618637e-07,
+      "loss": 0.3646,
+      "step": 8362
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.32295731170163944,
+      "learning_rate": 4.703777409110371e-07,
+      "loss": 0.2454,
+      "step": 8363
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2450631581739244,
+      "learning_rate": 4.702746490239554e-07,
+      "loss": 0.0479,
+      "step": 8364
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6994199215400096,
+      "learning_rate": 4.70171558405017e-07,
+      "loss": 0.2139,
+      "step": 8365
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9752995049467115,
+      "learning_rate": 4.7006846905861995e-07,
+      "loss": 0.3142,
+      "step": 8366
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.627713157969607,
+      "learning_rate": 4.6996538098916205e-07,
+      "loss": 0.294,
+      "step": 8367
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.41530161662137777,
+      "learning_rate": 4.698622942010416e-07,
+      "loss": 0.0899,
+      "step": 8368
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5522485878580587,
+      "learning_rate": 4.69759208698656e-07,
+      "loss": 0.1956,
+      "step": 8369
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5530429781071238,
+      "learning_rate": 4.6965612448640354e-07,
+      "loss": 0.1169,
+      "step": 8370
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.636913940183296,
+      "learning_rate": 4.6955304156868165e-07,
+      "loss": 0.0016,
+      "step": 8371
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.949578232853312,
+      "learning_rate": 4.694499599498884e-07,
+      "loss": 0.1755,
+      "step": 8372
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6750312131157368,
+      "learning_rate": 4.69346879634421e-07,
+      "loss": 0.2282,
+      "step": 8373
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.723795459386764,
+      "learning_rate": 4.692438006266775e-07,
+      "loss": 0.0811,
+      "step": 8374
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.39597401424435147,
+      "learning_rate": 4.6914072293105523e-07,
+      "loss": 0.0521,
+      "step": 8375
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5125918304620144,
+      "learning_rate": 4.690376465519519e-07,
+      "loss": 0.1816,
+      "step": 8376
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.15254802116096794,
+      "learning_rate": 4.6893457149376464e-07,
+      "loss": 0.0027,
+      "step": 8377
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1792409719099959,
+      "learning_rate": 4.688314977608912e-07,
+      "loss": 0.2804,
+      "step": 8378
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.00172599190676,
+      "learning_rate": 4.6872842535772874e-07,
+      "loss": 0.2656,
+      "step": 8379
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8731598878152553,
+      "learning_rate": 4.686253542886743e-07,
+      "loss": 0.132,
+      "step": 8380
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0209030059932318,
+      "learning_rate": 4.6852228455812535e-07,
+      "loss": 0.0621,
+      "step": 8381
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5910051250160278,
+      "learning_rate": 4.684192161704791e-07,
+      "loss": 0.3391,
+      "step": 8382
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5875092722322819,
+      "learning_rate": 4.683161491301326e-07,
+      "loss": 0.3366,
+      "step": 8383
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5832849105154081,
+      "learning_rate": 4.6821308344148267e-07,
+      "loss": 0.1486,
+      "step": 8384
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2557154629724138,
+      "learning_rate": 4.681100191089268e-07,
+      "loss": 0.185,
+      "step": 8385
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8808627732116225,
+      "learning_rate": 4.6800695613686135e-07,
+      "loss": 0.257,
+      "step": 8386
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.27462422795780117,
+      "learning_rate": 4.679038945296837e-07,
+      "loss": 0.1196,
+      "step": 8387
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6125359784602092,
+      "learning_rate": 4.6780083429179025e-07,
+      "loss": 0.2423,
+      "step": 8388
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.38922478139869265,
+      "learning_rate": 4.676977754275781e-07,
+      "loss": 0.0041,
+      "step": 8389
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8147472227342312,
+      "learning_rate": 4.6759471794144354e-07,
+      "loss": 0.2865,
+      "step": 8390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8615916354838382,
+      "learning_rate": 4.6749166183778367e-07,
+      "loss": 0.281,
+      "step": 8391
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6019648112880962,
+      "learning_rate": 4.673886071209948e-07,
+      "loss": 0.0172,
+      "step": 8392
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7827091748181172,
+      "learning_rate": 4.6728555379547353e-07,
+      "loss": 0.1966,
+      "step": 8393
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.40517344795530413,
+      "learning_rate": 4.6718250186561637e-07,
+      "loss": 0.1911,
+      "step": 8394
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.37609331752727443,
+      "learning_rate": 4.6707945133581944e-07,
+      "loss": 0.0497,
+      "step": 8395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8233073465419235,
+      "learning_rate": 4.669764022104795e-07,
+      "loss": 0.1986,
+      "step": 8396
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7806807328458075,
+      "learning_rate": 4.6687335449399246e-07,
+      "loss": 0.223,
+      "step": 8397
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9744470360724538,
+      "learning_rate": 4.6677030819075494e-07,
+      "loss": 0.1456,
+      "step": 8398
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9128355771847734,
+      "learning_rate": 4.666672633051625e-07,
+      "loss": 0.3131,
+      "step": 8399
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1988994555552688,
+      "learning_rate": 4.6656421984161185e-07,
+      "loss": 0.2688,
+      "step": 8400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7766678877976095,
+      "learning_rate": 4.6646117780449875e-07,
+      "loss": 0.2256,
+      "step": 8401
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7061747127141389,
+      "learning_rate": 4.663581371982194e-07,
+      "loss": 0.1532,
+      "step": 8402
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.173840406685293,
+      "learning_rate": 4.662550980271692e-07,
+      "loss": 0.2169,
+      "step": 8403
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6232501211624247,
+      "learning_rate": 4.661520602957447e-07,
+      "loss": 0.0222,
+      "step": 8404
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.51348777533686,
+      "learning_rate": 4.6604902400834113e-07,
+      "loss": 0.3304,
+      "step": 8405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5824220708737986,
+      "learning_rate": 4.659459891693546e-07,
+      "loss": 0.3077,
+      "step": 8406
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5160647659160633,
+      "learning_rate": 4.6584295578318055e-07,
+      "loss": 0.1435,
+      "step": 8407
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3249734204797299,
+      "learning_rate": 4.657399238542147e-07,
+      "loss": 0.1361,
+      "step": 8408
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.601539827843137,
+      "learning_rate": 4.6563689338685253e-07,
+      "loss": 0.2454,
+      "step": 8409
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4928277164362475,
+      "learning_rate": 4.6553386438548946e-07,
+      "loss": 0.0144,
+      "step": 8410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1178534249758558,
+      "learning_rate": 4.6543083685452127e-07,
+      "loss": 0.4038,
+      "step": 8411
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.69181954222864,
+      "learning_rate": 4.6532781079834284e-07,
+      "loss": 0.1806,
+      "step": 8412
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.023451811323160213,
+      "learning_rate": 4.652247862213499e-07,
+      "loss": 0.0003,
+      "step": 8413
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7853279716942542,
+      "learning_rate": 4.6512176312793735e-07,
+      "loss": 0.0061,
+      "step": 8414
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6417682889599597,
+      "learning_rate": 4.6501874152250064e-07,
+      "loss": 0.1982,
+      "step": 8415
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5193905532521556,
+      "learning_rate": 4.6491572140943455e-07,
+      "loss": 0.1116,
+      "step": 8416
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7000523213032452,
+      "learning_rate": 4.6481270279313433e-07,
+      "loss": 0.304,
+      "step": 8417
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8775827363867574,
+      "learning_rate": 4.647096856779951e-07,
+      "loss": 0.1525,
+      "step": 8418
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.257623169315283,
+      "learning_rate": 4.646066700684115e-07,
+      "loss": 0.0711,
+      "step": 8419
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3113897164922417,
+      "learning_rate": 4.6450365596877867e-07,
+      "loss": 0.059,
+      "step": 8420
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7721374090663957,
+      "learning_rate": 4.6440064338349094e-07,
+      "loss": 0.1587,
+      "step": 8421
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2158504599664803,
+      "learning_rate": 4.642976323169436e-07,
+      "loss": 0.1179,
+      "step": 8422
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3499208462900647,
+      "learning_rate": 4.6419462277353083e-07,
+      "loss": 0.2684,
+      "step": 8423
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.922887618861537,
+      "learning_rate": 4.640916147576477e-07,
+      "loss": 0.0658,
+      "step": 8424
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5540538560114585,
+      "learning_rate": 4.6398860827368814e-07,
+      "loss": 0.1489,
+      "step": 8425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0657525259003775,
+      "learning_rate": 4.638856033260471e-07,
+      "loss": 0.2786,
+      "step": 8426
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.38324265983943734,
+      "learning_rate": 4.6378259991911883e-07,
+      "loss": 0.0047,
+      "step": 8427
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9656010549995666,
+      "learning_rate": 4.6367959805729765e-07,
+      "loss": 0.226,
+      "step": 8428
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.998603730798484,
+      "learning_rate": 4.635765977449777e-07,
+      "loss": 0.094,
+      "step": 8429
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.548560762199273,
+      "learning_rate": 4.634735989865536e-07,
+      "loss": 0.0562,
+      "step": 8430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4566069762684165,
+      "learning_rate": 4.633706017864189e-07,
+      "loss": 0.114,
+      "step": 8431
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8559033881437175,
+      "learning_rate": 4.632676061489682e-07,
+      "loss": 0.1003,
+      "step": 8432
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8142044444980603,
+      "learning_rate": 4.6316461207859505e-07,
+      "loss": 0.1296,
+      "step": 8433
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5067751983888892,
+      "learning_rate": 4.630616195796938e-07,
+      "loss": 0.0145,
+      "step": 8434
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1498546862355845,
+      "learning_rate": 4.62958628656658e-07,
+      "loss": 0.3408,
+      "step": 8435
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.59984117843711,
+      "learning_rate": 4.628556393138815e-07,
+      "loss": 0.1739,
+      "step": 8436
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4816842370205916,
+      "learning_rate": 4.6275265155575824e-07,
+      "loss": 0.0509,
+      "step": 8437
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8461256816270033,
+      "learning_rate": 4.626496653866816e-07,
+      "loss": 0.3231,
+      "step": 8438
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0689898407931318,
+      "learning_rate": 4.6254668081104553e-07,
+      "loss": 0.3156,
+      "step": 8439
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.154742449093417,
+      "learning_rate": 4.624436978332431e-07,
+      "loss": 0.2319,
+      "step": 8440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7100348437553234,
+      "learning_rate": 4.6234071645766826e-07,
+      "loss": 0.2873,
+      "step": 8441
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5191890453423041,
+      "learning_rate": 4.622377366887139e-07,
+      "loss": 0.0466,
+      "step": 8442
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8436845098887154,
+      "learning_rate": 4.621347585307739e-07,
+      "loss": 0.2187,
+      "step": 8443
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4120815992148151,
+      "learning_rate": 4.620317819882409e-07,
+      "loss": 0.0053,
+      "step": 8444
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.656772350978441,
+      "learning_rate": 4.619288070655086e-07,
+      "loss": 0.1531,
+      "step": 8445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6586417685936933,
+      "learning_rate": 4.618258337669698e-07,
+      "loss": 0.239,
+      "step": 8446
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5419891264394243,
+      "learning_rate": 4.617228620970178e-07,
+      "loss": 0.1489,
+      "step": 8447
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8317930629719872,
+      "learning_rate": 4.6161989206004554e-07,
+      "loss": 0.2021,
+      "step": 8448
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.38792223643785884,
+      "learning_rate": 4.615169236604456e-07,
+      "loss": 0.0915,
+      "step": 8449
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9317987377358307,
+      "learning_rate": 4.614139569026113e-07,
+      "loss": 0.1291,
+      "step": 8450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.964238468234055,
+      "learning_rate": 4.613109917909349e-07,
+      "loss": 0.1983,
+      "step": 8451
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.504604358323638,
+      "learning_rate": 4.6120802832980946e-07,
+      "loss": 0.1325,
+      "step": 8452
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.193223609694503,
+      "learning_rate": 4.611050665236276e-07,
+      "loss": 0.5047,
+      "step": 8453
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.584229022981351,
+      "learning_rate": 4.610021063767817e-07,
+      "loss": 0.2107,
+      "step": 8454
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7681234642862357,
+      "learning_rate": 4.608991478936643e-07,
+      "loss": 0.1975,
+      "step": 8455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9747994156063038,
+      "learning_rate": 4.60796191078668e-07,
+      "loss": 0.2907,
+      "step": 8456
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6442174283300625,
+      "learning_rate": 4.6069323593618487e-07,
+      "loss": 0.1747,
+      "step": 8457
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0394665216571475,
+      "learning_rate": 4.605902824706074e-07,
+      "loss": 0.2331,
+      "step": 8458
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7345764613771602,
+      "learning_rate": 4.6048733068632763e-07,
+      "loss": 0.0262,
+      "step": 8459
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.364427066048286,
+      "learning_rate": 4.6038438058773795e-07,
+      "loss": 0.0444,
+      "step": 8460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7958588620663453,
+      "learning_rate": 4.6028143217922994e-07,
+      "loss": 0.1346,
+      "step": 8461
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.57771632222782,
+      "learning_rate": 4.60178485465196e-07,
+      "loss": 0.1572,
+      "step": 8462
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4540274020577097,
+      "learning_rate": 4.6007554045002807e-07,
+      "loss": 0.2118,
+      "step": 8463
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4574318972870368,
+      "learning_rate": 4.599725971381176e-07,
+      "loss": 0.0756,
+      "step": 8464
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2494177439012157,
+      "learning_rate": 4.598696555338569e-07,
+      "loss": 0.0699,
+      "step": 8465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.324517809174406,
+      "learning_rate": 4.5976671564163703e-07,
+      "loss": 0.1505,
+      "step": 8466
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8563665290211504,
+      "learning_rate": 4.5966377746585024e-07,
+      "loss": 0.0799,
+      "step": 8467
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.307458669269294,
+      "learning_rate": 4.5956084101088755e-07,
+      "loss": 0.3482,
+      "step": 8468
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8117454115871394,
+      "learning_rate": 4.594579062811409e-07,
+      "loss": 0.2052,
+      "step": 8469
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 22.86468592982254,
+      "learning_rate": 4.593549732810012e-07,
+      "loss": 0.2059,
+      "step": 8470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9233915420970702,
+      "learning_rate": 4.592520420148602e-07,
+      "loss": 0.1638,
+      "step": 8471
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9841653192358117,
+      "learning_rate": 4.591491124871089e-07,
+      "loss": 0.2057,
+      "step": 8472
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6514008796100588,
+      "learning_rate": 4.590461847021388e-07,
+      "loss": 0.1183,
+      "step": 8473
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.115142321159001,
+      "learning_rate": 4.589432586643405e-07,
+      "loss": 0.263,
+      "step": 8474
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6404165934906061,
+      "learning_rate": 4.5884033437810565e-07,
+      "loss": 0.388,
+      "step": 8475
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8905150331337415,
+      "learning_rate": 4.5873741184782484e-07,
+      "loss": 0.1492,
+      "step": 8476
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3374940481431025,
+      "learning_rate": 4.5863449107788877e-07,
+      "loss": 0.0189,
+      "step": 8477
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.837287076567836,
+      "learning_rate": 4.585315720726885e-07,
+      "loss": 0.5004,
+      "step": 8478
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.45436173566265,
+      "learning_rate": 4.584286548366148e-07,
+      "loss": 0.0123,
+      "step": 8479
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.07608838150686013,
+      "learning_rate": 4.583257393740583e-07,
+      "loss": 0.0126,
+      "step": 8480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.298241936440913,
+      "learning_rate": 4.582228256894093e-07,
+      "loss": 0.1864,
+      "step": 8481
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9557106428726254,
+      "learning_rate": 4.581199137870588e-07,
+      "loss": 0.1747,
+      "step": 8482
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4943259769523899,
+      "learning_rate": 4.580170036713968e-07,
+      "loss": 0.2299,
+      "step": 8483
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.991030265560546,
+      "learning_rate": 4.57914095346814e-07,
+      "loss": 0.1538,
+      "step": 8484
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0107885142531865,
+      "learning_rate": 4.578111888177003e-07,
+      "loss": 0.3783,
+      "step": 8485
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0740041697591793,
+      "learning_rate": 4.577082840884463e-07,
+      "loss": 0.211,
+      "step": 8486
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6396555243855557,
+      "learning_rate": 4.5760538116344174e-07,
+      "loss": 0.2172,
+      "step": 8487
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6631907456910964,
+      "learning_rate": 4.57502480047077e-07,
+      "loss": 0.264,
+      "step": 8488
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1861297116872045,
+      "learning_rate": 4.573995807437418e-07,
+      "loss": 0.0859,
+      "step": 8489
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.233467100768586,
+      "learning_rate": 4.57296683257826e-07,
+      "loss": 0.2024,
+      "step": 8490
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.562863703698288,
+      "learning_rate": 4.571937875937198e-07,
+      "loss": 0.1103,
+      "step": 8491
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3014767160948295,
+      "learning_rate": 4.570908937558124e-07,
+      "loss": 0.2501,
+      "step": 8492
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3948229887777503,
+      "learning_rate": 4.56988001748494e-07,
+      "loss": 0.275,
+      "step": 8493
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9564443585401297,
+      "learning_rate": 4.568851115761537e-07,
+      "loss": 0.368,
+      "step": 8494
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.29839740125089426,
+      "learning_rate": 4.5678222324318137e-07,
+      "loss": 0.1443,
+      "step": 8495
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6743428581848108,
+      "learning_rate": 4.5667933675396606e-07,
+      "loss": 0.2141,
+      "step": 8496
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7459111072513056,
+      "learning_rate": 4.5657645211289744e-07,
+      "loss": 0.1226,
+      "step": 8497
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7824845827003406,
+      "learning_rate": 4.564735693243647e-07,
+      "loss": 0.1287,
+      "step": 8498
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4834984827118423,
+      "learning_rate": 4.56370688392757e-07,
+      "loss": 0.1547,
+      "step": 8499
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0938361165736732,
+      "learning_rate": 4.562678093224633e-07,
+      "loss": 0.3256,
+      "step": 8500
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 29.55789695253735,
+      "learning_rate": 4.5616493211787297e-07,
+      "loss": 0.2399,
+      "step": 8501
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6838882822874129,
+      "learning_rate": 4.560620567833746e-07,
+      "loss": 0.076,
+      "step": 8502
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.24447589420637153,
+      "learning_rate": 4.5595918332335744e-07,
+      "loss": 0.069,
+      "step": 8503
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7864297919528537,
+      "learning_rate": 4.5585631174220994e-07,
+      "loss": 0.2658,
+      "step": 8504
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3257562561475966,
+      "learning_rate": 4.557534420443208e-07,
+      "loss": 0.0941,
+      "step": 8505
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9571794965312268,
+      "learning_rate": 4.556505742340789e-07,
+      "loss": 0.0663,
+      "step": 8506
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7530598028222795,
+      "learning_rate": 4.555477083158725e-07,
+      "loss": 0.3032,
+      "step": 8507
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.675894191460296,
+      "learning_rate": 4.554448442940904e-07,
+      "loss": 0.124,
+      "step": 8508
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5352661522361948,
+      "learning_rate": 4.5534198217312066e-07,
+      "loss": 0.2388,
+      "step": 8509
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1325929422909422,
+      "learning_rate": 4.552391219573519e-07,
+      "loss": 0.2196,
+      "step": 8510
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1595632651429615,
+      "learning_rate": 4.5513626365117196e-07,
+      "loss": 0.1096,
+      "step": 8511
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.935187183481488,
+      "learning_rate": 4.5503340725896946e-07,
+      "loss": 0.2327,
+      "step": 8512
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6270097308649183,
+      "learning_rate": 4.5493055278513194e-07,
+      "loss": 0.1127,
+      "step": 8513
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2333554916025773,
+      "learning_rate": 4.548277002340479e-07,
+      "loss": 0.0802,
+      "step": 8514
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3078384653044468,
+      "learning_rate": 4.547248496101047e-07,
+      "loss": 0.0836,
+      "step": 8515
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8529922203258207,
+      "learning_rate": 4.546220009176906e-07,
+      "loss": 0.2505,
+      "step": 8516
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5147821592586467,
+      "learning_rate": 4.5451915416119325e-07,
+      "loss": 0.1138,
+      "step": 8517
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7878131179164628,
+      "learning_rate": 4.54416309345e-07,
+      "loss": 0.0269,
+      "step": 8518
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.107587381449914,
+      "learning_rate": 4.5431346647349885e-07,
+      "loss": 0.0914,
+      "step": 8519
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3688615720946276,
+      "learning_rate": 4.542106255510768e-07,
+      "loss": 0.3,
+      "step": 8520
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9529514095880248,
+      "learning_rate": 4.541077865821218e-07,
+      "loss": 0.3298,
+      "step": 8521
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.51656489930236,
+      "learning_rate": 4.540049495710206e-07,
+      "loss": 0.3581,
+      "step": 8522
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0633316442621703,
+      "learning_rate": 4.5390211452216087e-07,
+      "loss": 0.2921,
+      "step": 8523
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6073632241863104,
+      "learning_rate": 4.537992814399296e-07,
+      "loss": 0.0691,
+      "step": 8524
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0659607195223204,
+      "learning_rate": 4.5369645032871377e-07,
+      "loss": 0.0703,
+      "step": 8525
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9100748578867685,
+      "learning_rate": 4.5359362119290046e-07,
+      "loss": 0.2182,
+      "step": 8526
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5451664741831638,
+      "learning_rate": 4.534907940368767e-07,
+      "loss": 0.1903,
+      "step": 8527
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2983460906523192,
+      "learning_rate": 4.5338796886502896e-07,
+      "loss": 0.2098,
+      "step": 8528
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9268620866095381,
+      "learning_rate": 4.5328514568174444e-07,
+      "loss": 0.1282,
+      "step": 8529
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6468324519500694,
+      "learning_rate": 4.531823244914094e-07,
+      "loss": 0.2481,
+      "step": 8530
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.875258197524188,
+      "learning_rate": 4.530795052984104e-07,
+      "loss": 0.1138,
+      "step": 8531
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6116136267177559,
+      "learning_rate": 4.529766881071341e-07,
+      "loss": 0.0849,
+      "step": 8532
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8706325022627655,
+      "learning_rate": 4.528738729219667e-07,
+      "loss": 0.3316,
+      "step": 8533
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.477585535082431,
+      "learning_rate": 4.5277105974729484e-07,
+      "loss": 0.1506,
+      "step": 8534
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.119045564193954,
+      "learning_rate": 4.526682485875043e-07,
+      "loss": 0.469,
+      "step": 8535
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.147142349762046,
+      "learning_rate": 4.525654394469816e-07,
+      "loss": 0.135,
+      "step": 8536
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8146832885122889,
+      "learning_rate": 4.5246263233011244e-07,
+      "loss": 0.1504,
+      "step": 8537
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4546010503383642,
+      "learning_rate": 4.523598272412832e-07,
+      "loss": 0.1102,
+      "step": 8538
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6218698456156957,
+      "learning_rate": 4.522570241848792e-07,
+      "loss": 0.1513,
+      "step": 8539
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5445552367923363,
+      "learning_rate": 4.521542231652868e-07,
+      "loss": 0.2466,
+      "step": 8540
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5778055370867833,
+      "learning_rate": 4.520514241868912e-07,
+      "loss": 0.1345,
+      "step": 8541
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7511853910031592,
+      "learning_rate": 4.519486272540783e-07,
+      "loss": 0.1383,
+      "step": 8542
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3388616325480788,
+      "learning_rate": 4.5184583237123356e-07,
+      "loss": 0.1855,
+      "step": 8543
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7383229291815947,
+      "learning_rate": 4.517430395427424e-07,
+      "loss": 0.2364,
+      "step": 8544
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8153944375237804,
+      "learning_rate": 4.5164024877299034e-07,
+      "loss": 0.2166,
+      "step": 8545
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2759299075411263,
+      "learning_rate": 4.515374600663623e-07,
+      "loss": 0.3461,
+      "step": 8546
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6106474779064373,
+      "learning_rate": 4.514346734272437e-07,
+      "loss": 0.2649,
+      "step": 8547
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8904849809586077,
+      "learning_rate": 4.5133188886001947e-07,
+      "loss": 0.2382,
+      "step": 8548
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6980647809570883,
+      "learning_rate": 4.512291063690749e-07,
+      "loss": 0.2302,
+      "step": 8549
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1361973622331778,
+      "learning_rate": 4.5112632595879444e-07,
+      "loss": 0.2255,
+      "step": 8550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.380638977705818,
+      "learning_rate": 4.510235476335633e-07,
+      "loss": 0.127,
+      "step": 8551
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.98530149233771,
+      "learning_rate": 4.5092077139776594e-07,
+      "loss": 0.1807,
+      "step": 8552
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8875602497822686,
+      "learning_rate": 4.508179972557874e-07,
+      "loss": 0.3445,
+      "step": 8553
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.0839262924190445,
+      "learning_rate": 4.507152252120117e-07,
+      "loss": 0.1327,
+      "step": 8554
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6907390655670341,
+      "learning_rate": 4.506124552708238e-07,
+      "loss": 0.1258,
+      "step": 8555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7974425763261175,
+      "learning_rate": 4.505096874366077e-07,
+      "loss": 0.0369,
+      "step": 8556
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.45529218864764853,
+      "learning_rate": 4.5040692171374794e-07,
+      "loss": 0.1797,
+      "step": 8557
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.794674730828558,
+      "learning_rate": 4.503041581066285e-07,
+      "loss": 0.1448,
+      "step": 8558
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3580680448302769,
+      "learning_rate": 4.502013966196336e-07,
+      "loss": 0.1913,
+      "step": 8559
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8710498541525413,
+      "learning_rate": 4.500986372571472e-07,
+      "loss": 0.3488,
+      "step": 8560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1796187287502056,
+      "learning_rate": 4.4999588002355314e-07,
+      "loss": 0.4305,
+      "step": 8561
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5076527385948689,
+      "learning_rate": 4.498931249232356e-07,
+      "loss": 0.1952,
+      "step": 8562
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1039826500375853,
+      "learning_rate": 4.4979037196057785e-07,
+      "loss": 0.2341,
+      "step": 8563
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3838415985195842,
+      "learning_rate": 4.49687621139964e-07,
+      "loss": 0.1882,
+      "step": 8564
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7075439609216445,
+      "learning_rate": 4.4958487246577714e-07,
+      "loss": 0.2253,
+      "step": 8565
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8971257055841362,
+      "learning_rate": 4.4948212594240113e-07,
+      "loss": 0.0992,
+      "step": 8566
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3838648711265849,
+      "learning_rate": 4.493793815742191e-07,
+      "loss": 0.0273,
+      "step": 8567
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6225953300300522,
+      "learning_rate": 4.492766393656144e-07,
+      "loss": 0.3109,
+      "step": 8568
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7097647397292058,
+      "learning_rate": 4.4917389932097027e-07,
+      "loss": 0.1853,
+      "step": 8569
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5155582573603247,
+      "learning_rate": 4.4907116144466976e-07,
+      "loss": 0.0226,
+      "step": 8570
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1948258985476277,
+      "learning_rate": 4.489684257410958e-07,
+      "loss": 0.0056,
+      "step": 8571
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28535405179376255,
+      "learning_rate": 4.4886569221463154e-07,
+      "loss": 0.1475,
+      "step": 8572
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.021488884601238,
+      "learning_rate": 4.487629608696597e-07,
+      "loss": 0.273,
+      "step": 8573
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9495205928322551,
+      "learning_rate": 4.4866023171056274e-07,
+      "loss": 0.1342,
+      "step": 8574
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4904094764027969,
+      "learning_rate": 4.485575047417237e-07,
+      "loss": 0.1998,
+      "step": 8575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.02893706203771,
+      "learning_rate": 4.484547799675247e-07,
+      "loss": 0.1373,
+      "step": 8576
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.26415931682304167,
+      "learning_rate": 4.483520573923485e-07,
+      "loss": 0.165,
+      "step": 8577
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.11699404437693192,
+      "learning_rate": 4.4824933702057727e-07,
+      "loss": 0.0034,
+      "step": 8578
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9777991146970927,
+      "learning_rate": 4.4814661885659353e-07,
+      "loss": 0.2992,
+      "step": 8579
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6497567332125923,
+      "learning_rate": 4.480439029047791e-07,
+      "loss": 0.3313,
+      "step": 8580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7355528914886902,
+      "learning_rate": 4.4794118916951644e-07,
+      "loss": 0.3799,
+      "step": 8581
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8520051342515318,
+      "learning_rate": 4.478384776551871e-07,
+      "loss": 0.1371,
+      "step": 8582
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2076017304051176,
+      "learning_rate": 4.477357683661733e-07,
+      "loss": 0.2908,
+      "step": 8583
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7971270527375104,
+      "learning_rate": 4.476330613068565e-07,
+      "loss": 0.1446,
+      "step": 8584
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8498952281979963,
+      "learning_rate": 4.4753035648161886e-07,
+      "loss": 0.3229,
+      "step": 8585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.084276383353148,
+      "learning_rate": 4.474276538948415e-07,
+      "loss": 0.128,
+      "step": 8586
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.779160114721411,
+      "learning_rate": 4.4732495355090597e-07,
+      "loss": 0.228,
+      "step": 8587
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4506411749383888,
+      "learning_rate": 4.4722225545419413e-07,
+      "loss": 0.0505,
+      "step": 8588
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1525618354432954,
+      "learning_rate": 4.471195596090867e-07,
+      "loss": 0.2949,
+      "step": 8589
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0802060735415184,
+      "learning_rate": 4.470168660199654e-07,
+      "loss": 0.1355,
+      "step": 8590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5992702188904998,
+      "learning_rate": 4.4691417469121083e-07,
+      "loss": 0.2294,
+      "step": 8591
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5559914605425803,
+      "learning_rate": 4.4681148562720455e-07,
+      "loss": 0.3393,
+      "step": 8592
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.390465442380409,
+      "learning_rate": 4.4670879883232696e-07,
+      "loss": 0.2481,
+      "step": 8593
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8133634865613211,
+      "learning_rate": 4.466061143109593e-07,
+      "loss": 0.174,
+      "step": 8594
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3427987143616134,
+      "learning_rate": 4.4650343206748215e-07,
+      "loss": 0.1598,
+      "step": 8595
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5185805257579341,
+      "learning_rate": 4.464007521062761e-07,
+      "loss": 0.0469,
+      "step": 8596
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6905742917442251,
+      "learning_rate": 4.462980744317216e-07,
+      "loss": 0.1262,
+      "step": 8597
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0559210717523206,
+      "learning_rate": 4.461953990481994e-07,
+      "loss": 0.2988,
+      "step": 8598
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0591882872068195,
+      "learning_rate": 4.460927259600897e-07,
+      "loss": 0.3202,
+      "step": 8599
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8848992748142672,
+      "learning_rate": 4.459900551717723e-07,
+      "loss": 0.4407,
+      "step": 8600
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.730913811847409,
+      "learning_rate": 4.4588738668762815e-07,
+      "loss": 0.1821,
+      "step": 8601
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.541553882484758,
+      "learning_rate": 4.4578472051203653e-07,
+      "loss": 0.2061,
+      "step": 8602
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8020017672337625,
+      "learning_rate": 4.4568205664937786e-07,
+      "loss": 0.3005,
+      "step": 8603
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2069163581798041,
+      "learning_rate": 4.455793951040318e-07,
+      "loss": 0.2515,
+      "step": 8604
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 13.947038719264075,
+      "learning_rate": 4.454767358803782e-07,
+      "loss": 0.1206,
+      "step": 8605
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2260616838684422,
+      "learning_rate": 4.453740789827965e-07,
+      "loss": 0.1138,
+      "step": 8606
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9046499230171707,
+      "learning_rate": 4.452714244156667e-07,
+      "loss": 0.097,
+      "step": 8607
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.356366811719842,
+      "learning_rate": 4.451687721833676e-07,
+      "loss": 0.2497,
+      "step": 8608
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8806676060896952,
+      "learning_rate": 4.450661222902792e-07,
+      "loss": 0.2198,
+      "step": 8609
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.340212205460144,
+      "learning_rate": 4.4496347474078027e-07,
+      "loss": 0.2238,
+      "step": 8610
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5907989412480348,
+      "learning_rate": 4.448608295392503e-07,
+      "loss": 0.3913,
+      "step": 8611
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.33591879836840965,
+      "learning_rate": 4.4475818669006806e-07,
+      "loss": 0.1539,
+      "step": 8612
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.481818672318781,
+      "learning_rate": 4.4465554619761263e-07,
+      "loss": 0.1277,
+      "step": 8613
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0167631925411578,
+      "learning_rate": 4.44552908066263e-07,
+      "loss": 0.1213,
+      "step": 8614
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6183022367609539,
+      "learning_rate": 4.4445027230039755e-07,
+      "loss": 0.1888,
+      "step": 8615
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6780775402626844,
+      "learning_rate": 4.4434763890439543e-07,
+      "loss": 0.2843,
+      "step": 8616
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 16.540078901841717,
+      "learning_rate": 4.442450078826346e-07,
+      "loss": 0.1059,
+      "step": 8617
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.022785206750274,
+      "learning_rate": 4.4414237923949407e-07,
+      "loss": 0.3227,
+      "step": 8618
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.721530692035682,
+      "learning_rate": 4.440397529793517e-07,
+      "loss": 0.0032,
+      "step": 8619
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7044104800798273,
+      "learning_rate": 4.439371291065862e-07,
+      "loss": 0.3409,
+      "step": 8620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5552772627363022,
+      "learning_rate": 4.438345076255753e-07,
+      "loss": 0.1829,
+      "step": 8621
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.17930898073871132,
+      "learning_rate": 4.4373188854069727e-07,
+      "loss": 0.0717,
+      "step": 8622
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5092543984358266,
+      "learning_rate": 4.436292718563299e-07,
+      "loss": 0.0758,
+      "step": 8623
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8913321612001055,
+      "learning_rate": 4.4352665757685125e-07,
+      "loss": 0.3083,
+      "step": 8624
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.119974980136262,
+      "learning_rate": 4.4342404570663876e-07,
+      "loss": 0.2503,
+      "step": 8625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.439870508423816,
+      "learning_rate": 4.4332143625007043e-07,
+      "loss": 0.1707,
+      "step": 8626
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.48170564663966,
+      "learning_rate": 4.432188292115236e-07,
+      "loss": 0.1542,
+      "step": 8627
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.449310572381116,
+      "learning_rate": 4.431162245953754e-07,
+      "loss": 0.0982,
+      "step": 8628
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2178805021997257,
+      "learning_rate": 4.430136224060035e-07,
+      "loss": 0.1398,
+      "step": 8629
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7279167315476547,
+      "learning_rate": 4.4291102264778506e-07,
+      "loss": 0.157,
+      "step": 8630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.093104599073745,
+      "learning_rate": 4.4280842532509716e-07,
+      "loss": 0.2506,
+      "step": 8631
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7498838727386716,
+      "learning_rate": 4.4270583044231667e-07,
+      "loss": 0.1681,
+      "step": 8632
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.916312673192367,
+      "learning_rate": 4.4260323800382086e-07,
+      "loss": 0.0876,
+      "step": 8633
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.73006049574467,
+      "learning_rate": 4.4250064801398607e-07,
+      "loss": 0.0755,
+      "step": 8634
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.47755745843624,
+      "learning_rate": 4.4239806047718945e-07,
+      "loss": 0.1912,
+      "step": 8635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3343599388855688,
+      "learning_rate": 4.4229547539780715e-07,
+      "loss": 0.1178,
+      "step": 8636
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.40355819689314826,
+      "learning_rate": 4.421928927802161e-07,
+      "loss": 0.1157,
+      "step": 8637
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.17095910480109017,
+      "learning_rate": 4.420903126287922e-07,
+      "loss": 0.0202,
+      "step": 8638
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5340901959281243,
+      "learning_rate": 4.419877349479121e-07,
+      "loss": 0.3271,
+      "step": 8639
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0797816216778933,
+      "learning_rate": 4.41885159741952e-07,
+      "loss": 0.2782,
+      "step": 8640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5183601064212755,
+      "learning_rate": 4.4178258701528756e-07,
+      "loss": 0.1911,
+      "step": 8641
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.65818756265992,
+      "learning_rate": 4.4168001677229523e-07,
+      "loss": 0.1366,
+      "step": 8642
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9687773782908738,
+      "learning_rate": 4.4157744901735033e-07,
+      "loss": 0.1931,
+      "step": 8643
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6543510595751106,
+      "learning_rate": 4.4147488375482915e-07,
+      "loss": 0.1133,
+      "step": 8644
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1118765521249256,
+      "learning_rate": 4.4137232098910693e-07,
+      "loss": 0.3299,
+      "step": 8645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24685852565012192,
+      "learning_rate": 4.4126976072455946e-07,
+      "loss": 0.1877,
+      "step": 8646
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2573644076145768,
+      "learning_rate": 4.411672029655619e-07,
+      "loss": 0.1783,
+      "step": 8647
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.250444386789134,
+      "learning_rate": 4.410646477164899e-07,
+      "loss": 0.3069,
+      "step": 8648
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1789709245803754,
+      "learning_rate": 4.4096209498171825e-07,
+      "loss": 0.4047,
+      "step": 8649
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1773938457912942,
+      "learning_rate": 4.4085954476562263e-07,
+      "loss": 0.3411,
+      "step": 8650
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.609434498904355,
+      "learning_rate": 4.4075699707257754e-07,
+      "loss": 0.2038,
+      "step": 8651
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7033537492918164,
+      "learning_rate": 4.4065445190695816e-07,
+      "loss": 0.2254,
+      "step": 8652
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.534189575717024,
+      "learning_rate": 4.40551909273139e-07,
+      "loss": 0.36,
+      "step": 8653
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.42367573297646766,
+      "learning_rate": 4.4044936917549505e-07,
+      "loss": 0.1442,
+      "step": 8654
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1640803618953181,
+      "learning_rate": 4.403468316184007e-07,
+      "loss": 0.0206,
+      "step": 8655
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.35458751937588334,
+      "learning_rate": 4.402442966062304e-07,
+      "loss": 0.0076,
+      "step": 8656
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5175948670631023,
+      "learning_rate": 4.4014176414335854e-07,
+      "loss": 0.1836,
+      "step": 8657
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6565120566952919,
+      "learning_rate": 4.4003923423415923e-07,
+      "loss": 0.1064,
+      "step": 8658
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5703494333846566,
+      "learning_rate": 4.3993670688300694e-07,
+      "loss": 0.3563,
+      "step": 8659
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.451116982674173,
+      "learning_rate": 4.3983418209427526e-07,
+      "loss": 0.1283,
+      "step": 8660
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7066977888382018,
+      "learning_rate": 4.397316598723385e-07,
+      "loss": 0.1661,
+      "step": 8661
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.861787167085259,
+      "learning_rate": 4.3962914022157014e-07,
+      "loss": 0.222,
+      "step": 8662
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5970506627413962,
+      "learning_rate": 4.395266231463442e-07,
+      "loss": 0.2229,
+      "step": 8663
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1996669243248315,
+      "learning_rate": 4.3942410865103384e-07,
+      "loss": 0.1183,
+      "step": 8664
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6991684169332326,
+      "learning_rate": 4.3932159674001305e-07,
+      "loss": 0.0926,
+      "step": 8665
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8945252113205767,
+      "learning_rate": 4.392190874176547e-07,
+      "loss": 0.1876,
+      "step": 8666
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.472981795237423,
+      "learning_rate": 4.3911658068833234e-07,
+      "loss": 0.2331,
+      "step": 8667
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5292525294377366,
+      "learning_rate": 4.3901407655641915e-07,
+      "loss": 0.3686,
+      "step": 8668
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6746245898447345,
+      "learning_rate": 4.389115750262878e-07,
+      "loss": 0.0777,
+      "step": 8669
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9496036799051835,
+      "learning_rate": 4.3880907610231177e-07,
+      "loss": 0.0932,
+      "step": 8670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1429990801985839,
+      "learning_rate": 4.387065797888633e-07,
+      "loss": 0.3681,
+      "step": 8671
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5704570566856747,
+      "learning_rate": 4.386040860903156e-07,
+      "loss": 0.2151,
+      "step": 8672
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 40.21932685715568,
+      "learning_rate": 4.3850159501104073e-07,
+      "loss": 0.2563,
+      "step": 8673
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.817384533150433,
+      "learning_rate": 4.3839910655541157e-07,
+      "loss": 0.0577,
+      "step": 8674
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.46772570645501416,
+      "learning_rate": 4.382966207278004e-07,
+      "loss": 0.1736,
+      "step": 8675
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24529039893710539,
+      "learning_rate": 4.381941375325794e-07,
+      "loss": 0.1944,
+      "step": 8676
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8053000301213352,
+      "learning_rate": 4.380916569741206e-07,
+      "loss": 0.2553,
+      "step": 8677
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2459487746508058,
+      "learning_rate": 4.379891790567964e-07,
+      "loss": 0.2528,
+      "step": 8678
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5786673703935769,
+      "learning_rate": 4.3788670378497827e-07,
+      "loss": 0.0103,
+      "step": 8679
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6131249040544321,
+      "learning_rate": 4.377842311630385e-07,
+      "loss": 0.1538,
+      "step": 8680
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.43975674968747863,
+      "learning_rate": 4.3768176119534816e-07,
+      "loss": 0.216,
+      "step": 8681
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9238939584516679,
+      "learning_rate": 4.3757929388627943e-07,
+      "loss": 0.1618,
+      "step": 8682
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5606365831651372,
+      "learning_rate": 4.3747682924020346e-07,
+      "loss": 0.275,
+      "step": 8683
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2871099484383887,
+      "learning_rate": 4.373743672614915e-07,
+      "loss": 0.2205,
+      "step": 8684
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2847660107718668,
+      "learning_rate": 4.3727190795451506e-07,
+      "loss": 0.2713,
+      "step": 8685
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6266697475905256,
+      "learning_rate": 4.3716945132364495e-07,
+      "loss": 0.0137,
+      "step": 8686
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.09043999399511815,
+      "learning_rate": 4.3706699737325263e-07,
+      "loss": 0.0024,
+      "step": 8687
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.773539277771456,
+      "learning_rate": 4.3696454610770843e-07,
+      "loss": 0.1742,
+      "step": 8688
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3623365393024069,
+      "learning_rate": 4.368620975313837e-07,
+      "loss": 0.0477,
+      "step": 8689
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7092943270773006,
+      "learning_rate": 4.367596516486485e-07,
+      "loss": 0.2816,
+      "step": 8690
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.42632053701819134,
+      "learning_rate": 4.3665720846387397e-07,
+      "loss": 0.0572,
+      "step": 8691
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2135248255203528,
+      "learning_rate": 4.3655476798143003e-07,
+      "loss": 0.419,
+      "step": 8692
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8067930471012168,
+      "learning_rate": 4.3645233020568734e-07,
+      "loss": 0.1214,
+      "step": 8693
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5990983382871302,
+      "learning_rate": 4.363498951410159e-07,
+      "loss": 0.1371,
+      "step": 8694
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8936630579336159,
+      "learning_rate": 4.3624746279178603e-07,
+      "loss": 0.1692,
+      "step": 8695
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5441647656795372,
+      "learning_rate": 4.361450331623676e-07,
+      "loss": 0.2733,
+      "step": 8696
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6118546178471441,
+      "learning_rate": 4.3604260625713025e-07,
+      "loss": 0.3282,
+      "step": 8697
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.39873167931542874,
+      "learning_rate": 4.359401820804441e-07,
+      "loss": 0.3149,
+      "step": 8698
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.37182408010870244,
+      "learning_rate": 4.358377606366783e-07,
+      "loss": 0.0987,
+      "step": 8699
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.543440377640531,
+      "learning_rate": 4.3573534193020274e-07,
+      "loss": 0.0189,
+      "step": 8700
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7683463471750335,
+      "learning_rate": 4.3563292596538665e-07,
+      "loss": 0.1435,
+      "step": 8701
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.063278332037879,
+      "learning_rate": 4.3553051274659934e-07,
+      "loss": 0.1295,
+      "step": 8702
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.254262210336867,
+      "learning_rate": 4.3542810227820985e-07,
+      "loss": 0.3357,
+      "step": 8703
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.481368459596217,
+      "learning_rate": 4.3532569456458755e-07,
+      "loss": 0.3109,
+      "step": 8704
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2236216102368147,
+      "learning_rate": 4.352232896101009e-07,
+      "loss": 0.1365,
+      "step": 8705
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 14.940685031864735,
+      "learning_rate": 4.351208874191192e-07,
+      "loss": 0.2594,
+      "step": 8706
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5902382366757902,
+      "learning_rate": 4.350184879960106e-07,
+      "loss": 0.0927,
+      "step": 8707
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.597275120931913,
+      "learning_rate": 4.3491609134514416e-07,
+      "loss": 0.0874,
+      "step": 8708
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5006277786000783,
+      "learning_rate": 4.3481369747088804e-07,
+      "loss": 0.2512,
+      "step": 8709
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.914182112445155,
+      "learning_rate": 4.3471130637761044e-07,
+      "loss": 0.0853,
+      "step": 8710
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2741050102783117,
+      "learning_rate": 4.3460891806968004e-07,
+      "loss": 0.2112,
+      "step": 8711
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6388195877724325,
+      "learning_rate": 4.3450653255146436e-07,
+      "loss": 0.2596,
+      "step": 8712
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3621628629849335,
+      "learning_rate": 4.344041498273319e-07,
+      "loss": 0.069,
+      "step": 8713
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.377470267897786,
+      "learning_rate": 4.343017699016501e-07,
+      "loss": 0.1486,
+      "step": 8714
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.27535344211405466,
+      "learning_rate": 4.3419939277878705e-07,
+      "loss": 0.0665,
+      "step": 8715
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.40089579548440857,
+      "learning_rate": 4.340970184631099e-07,
+      "loss": 0.3926,
+      "step": 8716
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6938331545369534,
+      "learning_rate": 4.339946469589867e-07,
+      "loss": 0.1051,
+      "step": 8717
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.462107263236189,
+      "learning_rate": 4.3389227827078434e-07,
+      "loss": 0.1912,
+      "step": 8718
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.242599660882346,
+      "learning_rate": 4.3378991240287044e-07,
+      "loss": 0.2046,
+      "step": 8719
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.803931785433432,
+      "learning_rate": 4.336875493596119e-07,
+      "loss": 0.3035,
+      "step": 8720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.34555717096346095,
+      "learning_rate": 4.3358518914537587e-07,
+      "loss": 0.0487,
+      "step": 8721
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2831189267608031,
+      "learning_rate": 4.33482831764529e-07,
+      "loss": 0.1161,
+      "step": 8722
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.785665893365899,
+      "learning_rate": 4.333804772214385e-07,
+      "loss": 0.2001,
+      "step": 8723
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1015902546268215,
+      "learning_rate": 4.332781255204707e-07,
+      "loss": 0.0564,
+      "step": 8724
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.897124187273214,
+      "learning_rate": 4.331757766659921e-07,
+      "loss": 0.2349,
+      "step": 8725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7622424133876797,
+      "learning_rate": 4.3307343066236934e-07,
+      "loss": 0.0969,
+      "step": 8726
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0066861931099265,
+      "learning_rate": 4.329710875139684e-07,
+      "loss": 0.2512,
+      "step": 8727
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.171090718565011,
+      "learning_rate": 4.3286874722515564e-07,
+      "loss": 0.2247,
+      "step": 8728
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2886861799314343,
+      "learning_rate": 4.32766409800297e-07,
+      "loss": 0.1011,
+      "step": 8729
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7881433383426815,
+      "learning_rate": 4.326640752437587e-07,
+      "loss": 0.156,
+      "step": 8730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.36662181002272237,
+      "learning_rate": 4.325617435599061e-07,
+      "loss": 0.1765,
+      "step": 8731
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8300046605081782,
+      "learning_rate": 4.3245941475310535e-07,
+      "loss": 0.1485,
+      "step": 8732
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.08197081239682,
+      "learning_rate": 4.323570888277214e-07,
+      "loss": 0.1513,
+      "step": 8733
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6685893549406288,
+      "learning_rate": 4.322547657881203e-07,
+      "loss": 0.1808,
+      "step": 8734
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7791728144281073,
+      "learning_rate": 4.321524456386669e-07,
+      "loss": 0.2814,
+      "step": 8735
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5645565736405153,
+      "learning_rate": 4.320501283837267e-07,
+      "loss": 0.2956,
+      "step": 8736
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.139665768483349,
+      "learning_rate": 4.3194781402766464e-07,
+      "loss": 0.0679,
+      "step": 8737
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7585261044644834,
+      "learning_rate": 4.318455025748454e-07,
+      "loss": 0.4393,
+      "step": 8738
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6420015229006227,
+      "learning_rate": 4.3174319402963426e-07,
+      "loss": 0.3076,
+      "step": 8739
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.053093193924328,
+      "learning_rate": 4.316408883963955e-07,
+      "loss": 0.0895,
+      "step": 8740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1766155095962807,
+      "learning_rate": 4.315385856794941e-07,
+      "loss": 0.3237,
+      "step": 8741
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.28378767122573545,
+      "learning_rate": 4.31436285883294e-07,
+      "loss": 0.0871,
+      "step": 8742
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8100632778598453,
+      "learning_rate": 4.3133398901216e-07,
+      "loss": 0.0596,
+      "step": 8743
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.917157902373197,
+      "learning_rate": 4.3123169507045587e-07,
+      "loss": 0.2018,
+      "step": 8744
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9164757278327148,
+      "learning_rate": 4.3112940406254595e-07,
+      "loss": 0.1257,
+      "step": 8745
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3863658506835133,
+      "learning_rate": 4.3102711599279406e-07,
+      "loss": 0.1335,
+      "step": 8746
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5603958703346787,
+      "learning_rate": 4.309248308655641e-07,
+      "loss": 0.2087,
+      "step": 8747
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8364777548053638,
+      "learning_rate": 4.308225486852196e-07,
+      "loss": 0.1572,
+      "step": 8748
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6659128864748425,
+      "learning_rate": 4.307202694561245e-07,
+      "loss": 0.2157,
+      "step": 8749
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4506071015548738,
+      "learning_rate": 4.306179931826419e-07,
+      "loss": 0.2602,
+      "step": 8750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.109939051352687,
+      "learning_rate": 4.3051571986913506e-07,
+      "loss": 0.316,
+      "step": 8751
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.43624671986174895,
+      "learning_rate": 4.304134495199674e-07,
+      "loss": 0.0085,
+      "step": 8752
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.181416783377148,
+      "learning_rate": 4.3031118213950166e-07,
+      "loss": 0.1573,
+      "step": 8753
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 19.120870356589716,
+      "learning_rate": 4.302089177321012e-07,
+      "loss": 0.2051,
+      "step": 8754
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8544548483466076,
+      "learning_rate": 4.3010665630212836e-07,
+      "loss": 0.1524,
+      "step": 8755
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7709835200098785,
+      "learning_rate": 4.3000439785394646e-07,
+      "loss": 0.2253,
+      "step": 8756
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.236281598019691,
+      "learning_rate": 4.299021423919173e-07,
+      "loss": 0.2094,
+      "step": 8757
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2272571124262694,
+      "learning_rate": 4.297998899204039e-07,
+      "loss": 0.0809,
+      "step": 8758
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7594296087296805,
+      "learning_rate": 4.296976404437681e-07,
+      "loss": 0.1135,
+      "step": 8759
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0453715609503011,
+      "learning_rate": 4.295953939663726e-07,
+      "loss": 0.1734,
+      "step": 8760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8404093890406891,
+      "learning_rate": 4.2949315049257873e-07,
+      "loss": 0.3088,
+      "step": 8761
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3165393104660033,
+      "learning_rate": 4.293909100267491e-07,
+      "loss": 0.2291,
+      "step": 8762
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8868526873780667,
+      "learning_rate": 4.2928867257324495e-07,
+      "loss": 0.0805,
+      "step": 8763
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5407469980793979,
+      "learning_rate": 4.2918643813642836e-07,
+      "loss": 0.2776,
+      "step": 8764
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0047903624271113,
+      "learning_rate": 4.2908420672066075e-07,
+      "loss": 0.2345,
+      "step": 8765
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4653994504140275,
+      "learning_rate": 4.289819783303032e-07,
+      "loss": 0.1555,
+      "step": 8766
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.46737109688055,
+      "learning_rate": 4.288797529697175e-07,
+      "loss": 0.3302,
+      "step": 8767
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6238645780029942,
+      "learning_rate": 4.287775306432642e-07,
+      "loss": 0.1174,
+      "step": 8768
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8079535255908654,
+      "learning_rate": 4.2867531135530484e-07,
+      "loss": 0.3148,
+      "step": 8769
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.43723374040357554,
+      "learning_rate": 4.285730951101999e-07,
+      "loss": 0.1074,
+      "step": 8770
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.38881263524925763,
+      "learning_rate": 4.284708819123104e-07,
+      "loss": 0.01,
+      "step": 8771
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1943730728812147,
+      "learning_rate": 4.2836867176599684e-07,
+      "loss": 0.0127,
+      "step": 8772
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.48869727062370727,
+      "learning_rate": 4.282664646756198e-07,
+      "loss": 0.1828,
+      "step": 8773
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0197104140215727,
+      "learning_rate": 4.2816426064553946e-07,
+      "loss": 0.1231,
+      "step": 8774
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5393575259996369,
+      "learning_rate": 4.280620596801163e-07,
+      "loss": 0.0497,
+      "step": 8775
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.05371980771129873,
+      "learning_rate": 4.279598617837102e-07,
+      "loss": 0.0002,
+      "step": 8776
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5745978682807066,
+      "learning_rate": 4.278576669606814e-07,
+      "loss": 0.1843,
+      "step": 8777
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9278652412815943,
+      "learning_rate": 4.277554752153895e-07,
+      "loss": 0.0959,
+      "step": 8778
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5250453545431085,
+      "learning_rate": 4.27653286552194e-07,
+      "loss": 0.2151,
+      "step": 8779
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9632878342014657,
+      "learning_rate": 4.2755110097545495e-07,
+      "loss": 0.023,
+      "step": 8780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6455365855607985,
+      "learning_rate": 4.274489184895315e-07,
+      "loss": 0.1282,
+      "step": 8781
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.19205089455886,
+      "learning_rate": 4.273467390987831e-07,
+      "loss": 0.0743,
+      "step": 8782
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.854499994034134,
+      "learning_rate": 4.272445628075687e-07,
+      "loss": 0.216,
+      "step": 8783
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6115082054764824,
+      "learning_rate": 4.2714238962024775e-07,
+      "loss": 0.1696,
+      "step": 8784
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2685461559803914,
+      "learning_rate": 4.270402195411787e-07,
+      "loss": 0.1875,
+      "step": 8785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8428912491270012,
+      "learning_rate": 4.269380525747207e-07,
+      "loss": 0.1702,
+      "step": 8786
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4900605117836085,
+      "learning_rate": 4.2683588872523214e-07,
+      "loss": 0.1229,
+      "step": 8787
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9344322384855469,
+      "learning_rate": 4.267337279970718e-07,
+      "loss": 0.2983,
+      "step": 8788
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3308066950851476,
+      "learning_rate": 4.2663157039459774e-07,
+      "loss": 0.1737,
+      "step": 8789
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4536736858630486,
+      "learning_rate": 4.265294159221684e-07,
+      "loss": 0.1757,
+      "step": 8790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2401233737061514,
+      "learning_rate": 4.264272645841419e-07,
+      "loss": 0.2111,
+      "step": 8791
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7235916896888939,
+      "learning_rate": 4.263251163848762e-07,
+      "loss": 0.3137,
+      "step": 8792
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8752625380095372,
+      "learning_rate": 4.262229713287293e-07,
+      "loss": 0.5314,
+      "step": 8793
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.771569810967823,
+      "learning_rate": 4.261208294200585e-07,
+      "loss": 0.196,
+      "step": 8794
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9800783063616554,
+      "learning_rate": 4.260186906632219e-07,
+      "loss": 0.2956,
+      "step": 8795
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9444654094315114,
+      "learning_rate": 4.259165550625764e-07,
+      "loss": 0.2544,
+      "step": 8796
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5934749999704413,
+      "learning_rate": 4.258144226224798e-07,
+      "loss": 0.1408,
+      "step": 8797
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3570317708605626,
+      "learning_rate": 4.257122933472889e-07,
+      "loss": 0.0839,
+      "step": 8798
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2242800026870266,
+      "learning_rate": 4.256101672413609e-07,
+      "loss": 0.1207,
+      "step": 8799
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.241545716452189,
+      "learning_rate": 4.255080443090527e-07,
+      "loss": 0.4852,
+      "step": 8800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.35390348831937746,
+      "learning_rate": 4.254059245547212e-07,
+      "loss": 0.1465,
+      "step": 8801
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3990140449314784,
+      "learning_rate": 4.2530380798272283e-07,
+      "loss": 0.0983,
+      "step": 8802
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4236219228074876,
+      "learning_rate": 4.2520169459741427e-07,
+      "loss": 0.1796,
+      "step": 8803
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.57621134386628,
+      "learning_rate": 4.250995844031516e-07,
+      "loss": 0.1206,
+      "step": 8804
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0245743971573507,
+      "learning_rate": 4.249974774042915e-07,
+      "loss": 0.3929,
+      "step": 8805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9511849609887367,
+      "learning_rate": 4.2489537360518963e-07,
+      "loss": 0.1818,
+      "step": 8806
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.36073045919090885,
+      "learning_rate": 4.2479327301020214e-07,
+      "loss": 0.275,
+      "step": 8807
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7016519427161496,
+      "learning_rate": 4.246911756236847e-07,
+      "loss": 0.1796,
+      "step": 8808
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.261643534370843,
+      "learning_rate": 4.245890814499931e-07,
+      "loss": 0.2649,
+      "step": 8809
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.17211187922867002,
+      "learning_rate": 4.244869904934831e-07,
+      "loss": 0.0906,
+      "step": 8810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0630988415737792,
+      "learning_rate": 4.243849027585096e-07,
+      "loss": 0.0131,
+      "step": 8811
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9457028326375878,
+      "learning_rate": 4.2428281824942843e-07,
+      "loss": 0.4614,
+      "step": 8812
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3354028111491003,
+      "learning_rate": 4.2418073697059414e-07,
+      "loss": 0.1195,
+      "step": 8813
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.486547982467088,
+      "learning_rate": 4.240786589263623e-07,
+      "loss": 0.3201,
+      "step": 8814
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.608709704000889,
+      "learning_rate": 4.239765841210873e-07,
+      "loss": 0.0223,
+      "step": 8815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8671122060277289,
+      "learning_rate": 4.2387451255912407e-07,
+      "loss": 0.2541,
+      "step": 8816
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.11364371117671,
+      "learning_rate": 4.2377244424482726e-07,
+      "loss": 0.2379,
+      "step": 8817
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5695881297848178,
+      "learning_rate": 4.236703791825512e-07,
+      "loss": 0.0693,
+      "step": 8818
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2860772628905273,
+      "learning_rate": 4.235683173766503e-07,
+      "loss": 0.148,
+      "step": 8819
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9321914860380835,
+      "learning_rate": 4.234662588314784e-07,
+      "loss": 0.2268,
+      "step": 8820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8848982594913803,
+      "learning_rate": 4.2336420355139e-07,
+      "loss": 0.0244,
+      "step": 8821
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5598738834934773,
+      "learning_rate": 4.232621515407385e-07,
+      "loss": 0.2702,
+      "step": 8822
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4768426556401251,
+      "learning_rate": 4.2316010280387803e-07,
+      "loss": 0.1238,
+      "step": 8823
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5368905908160633,
+      "learning_rate": 4.230580573451619e-07,
+      "loss": 0.2323,
+      "step": 8824
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4680823082719106,
+      "learning_rate": 4.2295601516894387e-07,
+      "loss": 0.2658,
+      "step": 8825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30093073521511227,
+      "learning_rate": 4.2285397627957686e-07,
+      "loss": 0.0098,
+      "step": 8826
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6365285501257308,
+      "learning_rate": 4.227519406814146e-07,
+      "loss": 0.0815,
+      "step": 8827
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6121023613489245,
+      "learning_rate": 4.226499083788096e-07,
+      "loss": 0.3337,
+      "step": 8828
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7200664274228797,
+      "learning_rate": 4.225478793761153e-07,
+      "loss": 0.1372,
+      "step": 8829
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.995027730163176,
+      "learning_rate": 4.2244585367768387e-07,
+      "loss": 0.1968,
+      "step": 8830
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1568868680182585,
+      "learning_rate": 4.2234383128786847e-07,
+      "loss": 0.2059,
+      "step": 8831
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.065654123080128,
+      "learning_rate": 4.222418122110211e-07,
+      "loss": 0.0939,
+      "step": 8832
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.760090562541034,
+      "learning_rate": 4.2213979645149455e-07,
+      "loss": 0.304,
+      "step": 8833
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1222549533889943,
+      "learning_rate": 4.220377840136407e-07,
+      "loss": 0.2764,
+      "step": 8834
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2180117061610305,
+      "learning_rate": 4.2193577490181163e-07,
+      "loss": 0.1748,
+      "step": 8835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4405662940994313,
+      "learning_rate": 4.218337691203595e-07,
+      "loss": 0.1627,
+      "step": 8836
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9280427079148141,
+      "learning_rate": 4.217317666736357e-07,
+      "loss": 0.0307,
+      "step": 8837
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.040666841572136,
+      "learning_rate": 4.2162976756599234e-07,
+      "loss": 0.0824,
+      "step": 8838
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.408029469422152,
+      "learning_rate": 4.215277718017804e-07,
+      "loss": 0.3568,
+      "step": 8839
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.850396461252373,
+      "learning_rate": 4.214257793853517e-07,
+      "loss": 0.3457,
+      "step": 8840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.130899098865247,
+      "learning_rate": 4.2132379032105693e-07,
+      "loss": 0.0031,
+      "step": 8841
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9850970598885355,
+      "learning_rate": 4.212218046132476e-07,
+      "loss": 0.2646,
+      "step": 8842
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7836241026560231,
+      "learning_rate": 4.2111982226627433e-07,
+      "loss": 0.0229,
+      "step": 8843
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.300270262942005,
+      "learning_rate": 4.2101784328448814e-07,
+      "loss": 0.1831,
+      "step": 8844
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1229748622532114,
+      "learning_rate": 4.2091586767223936e-07,
+      "loss": 0.1617,
+      "step": 8845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6543210638658481,
+      "learning_rate": 4.2081389543387893e-07,
+      "loss": 0.1861,
+      "step": 8846
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.127326640082101,
+      "learning_rate": 4.2071192657375684e-07,
+      "loss": 0.056,
+      "step": 8847
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.479982180174863,
+      "learning_rate": 4.206099610962231e-07,
+      "loss": 0.2553,
+      "step": 8848
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.47943624358487036,
+      "learning_rate": 4.2050799900562826e-07,
+      "loss": 0.3061,
+      "step": 8849
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1833321862723516,
+      "learning_rate": 4.2040604030632176e-07,
+      "loss": 0.1738,
+      "step": 8850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2990509725056782,
+      "learning_rate": 4.203040850026537e-07,
+      "loss": 0.0038,
+      "step": 8851
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6951384879437232,
+      "learning_rate": 4.202021330989736e-07,
+      "loss": 0.1756,
+      "step": 8852
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8608802564102809,
+      "learning_rate": 4.201001845996309e-07,
+      "loss": 0.1009,
+      "step": 8853
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.575110980996379,
+      "learning_rate": 4.199982395089749e-07,
+      "loss": 0.1647,
+      "step": 8854
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.192043742401415,
+      "learning_rate": 4.19896297831355e-07,
+      "loss": 0.1525,
+      "step": 8855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5647354508329827,
+      "learning_rate": 4.197943595711198e-07,
+      "loss": 0.0924,
+      "step": 8856
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.690594689503533,
+      "learning_rate": 4.1969242473261873e-07,
+      "loss": 0.1563,
+      "step": 8857
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 15.318151122942464,
+      "learning_rate": 4.195904933202e-07,
+      "loss": 0.1284,
+      "step": 8858
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4100718526531126,
+      "learning_rate": 4.194885653382127e-07,
+      "loss": 0.1654,
+      "step": 8859
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6682643198475973,
+      "learning_rate": 4.1938664079100493e-07,
+      "loss": 0.1893,
+      "step": 8860
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.702906979463967,
+      "learning_rate": 4.192847196829251e-07,
+      "loss": 0.3181,
+      "step": 8861
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2794038402298713,
+      "learning_rate": 4.1918280201832145e-07,
+      "loss": 0.0579,
+      "step": 8862
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.43971315297669045,
+      "learning_rate": 4.1908088780154183e-07,
+      "loss": 0.1261,
+      "step": 8863
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.44545082915315964,
+      "learning_rate": 4.1897897703693435e-07,
+      "loss": 0.1466,
+      "step": 8864
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2037949421885682,
+      "learning_rate": 4.188770697288464e-07,
+      "loss": 0.1159,
+      "step": 8865
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2637857986268712,
+      "learning_rate": 4.1877516588162596e-07,
+      "loss": 0.4962,
+      "step": 8866
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8129489178667497,
+      "learning_rate": 4.1867326549962e-07,
+      "loss": 0.2276,
+      "step": 8867
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6055254303719817,
+      "learning_rate": 4.185713685871762e-07,
+      "loss": 0.2519,
+      "step": 8868
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7323659725268498,
+      "learning_rate": 4.184694751486414e-07,
+      "loss": 0.1028,
+      "step": 8869
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.684166843548358,
+      "learning_rate": 4.183675851883627e-07,
+      "loss": 0.2773,
+      "step": 8870
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1639504907915714,
+      "learning_rate": 4.1826569871068686e-07,
+      "loss": 0.0061,
+      "step": 8871
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8127810049258692,
+      "learning_rate": 4.1816381571996075e-07,
+      "loss": 0.1356,
+      "step": 8872
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9025274955719915,
+      "learning_rate": 4.180619362205306e-07,
+      "loss": 0.17,
+      "step": 8873
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8218113799740292,
+      "learning_rate": 4.1796006021674326e-07,
+      "loss": 0.2427,
+      "step": 8874
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.22969596510198548,
+      "learning_rate": 4.1785818771294456e-07,
+      "loss": 0.0627,
+      "step": 8875
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.013127341362521,
+      "learning_rate": 4.1775631871348056e-07,
+      "loss": 0.2197,
+      "step": 8876
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6531218599809984,
+      "learning_rate": 4.176544532226974e-07,
+      "loss": 0.1165,
+      "step": 8877
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.6029073170787513,
+      "learning_rate": 4.175525912449408e-07,
+      "loss": 0.1084,
+      "step": 8878
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9247928800834706,
+      "learning_rate": 4.174507327845564e-07,
+      "loss": 0.0345,
+      "step": 8879
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4511928593583288,
+      "learning_rate": 4.173488778458896e-07,
+      "loss": 0.3031,
+      "step": 8880
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.758286084876253,
+      "learning_rate": 4.17247026433286e-07,
+      "loss": 0.0033,
+      "step": 8881
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6638340939757755,
+      "learning_rate": 4.171451785510904e-07,
+      "loss": 0.419,
+      "step": 8882
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8397141346873115,
+      "learning_rate": 4.1704333420364827e-07,
+      "loss": 0.2656,
+      "step": 8883
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.48375765804415893,
+      "learning_rate": 4.16941493395304e-07,
+      "loss": 0.0618,
+      "step": 8884
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4125366962194494,
+      "learning_rate": 4.1683965613040293e-07,
+      "loss": 0.3128,
+      "step": 8885
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7514592078671956,
+      "learning_rate": 4.16737822413289e-07,
+      "loss": 0.2279,
+      "step": 8886
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4256378297724906,
+      "learning_rate": 4.166359922483071e-07,
+      "loss": 0.2878,
+      "step": 8887
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6441216191541835,
+      "learning_rate": 4.165341656398014e-07,
+      "loss": 0.1651,
+      "step": 8888
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2115471476169245,
+      "learning_rate": 4.164323425921158e-07,
+      "loss": 0.1172,
+      "step": 8889
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9850882113832111,
+      "learning_rate": 4.163305231095947e-07,
+      "loss": 0.003,
+      "step": 8890
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8093672278466705,
+      "learning_rate": 4.1622870719658145e-07,
+      "loss": 0.2152,
+      "step": 8891
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6736874441809302,
+      "learning_rate": 4.1612689485742013e-07,
+      "loss": 0.1791,
+      "step": 8892
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.01785573323617,
+      "learning_rate": 4.160250860964539e-07,
+      "loss": 0.1442,
+      "step": 8893
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8646779079782378,
+      "learning_rate": 4.1592328091802645e-07,
+      "loss": 0.2035,
+      "step": 8894
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3145613306400818,
+      "learning_rate": 4.158214793264807e-07,
+      "loss": 0.0532,
+      "step": 8895
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1389023483351055,
+      "learning_rate": 4.1571968132615996e-07,
+      "loss": 0.2556,
+      "step": 8896
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7143415719349784,
+      "learning_rate": 4.1561788692140707e-07,
+      "loss": 0.148,
+      "step": 8897
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.090866347674417,
+      "learning_rate": 4.1551609611656473e-07,
+      "loss": 0.1279,
+      "step": 8898
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6558849385962897,
+      "learning_rate": 4.1541430891597544e-07,
+      "loss": 0.3018,
+      "step": 8899
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.387434021177685,
+      "learning_rate": 4.15312525323982e-07,
+      "loss": 0.3073,
+      "step": 8900
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7054601023014728,
+      "learning_rate": 4.152107453449263e-07,
+      "loss": 0.0077,
+      "step": 8901
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5885797868645957,
+      "learning_rate": 4.151089689831508e-07,
+      "loss": 0.2542,
+      "step": 8902
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3560516496879205,
+      "learning_rate": 4.1500719624299734e-07,
+      "loss": 0.0821,
+      "step": 8903
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3467056027890014,
+      "learning_rate": 4.1490542712880754e-07,
+      "loss": 0.2704,
+      "step": 8904
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5524279259988052,
+      "learning_rate": 4.148036616449234e-07,
+      "loss": 0.112,
+      "step": 8905
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.740823832942431,
+      "learning_rate": 4.147018997956862e-07,
+      "loss": 0.2801,
+      "step": 8906
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0735528649293915,
+      "learning_rate": 4.146001415854377e-07,
+      "loss": 0.2887,
+      "step": 8907
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7113644386011595,
+      "learning_rate": 4.144983870185185e-07,
+      "loss": 0.051,
+      "step": 8908
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9173098476106,
+      "learning_rate": 4.1439663609927033e-07,
+      "loss": 0.3603,
+      "step": 8909
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0221068373924442,
+      "learning_rate": 4.1429488883203346e-07,
+      "loss": 0.5015,
+      "step": 8910
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9748475047797515,
+      "learning_rate": 4.141931452211492e-07,
+      "loss": 0.0713,
+      "step": 8911
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.877476302434798,
+      "learning_rate": 4.140914052709575e-07,
+      "loss": 0.3888,
+      "step": 8912
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.035962076885889,
+      "learning_rate": 4.139896689857995e-07,
+      "loss": 0.0879,
+      "step": 8913
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3054973153770853,
+      "learning_rate": 4.138879363700149e-07,
+      "loss": 0.1952,
+      "step": 8914
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.59316800489636,
+      "learning_rate": 4.1378620742794413e-07,
+      "loss": 0.2633,
+      "step": 8915
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9554768278231138,
+      "learning_rate": 4.136844821639272e-07,
+      "loss": 0.3325,
+      "step": 8916
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5777505092967516,
+      "learning_rate": 4.135827605823035e-07,
+      "loss": 0.2842,
+      "step": 8917
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4933923077079456,
+      "learning_rate": 4.134810426874131e-07,
+      "loss": 0.1007,
+      "step": 8918
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.41674791695581065,
+      "learning_rate": 4.1337932848359526e-07,
+      "loss": 0.0506,
+      "step": 8919
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6713354467851044,
+      "learning_rate": 4.1327761797518955e-07,
+      "loss": 0.2607,
+      "step": 8920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4323185746476104,
+      "learning_rate": 4.131759111665348e-07,
+      "loss": 0.0325,
+      "step": 8921
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0622512988011819,
+      "learning_rate": 4.1307420806197036e-07,
+      "loss": 0.2112,
+      "step": 8922
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4389750806798214,
+      "learning_rate": 4.129725086658349e-07,
+      "loss": 0.0492,
+      "step": 8923
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.22444360001990274,
+      "learning_rate": 4.1287081298246716e-07,
+      "loss": 0.0761,
+      "step": 8924
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7183421669202206,
+      "learning_rate": 4.1276912101620564e-07,
+      "loss": 0.0647,
+      "step": 8925
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7079498184301862,
+      "learning_rate": 4.1266743277138894e-07,
+      "loss": 0.2825,
+      "step": 8926
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7627827987883815,
+      "learning_rate": 4.125657482523549e-07,
+      "loss": 0.297,
+      "step": 8927
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9324609714137633,
+      "learning_rate": 4.12464067463442e-07,
+      "loss": 0.3738,
+      "step": 8928
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4042411237926974,
+      "learning_rate": 4.12362390408988e-07,
+      "loss": 0.2373,
+      "step": 8929
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.09771824803843174,
+      "learning_rate": 4.122607170933304e-07,
+      "loss": 0.0008,
+      "step": 8930
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6181588386577638,
+      "learning_rate": 4.1215904752080704e-07,
+      "loss": 0.0991,
+      "step": 8931
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4188993403668053,
+      "learning_rate": 4.120573816957552e-07,
+      "loss": 0.0067,
+      "step": 8932
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5478641896560902,
+      "learning_rate": 4.119557196225125e-07,
+      "loss": 0.278,
+      "step": 8933
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8758992113226542,
+      "learning_rate": 4.118540613054155e-07,
+      "loss": 0.372,
+      "step": 8934
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9157197980811437,
+      "learning_rate": 4.117524067488017e-07,
+      "loss": 0.3069,
+      "step": 8935
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.892030904545098,
+      "learning_rate": 4.116507559570074e-07,
+      "loss": 0.2113,
+      "step": 8936
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4467649615776646,
+      "learning_rate": 4.1154910893436966e-07,
+      "loss": 0.0266,
+      "step": 8937
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.770026343102141,
+      "learning_rate": 4.1144746568522457e-07,
+      "loss": 0.1038,
+      "step": 8938
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5433695625090664,
+      "learning_rate": 4.113458262139088e-07,
+      "loss": 0.2527,
+      "step": 8939
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8449894277474345,
+      "learning_rate": 4.112441905247581e-07,
+      "loss": 0.2222,
+      "step": 8940
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7217248814820465,
+      "learning_rate": 4.111425586221087e-07,
+      "loss": 0.3618,
+      "step": 8941
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8637140262740672,
+      "learning_rate": 4.1104093051029647e-07,
+      "loss": 0.1332,
+      "step": 8942
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5222302317550118,
+      "learning_rate": 4.109393061936569e-07,
+      "loss": 0.2407,
+      "step": 8943
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.46888449597675147,
+      "learning_rate": 4.108376856765257e-07,
+      "loss": 0.106,
+      "step": 8944
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2237282015396125,
+      "learning_rate": 4.107360689632379e-07,
+      "loss": 0.2106,
+      "step": 8945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.468060232104511,
+      "learning_rate": 4.1063445605812894e-07,
+      "loss": 0.2647,
+      "step": 8946
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.831866185765841,
+      "learning_rate": 4.105328469655336e-07,
+      "loss": 0.2963,
+      "step": 8947
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1882829784434609,
+      "learning_rate": 4.10431241689787e-07,
+      "loss": 0.0757,
+      "step": 8948
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9410353486862414,
+      "learning_rate": 4.103296402352236e-07,
+      "loss": 0.1798,
+      "step": 8949
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8360667018701529,
+      "learning_rate": 4.1022804260617805e-07,
+      "loss": 0.1084,
+      "step": 8950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.41446117427026946,
+      "learning_rate": 4.1012644880698455e-07,
+      "loss": 0.2458,
+      "step": 8951
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6538221590023136,
+      "learning_rate": 4.1002485884197765e-07,
+      "loss": 0.2322,
+      "step": 8952
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8797931882510255,
+      "learning_rate": 4.0992327271549087e-07,
+      "loss": 0.0174,
+      "step": 8953
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6596385415035146,
+      "learning_rate": 4.098216904318587e-07,
+      "loss": 0.146,
+      "step": 8954
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9258696029597476,
+      "learning_rate": 4.0972011199541414e-07,
+      "loss": 0.1677,
+      "step": 8955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9562414673037318,
+      "learning_rate": 4.0961853741049135e-07,
+      "loss": 0.214,
+      "step": 8956
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6610291632584135,
+      "learning_rate": 4.095169666814233e-07,
+      "loss": 0.1226,
+      "step": 8957
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8198860743687632,
+      "learning_rate": 4.0941539981254336e-07,
+      "loss": 0.204,
+      "step": 8958
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.905845755857973,
+      "learning_rate": 4.093138368081847e-07,
+      "loss": 0.1959,
+      "step": 8959
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.211781261974601,
+      "learning_rate": 4.0921227767267974e-07,
+      "loss": 0.2329,
+      "step": 8960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4863503643373439,
+      "learning_rate": 4.0911072241036185e-07,
+      "loss": 0.0428,
+      "step": 8961
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7256574497194364,
+      "learning_rate": 4.09009171025563e-07,
+      "loss": 0.1728,
+      "step": 8962
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 18.89114656717362,
+      "learning_rate": 4.089076235226161e-07,
+      "loss": 0.4744,
+      "step": 8963
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9527457468840271,
+      "learning_rate": 4.0880607990585286e-07,
+      "loss": 0.2439,
+      "step": 8964
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.748155632069046,
+      "learning_rate": 4.0870454017960577e-07,
+      "loss": 0.3099,
+      "step": 8965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5801864280752493,
+      "learning_rate": 4.0860300434820633e-07,
+      "loss": 0.2227,
+      "step": 8966
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7190986270264014,
+      "learning_rate": 4.0850147241598654e-07,
+      "loss": 0.2481,
+      "step": 8967
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5443676111902176,
+      "learning_rate": 4.0839994438727785e-07,
+      "loss": 0.0259,
+      "step": 8968
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1092656664628424,
+      "learning_rate": 4.082984202664118e-07,
+      "loss": 0.2081,
+      "step": 8969
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7488173781044702,
+      "learning_rate": 4.0819690005771947e-07,
+      "loss": 0.1968,
+      "step": 8970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4602423968032658,
+      "learning_rate": 4.080953837655317e-07,
+      "loss": 0.3081,
+      "step": 8971
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.613948406488765,
+      "learning_rate": 4.079938713941799e-07,
+      "loss": 0.1172,
+      "step": 8972
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3240785831914836,
+      "learning_rate": 4.078923629479942e-07,
+      "loss": 0.1644,
+      "step": 8973
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.49065242535318954,
+      "learning_rate": 4.077908584313058e-07,
+      "loss": 0.3614,
+      "step": 8974
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6036816709428935,
+      "learning_rate": 4.0768935784844444e-07,
+      "loss": 0.1832,
+      "step": 8975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.656090000346207,
+      "learning_rate": 4.0758786120374073e-07,
+      "loss": 0.1065,
+      "step": 8976
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1238109859775967,
+      "learning_rate": 4.0748636850152457e-07,
+      "loss": 0.4529,
+      "step": 8977
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.658608595335291,
+      "learning_rate": 4.0738487974612613e-07,
+      "loss": 0.3871,
+      "step": 8978
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.020767182452796,
+      "learning_rate": 4.0728339494187465e-07,
+      "loss": 0.2117,
+      "step": 8979
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.587761079292793,
+      "learning_rate": 4.071819140931002e-07,
+      "loss": 0.225,
+      "step": 8980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1746589604840403,
+      "learning_rate": 4.0708043720413157e-07,
+      "loss": 0.3331,
+      "step": 8981
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6432716813705855,
+      "learning_rate": 4.069789642792986e-07,
+      "loss": 0.009,
+      "step": 8982
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6342285927951091,
+      "learning_rate": 4.068774953229297e-07,
+      "loss": 0.0823,
+      "step": 8983
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.316941299192119,
+      "learning_rate": 4.067760303393543e-07,
+      "loss": 0.0673,
+      "step": 8984
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7977403896202583,
+      "learning_rate": 4.066745693329007e-07,
+      "loss": 0.0088,
+      "step": 8985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6602867804753216,
+      "learning_rate": 4.0657311230789757e-07,
+      "loss": 0.3205,
+      "step": 8986
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3875502905647615,
+      "learning_rate": 4.0647165926867345e-07,
+      "loss": 0.2616,
+      "step": 8987
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1612784449805174,
+      "learning_rate": 4.0637021021955615e-07,
+      "loss": 0.3196,
+      "step": 8988
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4985510654803668,
+      "learning_rate": 4.062687651648741e-07,
+      "loss": 0.0171,
+      "step": 8989
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27865756818428206,
+      "learning_rate": 4.061673241089547e-07,
+      "loss": 0.0365,
+      "step": 8990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3307056343237686,
+      "learning_rate": 4.060658870561262e-07,
+      "loss": 0.0844,
+      "step": 8991
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.243988055057629,
+      "learning_rate": 4.0596445401071547e-07,
+      "loss": 0.073,
+      "step": 8992
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2478361278616963,
+      "learning_rate": 4.058630249770504e-07,
+      "loss": 0.2899,
+      "step": 8993
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3104727381547325,
+      "learning_rate": 4.057615999594578e-07,
+      "loss": 0.2499,
+      "step": 8994
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.32810876773928077,
+      "learning_rate": 4.0566017896226485e-07,
+      "loss": 0.1811,
+      "step": 8995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8333586188888399,
+      "learning_rate": 4.0555876198979817e-07,
+      "loss": 0.3105,
+      "step": 8996
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2441767375955954,
+      "learning_rate": 4.054573490463848e-07,
+      "loss": 0.0094,
+      "step": 8997
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0921772024803698,
+      "learning_rate": 4.0535594013635093e-07,
+      "loss": 0.058,
+      "step": 8998
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2390422713542835,
+      "learning_rate": 4.0525453526402276e-07,
+      "loss": 0.1206,
+      "step": 8999
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1778885430337904,
+      "learning_rate": 4.0515313443372675e-07,
+      "loss": 0.2504,
+      "step": 9000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.08698806208004202,
+      "learning_rate": 4.050517376497885e-07,
+      "loss": 0.0028,
+      "step": 9001
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.46755935882190264,
+      "learning_rate": 4.049503449165341e-07,
+      "loss": 0.0989,
+      "step": 9002
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7650949509322319,
+      "learning_rate": 4.04848956238289e-07,
+      "loss": 0.1436,
+      "step": 9003
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2979398501256509,
+      "learning_rate": 4.0474757161937876e-07,
+      "loss": 0.141,
+      "step": 9004
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0035517416848905,
+      "learning_rate": 4.046461910641285e-07,
+      "loss": 0.2083,
+      "step": 9005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 24.001824481540783,
+      "learning_rate": 4.045448145768636e-07,
+      "loss": 0.3099,
+      "step": 9006
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.783926302428239,
+      "learning_rate": 4.0444344216190863e-07,
+      "loss": 0.2344,
+      "step": 9007
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.939785030272276,
+      "learning_rate": 4.0434207382358865e-07,
+      "loss": 0.1586,
+      "step": 9008
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7242003816517104,
+      "learning_rate": 4.042407095662279e-07,
+      "loss": 0.129,
+      "step": 9009
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8403834322257356,
+      "learning_rate": 4.0413934939415126e-07,
+      "loss": 0.3,
+      "step": 9010
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4074080079376277,
+      "learning_rate": 4.0403799331168243e-07,
+      "loss": 0.2928,
+      "step": 9011
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4771753974956194,
+      "learning_rate": 4.0393664132314576e-07,
+      "loss": 0.1581,
+      "step": 9012
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6087305966614077,
+      "learning_rate": 4.0383529343286525e-07,
+      "loss": 0.0752,
+      "step": 9013
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4078878006792705,
+      "learning_rate": 4.037339496451642e-07,
+      "loss": 0.0535,
+      "step": 9014
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3292486987075627,
+      "learning_rate": 4.0363260996436666e-07,
+      "loss": 0.0814,
+      "step": 9015
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6243070981600599,
+      "learning_rate": 4.035312743947954e-07,
+      "loss": 0.1725,
+      "step": 9016
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4245762187153739,
+      "learning_rate": 4.0342994294077414e-07,
+      "loss": 0.1674,
+      "step": 9017
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.22093514648280238,
+      "learning_rate": 4.0332861560662547e-07,
+      "loss": 0.0796,
+      "step": 9018
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7260767099304124,
+      "learning_rate": 4.0322729239667253e-07,
+      "loss": 0.1967,
+      "step": 9019
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7598463566865133,
+      "learning_rate": 4.031259733152377e-07,
+      "loss": 0.2083,
+      "step": 9020
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.661778512892847,
+      "learning_rate": 4.030246583666437e-07,
+      "loss": 0.211,
+      "step": 9021
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5116657482186662,
+      "learning_rate": 4.029233475552125e-07,
+      "loss": 0.2962,
+      "step": 9022
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4688982985790497,
+      "learning_rate": 4.0282204088526674e-07,
+      "loss": 0.1241,
+      "step": 9023
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.45749741855101583,
+      "learning_rate": 4.0272073836112786e-07,
+      "loss": 0.4121,
+      "step": 9024
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9434951366675195,
+      "learning_rate": 4.0261943998711803e-07,
+      "loss": 0.1673,
+      "step": 9025
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.20550577244305945,
+      "learning_rate": 4.025181457675587e-07,
+      "loss": 0.0588,
+      "step": 9026
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.428451287773308,
+      "learning_rate": 4.024168557067709e-07,
+      "loss": 0.1814,
+      "step": 9027
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8381884433585038,
+      "learning_rate": 4.023155698090764e-07,
+      "loss": 0.2038,
+      "step": 9028
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7965236383737759,
+      "learning_rate": 4.02214288078796e-07,
+      "loss": 0.1706,
+      "step": 9029
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6777519886371323,
+      "learning_rate": 4.021130105202507e-07,
+      "loss": 0.0079,
+      "step": 9030
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5632899213811604,
+      "learning_rate": 4.0201173713776105e-07,
+      "loss": 0.156,
+      "step": 9031
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9247967735208444,
+      "learning_rate": 4.0191046793564787e-07,
+      "loss": 0.3693,
+      "step": 9032
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5713296572055833,
+      "learning_rate": 4.0180920291823114e-07,
+      "loss": 0.3208,
+      "step": 9033
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5066304749829127,
+      "learning_rate": 4.017079420898314e-07,
+      "loss": 0.1027,
+      "step": 9034
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7147742170935585,
+      "learning_rate": 4.016066854547682e-07,
+      "loss": 0.1065,
+      "step": 9035
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3069685824116117,
+      "learning_rate": 4.015054330173618e-07,
+      "loss": 0.3056,
+      "step": 9036
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.667330074309844,
+      "learning_rate": 4.014041847819314e-07,
+      "loss": 0.1823,
+      "step": 9037
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9773945518802425,
+      "learning_rate": 4.013029407527968e-07,
+      "loss": 0.1436,
+      "step": 9038
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2996877867987215,
+      "learning_rate": 4.012017009342773e-07,
+      "loss": 0.1427,
+      "step": 9039
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4452488615195211,
+      "learning_rate": 4.0110046533069157e-07,
+      "loss": 0.3065,
+      "step": 9040
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.30645353995812,
+      "learning_rate": 4.00999233946359e-07,
+      "loss": 0.333,
+      "step": 9041
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4956683821175165,
+      "learning_rate": 4.0089800678559803e-07,
+      "loss": 0.1727,
+      "step": 9042
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.739914002270914,
+      "learning_rate": 4.007967838527274e-07,
+      "loss": 0.0503,
+      "step": 9043
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3603300123768585,
+      "learning_rate": 4.0069556515206527e-07,
+      "loss": 0.1889,
+      "step": 9044
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9760889355895916,
+      "learning_rate": 4.005943506879301e-07,
+      "loss": 0.077,
+      "step": 9045
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.081222799612338,
+      "learning_rate": 4.004931404646397e-07,
+      "loss": 0.3447,
+      "step": 9046
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4811212383192218,
+      "learning_rate": 4.003919344865119e-07,
+      "loss": 0.2423,
+      "step": 9047
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2905747419769944,
+      "learning_rate": 4.0029073275786436e-07,
+      "loss": 0.0074,
+      "step": 9048
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.09817365571362,
+      "learning_rate": 4.001895352830148e-07,
+      "loss": 0.0707,
+      "step": 9049
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9445935114472639,
+      "learning_rate": 4.000883420662801e-07,
+      "loss": 0.2193,
+      "step": 9050
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4456565270365584,
+      "learning_rate": 3.9998715311197783e-07,
+      "loss": 0.2212,
+      "step": 9051
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0246092475515236,
+      "learning_rate": 3.998859684244244e-07,
+      "loss": 0.2089,
+      "step": 9052
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.215196055215365,
+      "learning_rate": 3.99784788007937e-07,
+      "loss": 0.3612,
+      "step": 9053
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7285277373881145,
+      "learning_rate": 3.99683611866832e-07,
+      "loss": 0.1118,
+      "step": 9054
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.310467250687471,
+      "learning_rate": 3.9958244000542566e-07,
+      "loss": 0.2895,
+      "step": 9055
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.721133612694442,
+      "learning_rate": 3.9948127242803437e-07,
+      "loss": 0.1636,
+      "step": 9056
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.025961345957209,
+      "learning_rate": 3.993801091389739e-07,
+      "loss": 0.1483,
+      "step": 9057
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9314000741301101,
+      "learning_rate": 3.9927895014256053e-07,
+      "loss": 0.1196,
+      "step": 9058
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.420191387188993,
+      "learning_rate": 3.9917779544310935e-07,
+      "loss": 0.1613,
+      "step": 9059
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4235913058231204,
+      "learning_rate": 3.990766450449363e-07,
+      "loss": 0.1023,
+      "step": 9060
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2195491324790444,
+      "learning_rate": 3.989754989523563e-07,
+      "loss": 0.2947,
+      "step": 9061
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7037524693812025,
+      "learning_rate": 3.9887435716968484e-07,
+      "loss": 0.1254,
+      "step": 9062
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5073221242159822,
+      "learning_rate": 3.987732197012363e-07,
+      "loss": 0.2778,
+      "step": 9063
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8665317055890085,
+      "learning_rate": 3.9867208655132586e-07,
+      "loss": 0.0605,
+      "step": 9064
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.762887116434521,
+      "learning_rate": 3.9857095772426784e-07,
+      "loss": 0.0755,
+      "step": 9065
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9474658750577679,
+      "learning_rate": 3.9846983322437667e-07,
+      "loss": 0.1098,
+      "step": 9066
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.36958400160465527,
+      "learning_rate": 3.983687130559666e-07,
+      "loss": 0.1108,
+      "step": 9067
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7325477158345699,
+      "learning_rate": 3.982675972233514e-07,
+      "loss": 0.2775,
+      "step": 9068
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6345864761335269,
+      "learning_rate": 3.981664857308452e-07,
+      "loss": 0.07,
+      "step": 9069
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5382803473694683,
+      "learning_rate": 3.9806537858276124e-07,
+      "loss": 0.2797,
+      "step": 9070
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.163239855383378,
+      "learning_rate": 3.979642757834133e-07,
+      "loss": 0.1866,
+      "step": 9071
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5826955762141282,
+      "learning_rate": 3.9786317733711427e-07,
+      "loss": 0.1656,
+      "step": 9072
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8460086516001164,
+      "learning_rate": 3.9776208324817755e-07,
+      "loss": 0.096,
+      "step": 9073
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8705191724337407,
+      "learning_rate": 3.9766099352091587e-07,
+      "loss": 0.3196,
+      "step": 9074
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5142661317455479,
+      "learning_rate": 3.975599081596419e-07,
+      "loss": 0.156,
+      "step": 9075
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6104047618995742,
+      "learning_rate": 3.9745882716866813e-07,
+      "loss": 0.031,
+      "step": 9076
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7345028381558942,
+      "learning_rate": 3.9735775055230714e-07,
+      "loss": 0.355,
+      "step": 9077
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.588846068939626,
+      "learning_rate": 3.972566783148706e-07,
+      "loss": 0.1142,
+      "step": 9078
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.341741735597102,
+      "learning_rate": 3.9715561046067094e-07,
+      "loss": 0.1682,
+      "step": 9079
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1621309997240818,
+      "learning_rate": 3.9705454699401963e-07,
+      "loss": 0.0015,
+      "step": 9080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4465533222689928,
+      "learning_rate": 3.969534879192281e-07,
+      "loss": 0.2088,
+      "step": 9081
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3472903473872619,
+      "learning_rate": 3.9685243324060807e-07,
+      "loss": 0.1967,
+      "step": 9082
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.89663315989137,
+      "learning_rate": 3.9675138296247045e-07,
+      "loss": 0.0303,
+      "step": 9083
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6255986993763517,
+      "learning_rate": 3.966503370891266e-07,
+      "loss": 0.1166,
+      "step": 9084
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3405934575042138,
+      "learning_rate": 3.9654929562488693e-07,
+      "loss": 0.2685,
+      "step": 9085
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5722119128549745,
+      "learning_rate": 3.964482585740625e-07,
+      "loss": 0.1406,
+      "step": 9086
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.24404177857455117,
+      "learning_rate": 3.9634722594096325e-07,
+      "loss": 0.0997,
+      "step": 9087
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.681805078870619,
+      "learning_rate": 3.962461977299e-07,
+      "loss": 0.2331,
+      "step": 9088
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8243376873980711,
+      "learning_rate": 3.961451739451823e-07,
+      "loss": 0.1161,
+      "step": 9089
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7757202223089231,
+      "learning_rate": 3.960441545911204e-07,
+      "loss": 0.0637,
+      "step": 9090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7456104967572204,
+      "learning_rate": 3.959431396720237e-07,
+      "loss": 0.083,
+      "step": 9091
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1350634707790515,
+      "learning_rate": 3.9584212919220196e-07,
+      "loss": 0.2344,
+      "step": 9092
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.878038497687454,
+      "learning_rate": 3.957411231559642e-07,
+      "loss": 0.1174,
+      "step": 9093
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0787155370288095,
+      "learning_rate": 3.9564012156761994e-07,
+      "loss": 0.1609,
+      "step": 9094
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7154063988683648,
+      "learning_rate": 3.955391244314779e-07,
+      "loss": 0.05,
+      "step": 9095
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8266407603287446,
+      "learning_rate": 3.954381317518466e-07,
+      "loss": 0.168,
+      "step": 9096
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.122799645277665,
+      "learning_rate": 3.95337143533035e-07,
+      "loss": 0.3171,
+      "step": 9097
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6984134699333916,
+      "learning_rate": 3.9523615977935097e-07,
+      "loss": 0.3383,
+      "step": 9098
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1842718551636655,
+      "learning_rate": 3.951351804951032e-07,
+      "loss": 0.0621,
+      "step": 9099
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1785945127202282,
+      "learning_rate": 3.9503420568459936e-07,
+      "loss": 0.3907,
+      "step": 9100
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6463394683670811,
+      "learning_rate": 3.9493323535214737e-07,
+      "loss": 0.1419,
+      "step": 9101
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0666113227064726,
+      "learning_rate": 3.948322695020546e-07,
+      "loss": 0.158,
+      "step": 9102
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3029223594752769,
+      "learning_rate": 3.947313081386289e-07,
+      "loss": 0.0003,
+      "step": 9103
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2458121378905718,
+      "learning_rate": 3.9463035126617696e-07,
+      "loss": 0.1343,
+      "step": 9104
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5893834935333186,
+      "learning_rate": 3.945293988890062e-07,
+      "loss": 0.259,
+      "step": 9105
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3004152307314003,
+      "learning_rate": 3.944284510114232e-07,
+      "loss": 0.3243,
+      "step": 9106
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5497385037057883,
+      "learning_rate": 3.943275076377349e-07,
+      "loss": 0.1432,
+      "step": 9107
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7587435307983558,
+      "learning_rate": 3.942265687722474e-07,
+      "loss": 0.3035,
+      "step": 9108
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4121270975608575,
+      "learning_rate": 3.9412563441926706e-07,
+      "loss": 0.2223,
+      "step": 9109
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3152190806198152,
+      "learning_rate": 3.940247045831001e-07,
+      "loss": 0.3556,
+      "step": 9110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.077669034473468,
+      "learning_rate": 3.939237792680522e-07,
+      "loss": 0.3115,
+      "step": 9111
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5996167287361195,
+      "learning_rate": 3.938228584784292e-07,
+      "loss": 0.3228,
+      "step": 9112
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8498250818324076,
+      "learning_rate": 3.9372194221853636e-07,
+      "loss": 0.3961,
+      "step": 9113
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9441137172312675,
+      "learning_rate": 3.9362103049267934e-07,
+      "loss": 0.0193,
+      "step": 9114
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1079625665689703,
+      "learning_rate": 3.9352012330516275e-07,
+      "loss": 0.377,
+      "step": 9115
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5648126864513543,
+      "learning_rate": 3.9341922066029205e-07,
+      "loss": 0.1762,
+      "step": 9116
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2391696149516038,
+      "learning_rate": 3.9331832256237136e-07,
+      "loss": 0.0849,
+      "step": 9117
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 28.27204448115714,
+      "learning_rate": 3.932174290157056e-07,
+      "loss": 0.291,
+      "step": 9118
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1459072452225505,
+      "learning_rate": 3.9311654002459896e-07,
+      "loss": 0.0047,
+      "step": 9119
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 14.449940850162786,
+      "learning_rate": 3.9301565559335565e-07,
+      "loss": 0.3272,
+      "step": 9120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.907058481995523,
+      "learning_rate": 3.929147757262794e-07,
+      "loss": 0.099,
+      "step": 9121
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.601482362661088,
+      "learning_rate": 3.9281390042767423e-07,
+      "loss": 0.0294,
+      "step": 9122
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.695113172316583,
+      "learning_rate": 3.927130297018436e-07,
+      "loss": 0.2036,
+      "step": 9123
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4512606804656528,
+      "learning_rate": 3.9261216355309063e-07,
+      "loss": 0.1826,
+      "step": 9124
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.40888229553450833,
+      "learning_rate": 3.925113019857187e-07,
+      "loss": 0.0128,
+      "step": 9125
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.352895839425,
+      "learning_rate": 3.924104450040307e-07,
+      "loss": 0.2253,
+      "step": 9126
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8631112591496051,
+      "learning_rate": 3.9230959261232943e-07,
+      "loss": 0.1636,
+      "step": 9127
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9239130250737256,
+      "learning_rate": 3.9220874481491726e-07,
+      "loss": 0.4013,
+      "step": 9128
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8545318574040791,
+      "learning_rate": 3.92107901616097e-07,
+      "loss": 0.123,
+      "step": 9129
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6277111029733823,
+      "learning_rate": 3.920070630201703e-07,
+      "loss": 0.1028,
+      "step": 9130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.025476062430447,
+      "learning_rate": 3.9190622903143953e-07,
+      "loss": 0.0061,
+      "step": 9131
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.43370590315597163,
+      "learning_rate": 3.918053996542062e-07,
+      "loss": 0.0681,
+      "step": 9132
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.21568988514208015,
+      "learning_rate": 3.9170457489277227e-07,
+      "loss": 0.0934,
+      "step": 9133
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2134802163903566,
+      "learning_rate": 3.916037547514386e-07,
+      "loss": 0.48,
+      "step": 9134
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.44384820722314655,
+      "learning_rate": 3.9150293923450684e-07,
+      "loss": 0.2024,
+      "step": 9135
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5290360688562474,
+      "learning_rate": 3.914021283462777e-07,
+      "loss": 0.2915,
+      "step": 9136
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.199319608565975,
+      "learning_rate": 3.9130132209105207e-07,
+      "loss": 0.0644,
+      "step": 9137
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8110329089991334,
+      "learning_rate": 3.912005204731307e-07,
+      "loss": 0.2226,
+      "step": 9138
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6090097521712758,
+      "learning_rate": 3.9109972349681357e-07,
+      "loss": 0.1522,
+      "step": 9139
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0628346686616439,
+      "learning_rate": 3.909989311664014e-07,
+      "loss": 0.1917,
+      "step": 9140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0460528081900293,
+      "learning_rate": 3.9089814348619376e-07,
+      "loss": 0.4451,
+      "step": 9141
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8136537433216505,
+      "learning_rate": 3.907973604604908e-07,
+      "loss": 0.1841,
+      "step": 9142
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2886978776955402,
+      "learning_rate": 3.9069658209359186e-07,
+      "loss": 0.266,
+      "step": 9143
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.202882225819883,
+      "learning_rate": 3.905958083897965e-07,
+      "loss": 0.2374,
+      "step": 9144
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.500113987439878,
+      "learning_rate": 3.9049503935340386e-07,
+      "loss": 0.2706,
+      "step": 9145
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1938992054164177,
+      "learning_rate": 3.90394274988713e-07,
+      "loss": 0.143,
+      "step": 9146
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7248861178693087,
+      "learning_rate": 3.9029351530002264e-07,
+      "loss": 0.0849,
+      "step": 9147
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8164298406079303,
+      "learning_rate": 3.901927602916316e-07,
+      "loss": 0.0368,
+      "step": 9148
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7645650435096253,
+      "learning_rate": 3.900920099678382e-07,
+      "loss": 0.1669,
+      "step": 9149
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6462155620964484,
+      "learning_rate": 3.899912643329403e-07,
+      "loss": 0.0824,
+      "step": 9150
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6100529014449996,
+      "learning_rate": 3.898905233912365e-07,
+      "loss": 0.3558,
+      "step": 9151
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.44330946916766,
+      "learning_rate": 3.897897871470241e-07,
+      "loss": 0.004,
+      "step": 9152
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7378142353289706,
+      "learning_rate": 3.8968905560460095e-07,
+      "loss": 0.0732,
+      "step": 9153
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7624836196576208,
+      "learning_rate": 3.895883287682644e-07,
+      "loss": 0.2406,
+      "step": 9154
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1989838823723706,
+      "learning_rate": 3.8948760664231194e-07,
+      "loss": 0.2578,
+      "step": 9155
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.38854518237118085,
+      "learning_rate": 3.893868892310401e-07,
+      "loss": 0.0291,
+      "step": 9156
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6847271574043243,
+      "learning_rate": 3.8928617653874616e-07,
+      "loss": 0.3524,
+      "step": 9157
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4334590327411633,
+      "learning_rate": 3.891854685697263e-07,
+      "loss": 0.0711,
+      "step": 9158
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.411262064489176,
+      "learning_rate": 3.8908476532827727e-07,
+      "loss": 0.2449,
+      "step": 9159
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3337348207732892,
+      "learning_rate": 3.889840668186949e-07,
+      "loss": 0.1965,
+      "step": 9160
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.129793996447344,
+      "learning_rate": 3.8888337304527564e-07,
+      "loss": 0.3591,
+      "step": 9161
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4164599484664437,
+      "learning_rate": 3.8878268401231487e-07,
+      "loss": 0.1566,
+      "step": 9162
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1135343748933653,
+      "learning_rate": 3.8868199972410846e-07,
+      "loss": 0.186,
+      "step": 9163
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9535641093883805,
+      "learning_rate": 3.8858132018495184e-07,
+      "loss": 0.2839,
+      "step": 9164
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4248583334239953,
+      "learning_rate": 3.884806453991398e-07,
+      "loss": 0.1362,
+      "step": 9165
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.49284576972669875,
+      "learning_rate": 3.883799753709679e-07,
+      "loss": 0.3284,
+      "step": 9166
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.028064848108264,
+      "learning_rate": 3.882793101047304e-07,
+      "loss": 0.2369,
+      "step": 9167
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2460391072109868,
+      "learning_rate": 3.8817864960472233e-07,
+      "loss": 0.009,
+      "step": 9168
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.061349969952489876,
+      "learning_rate": 3.8807799387523765e-07,
+      "loss": 0.001,
+      "step": 9169
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5311125442366161,
+      "learning_rate": 3.8797734292057087e-07,
+      "loss": 0.0622,
+      "step": 9170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8721603956515364,
+      "learning_rate": 3.8787669674501575e-07,
+      "loss": 0.1645,
+      "step": 9171
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1911385618025718,
+      "learning_rate": 3.877760553528663e-07,
+      "loss": 0.2388,
+      "step": 9172
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.44380367715135,
+      "learning_rate": 3.876754187484157e-07,
+      "loss": 0.1058,
+      "step": 9173
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0451666161687792,
+      "learning_rate": 3.8757478693595777e-07,
+      "loss": 0.3121,
+      "step": 9174
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.14264998394508208,
+      "learning_rate": 3.874741599197852e-07,
+      "loss": 0.0794,
+      "step": 9175
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6246629504104158,
+      "learning_rate": 3.873735377041914e-07,
+      "loss": 0.085,
+      "step": 9176
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5304588390550291,
+      "learning_rate": 3.872729202934689e-07,
+      "loss": 0.1211,
+      "step": 9177
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2139916211796347,
+      "learning_rate": 3.8717230769191003e-07,
+      "loss": 0.1638,
+      "step": 9178
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4945753870794083,
+      "learning_rate": 3.870716999038074e-07,
+      "loss": 0.4496,
+      "step": 9179
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4775768718401756,
+      "learning_rate": 3.869710969334531e-07,
+      "loss": 0.1229,
+      "step": 9180
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7109870809856619,
+      "learning_rate": 3.86870498785139e-07,
+      "loss": 0.2483,
+      "step": 9181
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9287785944134873,
+      "learning_rate": 3.8676990546315676e-07,
+      "loss": 0.1111,
+      "step": 9182
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7729595500918812,
+      "learning_rate": 3.8666931697179816e-07,
+      "loss": 0.2111,
+      "step": 9183
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6794761483883275,
+      "learning_rate": 3.8656873331535413e-07,
+      "loss": 0.2492,
+      "step": 9184
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4381350380622578,
+      "learning_rate": 3.8646815449811617e-07,
+      "loss": 0.0783,
+      "step": 9185
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7720450624517092,
+      "learning_rate": 3.863675805243747e-07,
+      "loss": 0.1668,
+      "step": 9186
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4921296405996789,
+      "learning_rate": 3.8626701139842094e-07,
+      "loss": 0.217,
+      "step": 9187
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8177338474651337,
+      "learning_rate": 3.8616644712454486e-07,
+      "loss": 0.1114,
+      "step": 9188
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.166284876492476,
+      "learning_rate": 3.860658877070371e-07,
+      "loss": 0.1488,
+      "step": 9189
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.143827540947076,
+      "learning_rate": 3.8596533315018775e-07,
+      "loss": 0.0627,
+      "step": 9190
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8860199853316247,
+      "learning_rate": 3.858647834582863e-07,
+      "loss": 0.2058,
+      "step": 9191
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8097965018820006,
+      "learning_rate": 3.857642386356228e-07,
+      "loss": 0.3329,
+      "step": 9192
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.070774336932143,
+      "learning_rate": 3.8566369868648637e-07,
+      "loss": 0.1568,
+      "step": 9193
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2699900825217016,
+      "learning_rate": 3.855631636151666e-07,
+      "loss": 0.2194,
+      "step": 9194
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6469029270648288,
+      "learning_rate": 3.854626334259521e-07,
+      "loss": 0.2912,
+      "step": 9195
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5244764993476563,
+      "learning_rate": 3.853621081231321e-07,
+      "loss": 0.1268,
+      "step": 9196
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9962098492424587,
+      "learning_rate": 3.852615877109948e-07,
+      "loss": 0.2871,
+      "step": 9197
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.7032776124392495,
+      "learning_rate": 3.851610721938289e-07,
+      "loss": 0.315,
+      "step": 9198
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8505050123015702,
+      "learning_rate": 3.850605615759225e-07,
+      "loss": 0.0576,
+      "step": 9199
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.707335169164966,
+      "learning_rate": 3.849600558615637e-07,
+      "loss": 0.1038,
+      "step": 9200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0816445579365792,
+      "learning_rate": 3.8485955505504004e-07,
+      "loss": 0.075,
+      "step": 9201
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1298037125380591,
+      "learning_rate": 3.8475905916063947e-07,
+      "loss": 0.0017,
+      "step": 9202
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2499967898855564,
+      "learning_rate": 3.8465856818264873e-07,
+      "loss": 0.0452,
+      "step": 9203
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7829148141341262,
+      "learning_rate": 3.8455808212535566e-07,
+      "loss": 0.2444,
+      "step": 9204
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.45011960836309023,
+      "learning_rate": 3.844576009930468e-07,
+      "loss": 0.264,
+      "step": 9205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1994074093175575,
+      "learning_rate": 3.843571247900089e-07,
+      "loss": 0.0186,
+      "step": 9206
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3517055045129158,
+      "learning_rate": 3.8425665352052854e-07,
+      "loss": 0.4352,
+      "step": 9207
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7661274484486612,
+      "learning_rate": 3.841561871888919e-07,
+      "loss": 0.1375,
+      "step": 9208
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7206736025970967,
+      "learning_rate": 3.8405572579938545e-07,
+      "loss": 0.1006,
+      "step": 9209
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4045134212470072,
+      "learning_rate": 3.839552693562945e-07,
+      "loss": 0.1919,
+      "step": 9210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2670005349429003,
+      "learning_rate": 3.838548178639054e-07,
+      "loss": 0.191,
+      "step": 9211
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3860446118169985,
+      "learning_rate": 3.837543713265029e-07,
+      "loss": 0.1544,
+      "step": 9212
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6092347251192362,
+      "learning_rate": 3.8365392974837286e-07,
+      "loss": 0.0883,
+      "step": 9213
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7268909521142994,
+      "learning_rate": 3.8355349313379985e-07,
+      "loss": 0.507,
+      "step": 9214
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6860064248322543,
+      "learning_rate": 3.8345306148706903e-07,
+      "loss": 0.2763,
+      "step": 9215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5713052815692077,
+      "learning_rate": 3.833526348124649e-07,
+      "loss": 0.4124,
+      "step": 9216
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7284454410432929,
+      "learning_rate": 3.8325221311427187e-07,
+      "loss": 0.3609,
+      "step": 9217
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5045497683987311,
+      "learning_rate": 3.831517963967742e-07,
+      "loss": 0.1256,
+      "step": 9218
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7560206150188047,
+      "learning_rate": 3.830513846642556e-07,
+      "loss": 0.3329,
+      "step": 9219
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5132619562727003,
+      "learning_rate": 3.829509779210002e-07,
+      "loss": 0.2106,
+      "step": 9220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6174686597443946,
+      "learning_rate": 3.828505761712912e-07,
+      "loss": 0.2803,
+      "step": 9221
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8508200040860678,
+      "learning_rate": 3.827501794194123e-07,
+      "loss": 0.3196,
+      "step": 9222
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1476513717416568,
+      "learning_rate": 3.8264978766964627e-07,
+      "loss": 0.1843,
+      "step": 9223
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.49750688205076526,
+      "learning_rate": 3.825494009262763e-07,
+      "loss": 0.1874,
+      "step": 9224
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6492510892939429,
+      "learning_rate": 3.8244901919358486e-07,
+      "loss": 0.354,
+      "step": 9225
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6144163882567673,
+      "learning_rate": 3.823486424758548e-07,
+      "loss": 0.2499,
+      "step": 9226
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.003569649754277,
+      "learning_rate": 3.82248270777368e-07,
+      "loss": 0.2284,
+      "step": 9227
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2232116338891916,
+      "learning_rate": 3.8214790410240685e-07,
+      "loss": 0.2695,
+      "step": 9228
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5615557423366042,
+      "learning_rate": 3.820475424552527e-07,
+      "loss": 0.2816,
+      "step": 9229
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5172325479600991,
+      "learning_rate": 3.8194718584018787e-07,
+      "loss": 0.2323,
+      "step": 9230
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.339617527615744,
+      "learning_rate": 3.8184683426149315e-07,
+      "loss": 0.2372,
+      "step": 9231
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.635007534459703,
+      "learning_rate": 3.817464877234502e-07,
+      "loss": 0.2388,
+      "step": 9232
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1168657342703339,
+      "learning_rate": 3.816461462303397e-07,
+      "loss": 0.1081,
+      "step": 9233
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8265448121999057,
+      "learning_rate": 3.815458097864424e-07,
+      "loss": 0.3288,
+      "step": 9234
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4160038277515097,
+      "learning_rate": 3.8144547839603923e-07,
+      "loss": 0.0752,
+      "step": 9235
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2909499803288882,
+      "learning_rate": 3.8134515206341005e-07,
+      "loss": 0.0294,
+      "step": 9236
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9679006141827274,
+      "learning_rate": 3.8124483079283543e-07,
+      "loss": 0.2623,
+      "step": 9237
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1028104956836575,
+      "learning_rate": 3.8114451458859483e-07,
+      "loss": 0.2275,
+      "step": 9238
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.22196268196621946,
+      "learning_rate": 3.8104420345496837e-07,
+      "loss": 0.1058,
+      "step": 9239
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.993825894678243,
+      "learning_rate": 3.809438973962351e-07,
+      "loss": 0.2151,
+      "step": 9240
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0319915627413767,
+      "learning_rate": 3.8084359641667476e-07,
+      "loss": 0.076,
+      "step": 9241
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.07997144953621751,
+      "learning_rate": 3.8074330052056587e-07,
+      "loss": 0.0012,
+      "step": 9242
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7991306537992611,
+      "learning_rate": 3.806430097121876e-07,
+      "loss": 0.2895,
+      "step": 9243
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8036803252125937,
+      "learning_rate": 3.8054272399581844e-07,
+      "loss": 0.1936,
+      "step": 9244
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8102653795342827,
+      "learning_rate": 3.8044244337573694e-07,
+      "loss": 0.2513,
+      "step": 9245
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8675800393996959,
+      "learning_rate": 3.8034216785622125e-07,
+      "loss": 0.0464,
+      "step": 9246
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2588834721619737,
+      "learning_rate": 3.802418974415489e-07,
+      "loss": 0.3701,
+      "step": 9247
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0249739483080074,
+      "learning_rate": 3.8014163213599824e-07,
+      "loss": 0.0737,
+      "step": 9248
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 14.622398411848904,
+      "learning_rate": 3.800413719438463e-07,
+      "loss": 0.3292,
+      "step": 9249
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.40679487772653566,
+      "learning_rate": 3.7994111686937066e-07,
+      "loss": 0.2561,
+      "step": 9250
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2984174953844867,
+      "learning_rate": 3.798408669168483e-07,
+      "loss": 0.0028,
+      "step": 9251
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4858604040673141,
+      "learning_rate": 3.7974062209055615e-07,
+      "loss": 0.1167,
+      "step": 9252
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.563922569166749,
+      "learning_rate": 3.7964038239477065e-07,
+      "loss": 0.3193,
+      "step": 9253
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.650429162744675,
+      "learning_rate": 3.7954014783376865e-07,
+      "loss": 0.1074,
+      "step": 9254
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.61252308008292,
+      "learning_rate": 3.794399184118258e-07,
+      "loss": 0.1816,
+      "step": 9255
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6142621497538089,
+      "learning_rate": 3.793396941332187e-07,
+      "loss": 0.0874,
+      "step": 9256
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2884583985225477,
+      "learning_rate": 3.7923947500222256e-07,
+      "loss": 0.1975,
+      "step": 9257
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.110770340974891,
+      "learning_rate": 3.7913926102311337e-07,
+      "loss": 0.1255,
+      "step": 9258
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3867503123187839,
+      "learning_rate": 3.790390522001662e-07,
+      "loss": 0.08,
+      "step": 9259
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3152861298882226,
+      "learning_rate": 3.789388485376561e-07,
+      "loss": 0.1842,
+      "step": 9260
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6016674065452318,
+      "learning_rate": 3.7883865003985824e-07,
+      "loss": 0.3036,
+      "step": 9261
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0199353647149971,
+      "learning_rate": 3.78738456711047e-07,
+      "loss": 0.1298,
+      "step": 9262
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5743923291555721,
+      "learning_rate": 3.786382685554972e-07,
+      "loss": 0.2609,
+      "step": 9263
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1420267270789695,
+      "learning_rate": 3.785380855774826e-07,
+      "loss": 0.2284,
+      "step": 9264
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3823484926973197,
+      "learning_rate": 3.784379077812776e-07,
+      "loss": 0.1094,
+      "step": 9265
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6768785129797791,
+      "learning_rate": 3.783377351711556e-07,
+      "loss": 0.4861,
+      "step": 9266
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.493585718026856,
+      "learning_rate": 3.7823756775139064e-07,
+      "loss": 0.1025,
+      "step": 9267
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5133365280139011,
+      "learning_rate": 3.781374055262556e-07,
+      "loss": 0.1072,
+      "step": 9268
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6378779444642336,
+      "learning_rate": 3.7803724850002386e-07,
+      "loss": 0.3225,
+      "step": 9269
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6305097473210857,
+      "learning_rate": 3.779370966769683e-07,
+      "loss": 0.0843,
+      "step": 9270
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7640534442975933,
+      "learning_rate": 3.7783695006136166e-07,
+      "loss": 0.3095,
+      "step": 9271
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7972338856128529,
+      "learning_rate": 3.7773680865747614e-07,
+      "loss": 0.0946,
+      "step": 9272
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3971028010300925,
+      "learning_rate": 3.776366724695844e-07,
+      "loss": 0.3012,
+      "step": 9273
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.669258071981314,
+      "learning_rate": 3.775365415019581e-07,
+      "loss": 0.1948,
+      "step": 9274
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8696804014722496,
+      "learning_rate": 3.77436415758869e-07,
+      "loss": 0.0936,
+      "step": 9275
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9205455098351596,
+      "learning_rate": 3.773362952445889e-07,
+      "loss": 0.1226,
+      "step": 9276
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7484490139593507,
+      "learning_rate": 3.77236179963389e-07,
+      "loss": 0.2307,
+      "step": 9277
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7675191985092316,
+      "learning_rate": 3.771360699195404e-07,
+      "loss": 0.088,
+      "step": 9278
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.48411160950391235,
+      "learning_rate": 3.770359651173139e-07,
+      "loss": 0.1327,
+      "step": 9279
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.966301264369236,
+      "learning_rate": 3.769358655609807e-07,
+      "loss": 0.1546,
+      "step": 9280
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6779362212885075,
+      "learning_rate": 3.7683577125481053e-07,
+      "loss": 0.2624,
+      "step": 9281
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7683336809436724,
+      "learning_rate": 3.7673568220307417e-07,
+      "loss": 0.0865,
+      "step": 9282
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.36820261982633506,
+      "learning_rate": 3.766355984100412e-07,
+      "loss": 0.0808,
+      "step": 9283
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6542238951265417,
+      "learning_rate": 3.7653551987998185e-07,
+      "loss": 0.0406,
+      "step": 9284
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2466414346265247,
+      "learning_rate": 3.7643544661716514e-07,
+      "loss": 0.1965,
+      "step": 9285
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9634669945912673,
+      "learning_rate": 3.7633537862586085e-07,
+      "loss": 0.5376,
+      "step": 9286
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.43694322912159,
+      "learning_rate": 3.7623531591033795e-07,
+      "loss": 0.1466,
+      "step": 9287
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.7176798163861555,
+      "learning_rate": 3.76135258474865e-07,
+      "loss": 0.1206,
+      "step": 9288
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3216627653953292,
+      "learning_rate": 3.7603520632371127e-07,
+      "loss": 0.0452,
+      "step": 9289
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 20.910602243800536,
+      "learning_rate": 3.7593515946114457e-07,
+      "loss": 0.2689,
+      "step": 9290
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3322446981226637,
+      "learning_rate": 3.758351178914336e-07,
+      "loss": 0.2586,
+      "step": 9291
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.564679148620845,
+      "learning_rate": 3.7573508161884587e-07,
+      "loss": 0.0327,
+      "step": 9292
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7216668245555732,
+      "learning_rate": 3.7563505064764953e-07,
+      "loss": 0.2728,
+      "step": 9293
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9364007422457261,
+      "learning_rate": 3.755350249821118e-07,
+      "loss": 0.2889,
+      "step": 9294
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5974168987425299,
+      "learning_rate": 3.754350046265002e-07,
+      "loss": 0.1377,
+      "step": 9295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6638751878720911,
+      "learning_rate": 3.753349895850817e-07,
+      "loss": 0.2111,
+      "step": 9296
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4677716160403327,
+      "learning_rate": 3.7523497986212317e-07,
+      "loss": 0.0553,
+      "step": 9297
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1163588469681387,
+      "learning_rate": 3.7513497546189103e-07,
+      "loss": 0.1391,
+      "step": 9298
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3841680899630626,
+      "learning_rate": 3.750349763886521e-07,
+      "loss": 0.4298,
+      "step": 9299
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.948411230163331,
+      "learning_rate": 3.7493498264667234e-07,
+      "loss": 0.0064,
+      "step": 9300
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6125477209147805,
+      "learning_rate": 3.7483499424021737e-07,
+      "loss": 0.1465,
+      "step": 9301
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9717487418475622,
+      "learning_rate": 3.747350111735533e-07,
+      "loss": 0.2071,
+      "step": 9302
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6911711802058582,
+      "learning_rate": 3.7463503345094537e-07,
+      "loss": 0.246,
+      "step": 9303
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6588841672053187,
+      "learning_rate": 3.74535061076659e-07,
+      "loss": 0.0713,
+      "step": 9304
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1389535575238425,
+      "learning_rate": 3.7443509405495905e-07,
+      "loss": 0.1852,
+      "step": 9305
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6035440069910316,
+      "learning_rate": 3.7433513239011063e-07,
+      "loss": 0.2771,
+      "step": 9306
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8692971748620172,
+      "learning_rate": 3.742351760863778e-07,
+      "loss": 0.2165,
+      "step": 9307
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8784110511688423,
+      "learning_rate": 3.7413522514802544e-07,
+      "loss": 0.1849,
+      "step": 9308
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0342959203137772,
+      "learning_rate": 3.7403527957931713e-07,
+      "loss": 0.1187,
+      "step": 9309
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.35069950752885065,
+      "learning_rate": 3.7393533938451733e-07,
+      "loss": 0.0126,
+      "step": 9310
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8453374061680935,
+      "learning_rate": 3.738354045678891e-07,
+      "loss": 0.2379,
+      "step": 9311
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.790934286897221,
+      "learning_rate": 3.737354751336963e-07,
+      "loss": 0.195,
+      "step": 9312
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.04097460610923,
+      "learning_rate": 3.736355510862018e-07,
+      "loss": 0.0642,
+      "step": 9313
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9689038288395968,
+      "learning_rate": 3.7353563242966877e-07,
+      "loss": 0.1235,
+      "step": 9314
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9070018954853412,
+      "learning_rate": 3.7343571916836005e-07,
+      "loss": 0.1238,
+      "step": 9315
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5193318612328358,
+      "learning_rate": 3.733358113065378e-07,
+      "loss": 0.1348,
+      "step": 9316
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.589835579930599,
+      "learning_rate": 3.7323590884846454e-07,
+      "loss": 0.2745,
+      "step": 9317
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8207545384687512,
+      "learning_rate": 3.7313601179840214e-07,
+      "loss": 0.1382,
+      "step": 9318
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7543516932605283,
+      "learning_rate": 3.730361201606127e-07,
+      "loss": 0.2343,
+      "step": 9319
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.319509783700083,
+      "learning_rate": 3.7293623393935736e-07,
+      "loss": 0.2488,
+      "step": 9320
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9649618460249436,
+      "learning_rate": 3.7283635313889784e-07,
+      "loss": 0.2595,
+      "step": 9321
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8895959725612803,
+      "learning_rate": 3.7273647776349514e-07,
+      "loss": 0.1304,
+      "step": 9322
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6724856053530631,
+      "learning_rate": 3.726366078174101e-07,
+      "loss": 0.2315,
+      "step": 9323
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.494764728635633,
+      "learning_rate": 3.7253674330490324e-07,
+      "loss": 0.1345,
+      "step": 9324
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1241400815157059,
+      "learning_rate": 3.7243688423023546e-07,
+      "loss": 0.3665,
+      "step": 9325
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7634027749162353,
+      "learning_rate": 3.7233703059766643e-07,
+      "loss": 0.1472,
+      "step": 9326
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.35743247768133973,
+      "learning_rate": 3.722371824114564e-07,
+      "loss": 0.0067,
+      "step": 9327
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.19256407345676066,
+      "learning_rate": 3.7213733967586514e-07,
+      "loss": 0.1221,
+      "step": 9328
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8973184367805298,
+      "learning_rate": 3.720375023951517e-07,
+      "loss": 0.1271,
+      "step": 9329
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8442009928931453,
+      "learning_rate": 3.7193767057357577e-07,
+      "loss": 0.1489,
+      "step": 9330
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1138760101326204,
+      "learning_rate": 3.718378442153962e-07,
+      "loss": 0.0028,
+      "step": 9331
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8201737035455592,
+      "learning_rate": 3.7173802332487196e-07,
+      "loss": 0.1708,
+      "step": 9332
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4636916227535591,
+      "learning_rate": 3.716382079062613e-07,
+      "loss": 0.1483,
+      "step": 9333
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8163774505063767,
+      "learning_rate": 3.71538397963823e-07,
+      "loss": 0.352,
+      "step": 9334
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.40018528842627643,
+      "learning_rate": 3.7143859350181464e-07,
+      "loss": 0.2831,
+      "step": 9335
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.785518886776523,
+      "learning_rate": 3.7133879452449446e-07,
+      "loss": 0.4279,
+      "step": 9336
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7049248450393616,
+      "learning_rate": 3.712390010361198e-07,
+      "loss": 0.1501,
+      "step": 9337
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7123666464227902,
+      "learning_rate": 3.711392130409484e-07,
+      "loss": 0.2971,
+      "step": 9338
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.368363434577095,
+      "learning_rate": 3.710394305432371e-07,
+      "loss": 0.2618,
+      "step": 9339
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8423457662773516,
+      "learning_rate": 3.7093965354724286e-07,
+      "loss": 0.1768,
+      "step": 9340
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1982493547049378,
+      "learning_rate": 3.708398820572225e-07,
+      "loss": 0.0443,
+      "step": 9341
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8572009384772475,
+      "learning_rate": 3.707401160774325e-07,
+      "loss": 0.074,
+      "step": 9342
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.715077017393968,
+      "learning_rate": 3.70640355612129e-07,
+      "loss": 0.1425,
+      "step": 9343
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8019019532902004,
+      "learning_rate": 3.7054060066556793e-07,
+      "loss": 0.2665,
+      "step": 9344
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.791625294891656,
+      "learning_rate": 3.7044085124200514e-07,
+      "loss": 0.087,
+      "step": 9345
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.747912087089999,
+      "learning_rate": 3.703411073456959e-07,
+      "loss": 0.3365,
+      "step": 9346
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6407920231572916,
+      "learning_rate": 3.7024136898089576e-07,
+      "loss": 0.3517,
+      "step": 9347
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 15.630045302440479,
+      "learning_rate": 3.701416361518597e-07,
+      "loss": 0.0223,
+      "step": 9348
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3720280989672002,
+      "learning_rate": 3.700419088628425e-07,
+      "loss": 0.3173,
+      "step": 9349
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2971003551245055,
+      "learning_rate": 3.6994218711809856e-07,
+      "loss": 0.097,
+      "step": 9350
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.48855604220375287,
+      "learning_rate": 3.698424709218826e-07,
+      "loss": 0.2794,
+      "step": 9351
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.434706071754007,
+      "learning_rate": 3.697427602784483e-07,
+      "loss": 0.2833,
+      "step": 9352
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7130226436187235,
+      "learning_rate": 3.696430551920499e-07,
+      "loss": 0.1784,
+      "step": 9353
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6419149544861212,
+      "learning_rate": 3.6954335566694057e-07,
+      "loss": 0.2665,
+      "step": 9354
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7860592936408222,
+      "learning_rate": 3.6944366170737415e-07,
+      "loss": 0.1419,
+      "step": 9355
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3870030044112588,
+      "learning_rate": 3.693439733176035e-07,
+      "loss": 0.1992,
+      "step": 9356
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4643972104686778,
+      "learning_rate": 3.6924429050188157e-07,
+      "loss": 0.1586,
+      "step": 9357
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.778282250295543,
+      "learning_rate": 3.6914461326446104e-07,
+      "loss": 0.1099,
+      "step": 9358
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0465978772358504,
+      "learning_rate": 3.690449416095943e-07,
+      "loss": 0.1571,
+      "step": 9359
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5731659928726092,
+      "learning_rate": 3.6894527554153383e-07,
+      "loss": 0.3195,
+      "step": 9360
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2875317520675376,
+      "learning_rate": 3.688456150645311e-07,
+      "loss": 0.2305,
+      "step": 9361
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.029337424805925,
+      "learning_rate": 3.6874596018283833e-07,
+      "loss": 0.088,
+      "step": 9362
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8809815467086525,
+      "learning_rate": 3.6864631090070653e-07,
+      "loss": 0.3054,
+      "step": 9363
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.47611715679060396,
+      "learning_rate": 3.6854666722238737e-07,
+      "loss": 0.1209,
+      "step": 9364
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6557302681228317,
+      "learning_rate": 3.684470291521314e-07,
+      "loss": 0.0543,
+      "step": 9365
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.31106441265079243,
+      "learning_rate": 3.683473966941898e-07,
+      "loss": 0.0051,
+      "step": 9366
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.020456985395643,
+      "learning_rate": 3.682477698528128e-07,
+      "loss": 0.1233,
+      "step": 9367
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6984927839984826,
+      "learning_rate": 3.6814814863225083e-07,
+      "loss": 0.0894,
+      "step": 9368
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0236622478125843,
+      "learning_rate": 3.6804853303675407e-07,
+      "loss": 0.0866,
+      "step": 9369
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9520519063215946,
+      "learning_rate": 3.679489230705719e-07,
+      "loss": 0.1956,
+      "step": 9370
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9785352268345708,
+      "learning_rate": 3.6784931873795434e-07,
+      "loss": 0.2661,
+      "step": 9371
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5489439806777325,
+      "learning_rate": 3.677497200431503e-07,
+      "loss": 0.2252,
+      "step": 9372
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1660682395226845,
+      "learning_rate": 3.676501269904093e-07,
+      "loss": 0.2097,
+      "step": 9373
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9117422336753886,
+      "learning_rate": 3.6755053958397964e-07,
+      "loss": 0.1649,
+      "step": 9374
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3174871712267413,
+      "learning_rate": 3.6745095782811034e-07,
+      "loss": 0.1165,
+      "step": 9375
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1737490207513126,
+      "learning_rate": 3.673513817270496e-07,
+      "loss": 0.3228,
+      "step": 9376
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3779584403245971,
+      "learning_rate": 3.6725181128504566e-07,
+      "loss": 0.0662,
+      "step": 9377
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9040176754693658,
+      "learning_rate": 3.6715224650634623e-07,
+      "loss": 0.3983,
+      "step": 9378
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0949680380558715,
+      "learning_rate": 3.6705268739519916e-07,
+      "loss": 0.2969,
+      "step": 9379
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5585126973068211,
+      "learning_rate": 3.669531339558515e-07,
+      "loss": 0.0869,
+      "step": 9380
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8409218650921324,
+      "learning_rate": 3.668535861925509e-07,
+      "loss": 0.2927,
+      "step": 9381
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.401990657290293,
+      "learning_rate": 3.667540441095436e-07,
+      "loss": 0.0338,
+      "step": 9382
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.27809309699568374,
+      "learning_rate": 3.6665450771107697e-07,
+      "loss": 0.1003,
+      "step": 9383
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7875044896588642,
+      "learning_rate": 3.6655497700139693e-07,
+      "loss": 0.1111,
+      "step": 9384
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3888928888215062,
+      "learning_rate": 3.6645545198474973e-07,
+      "loss": 0.1958,
+      "step": 9385
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.46544895086304044,
+      "learning_rate": 3.6635593266538167e-07,
+      "loss": 0.2682,
+      "step": 9386
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5980123759044917,
+      "learning_rate": 3.662564190475379e-07,
+      "loss": 0.1744,
+      "step": 9387
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.738477076975794,
+      "learning_rate": 3.6615691113546445e-07,
+      "loss": 0.0572,
+      "step": 9388
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4997082116318408,
+      "learning_rate": 3.6605740893340596e-07,
+      "loss": 0.2915,
+      "step": 9389
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0518297815423197,
+      "learning_rate": 3.659579124456079e-07,
+      "loss": 0.2468,
+      "step": 9390
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8886369436167223,
+      "learning_rate": 3.6585842167631453e-07,
+      "loss": 0.3167,
+      "step": 9391
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7653665718881084,
+      "learning_rate": 3.6575893662977066e-07,
+      "loss": 0.171,
+      "step": 9392
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3841488873652306,
+      "learning_rate": 3.6565945731022036e-07,
+      "loss": 0.1029,
+      "step": 9393
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8641759029425111,
+      "learning_rate": 3.655599837219077e-07,
+      "loss": 0.2416,
+      "step": 9394
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9428812489335693,
+      "learning_rate": 3.6546051586907635e-07,
+      "loss": 0.1067,
+      "step": 9395
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3502908802967608,
+      "learning_rate": 3.6536105375596996e-07,
+      "loss": 0.0645,
+      "step": 9396
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8358373455391138,
+      "learning_rate": 3.652615973868317e-07,
+      "loss": 0.0722,
+      "step": 9397
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6973503012289795,
+      "learning_rate": 3.651621467659044e-07,
+      "loss": 0.1672,
+      "step": 9398
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.926952821654789,
+      "learning_rate": 3.6506270189743116e-07,
+      "loss": 0.3482,
+      "step": 9399
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0162155844066454,
+      "learning_rate": 3.649632627856541e-07,
+      "loss": 0.3422,
+      "step": 9400
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.170829378895211,
+      "learning_rate": 3.648638294348158e-07,
+      "loss": 0.0836,
+      "step": 9401
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.33666530053854105,
+      "learning_rate": 3.647644018491581e-07,
+      "loss": 0.1849,
+      "step": 9402
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.195738706653307,
+      "learning_rate": 3.646649800329231e-07,
+      "loss": 0.2266,
+      "step": 9403
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7221616589187154,
+      "learning_rate": 3.6456556399035186e-07,
+      "loss": 0.0995,
+      "step": 9404
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5300694572259874,
+      "learning_rate": 3.6446615372568616e-07,
+      "loss": 0.181,
+      "step": 9405
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 14.89805046405319,
+      "learning_rate": 3.643667492431666e-07,
+      "loss": 0.0775,
+      "step": 9406
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.38542962638776146,
+      "learning_rate": 3.642673505470344e-07,
+      "loss": 0.0659,
+      "step": 9407
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6983717939905597,
+      "learning_rate": 3.6416795764152964e-07,
+      "loss": 0.2694,
+      "step": 9408
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.47082656016469765,
+      "learning_rate": 3.64068570530893e-07,
+      "loss": 0.1048,
+      "step": 9409
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9109312430517122,
+      "learning_rate": 3.639691892193644e-07,
+      "loss": 0.1162,
+      "step": 9410
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6899362271898326,
+      "learning_rate": 3.638698137111835e-07,
+      "loss": 0.1932,
+      "step": 9411
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8027858842048811,
+      "learning_rate": 3.637704440105902e-07,
+      "loss": 0.1846,
+      "step": 9412
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.739760075312104,
+      "learning_rate": 3.636710801218235e-07,
+      "loss": 0.1182,
+      "step": 9413
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5382520406612367,
+      "learning_rate": 3.635717220491227e-07,
+      "loss": 0.2819,
+      "step": 9414
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0858281035483512,
+      "learning_rate": 3.6347236979672626e-07,
+      "loss": 0.3465,
+      "step": 9415
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9059460894166045,
+      "learning_rate": 3.633730233688732e-07,
+      "loss": 0.1408,
+      "step": 9416
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7336907181010045,
+      "learning_rate": 3.6327368276980145e-07,
+      "loss": 0.0576,
+      "step": 9417
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1336513371755648,
+      "learning_rate": 3.631743480037495e-07,
+      "loss": 0.2492,
+      "step": 9418
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4045115547659766,
+      "learning_rate": 3.630750190749546e-07,
+      "loss": 0.2621,
+      "step": 9419
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.44199810016468766,
+      "learning_rate": 3.6297569598765475e-07,
+      "loss": 0.0086,
+      "step": 9420
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2029416627921536,
+      "learning_rate": 3.6287637874608713e-07,
+      "loss": 0.3133,
+      "step": 9421
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1686022841441073,
+      "learning_rate": 3.6277706735448903e-07,
+      "loss": 0.2998,
+      "step": 9422
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.519242917613396,
+      "learning_rate": 3.6267776181709684e-07,
+      "loss": 0.3648,
+      "step": 9423
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8015422825334999,
+      "learning_rate": 3.6257846213814767e-07,
+      "loss": 0.0705,
+      "step": 9424
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5074650534128662,
+      "learning_rate": 3.6247916832187755e-07,
+      "loss": 0.0888,
+      "step": 9425
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4409350840386246,
+      "learning_rate": 3.6237988037252227e-07,
+      "loss": 0.3621,
+      "step": 9426
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.588334078053489,
+      "learning_rate": 3.622805982943181e-07,
+      "loss": 0.0433,
+      "step": 9427
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.690879103346311,
+      "learning_rate": 3.621813220915004e-07,
+      "loss": 0.2864,
+      "step": 9428
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.641539238707806,
+      "learning_rate": 3.620820517683046e-07,
+      "loss": 0.094,
+      "step": 9429
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.49798299721738,
+      "learning_rate": 3.619827873289657e-07,
+      "loss": 0.3174,
+      "step": 9430
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6268779811381258,
+      "learning_rate": 3.6188352877771865e-07,
+      "loss": 0.1791,
+      "step": 9431
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4414730057712469,
+      "learning_rate": 3.617842761187977e-07,
+      "loss": 0.3484,
+      "step": 9432
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.086231279791825,
+      "learning_rate": 3.6168502935643763e-07,
+      "loss": 0.019,
+      "step": 9433
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4089373888040658,
+      "learning_rate": 3.61585788494872e-07,
+      "loss": 0.0678,
+      "step": 9434
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4732372969873673,
+      "learning_rate": 3.614865535383351e-07,
+      "loss": 0.1703,
+      "step": 9435
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5186388228339889,
+      "learning_rate": 3.613873244910601e-07,
+      "loss": 0.2791,
+      "step": 9436
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9842137879472515,
+      "learning_rate": 3.612881013572806e-07,
+      "loss": 0.1501,
+      "step": 9437
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5935332181125628,
+      "learning_rate": 3.611888841412296e-07,
+      "loss": 0.2552,
+      "step": 9438
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.787286726230883,
+      "learning_rate": 3.6108967284713966e-07,
+      "loss": 0.193,
+      "step": 9439
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4755215820683798,
+      "learning_rate": 3.6099046747924373e-07,
+      "loss": 0.1659,
+      "step": 9440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6750708548489052,
+      "learning_rate": 3.6089126804177364e-07,
+      "loss": 0.1033,
+      "step": 9441
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7863623723317774,
+      "learning_rate": 3.60792074538962e-07,
+      "loss": 0.1735,
+      "step": 9442
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 16.245197149121086,
+      "learning_rate": 3.606928869750401e-07,
+      "loss": 0.2154,
+      "step": 9443
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6445252825729044,
+      "learning_rate": 3.605937053542398e-07,
+      "loss": 0.3043,
+      "step": 9444
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9737129284554178,
+      "learning_rate": 3.604945296807921e-07,
+      "loss": 0.3247,
+      "step": 9445
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3940290771510294,
+      "learning_rate": 3.6039535995892835e-07,
+      "loss": 0.0489,
+      "step": 9446
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.610151977789657,
+      "learning_rate": 3.6029619619287897e-07,
+      "loss": 0.0484,
+      "step": 9447
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3523942494620698,
+      "learning_rate": 3.60197038386875e-07,
+      "loss": 0.2164,
+      "step": 9448
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4857908866675192,
+      "learning_rate": 3.600978865451462e-07,
+      "loss": 0.0688,
+      "step": 9449
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7683698264502922,
+      "learning_rate": 3.59998740671923e-07,
+      "loss": 0.1496,
+      "step": 9450
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6713109763358255,
+      "learning_rate": 3.5989960077143475e-07,
+      "loss": 0.141,
+      "step": 9451
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.39428184273653033,
+      "learning_rate": 3.598004668479113e-07,
+      "loss": 0.1295,
+      "step": 9452
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6878095952961367,
+      "learning_rate": 3.597013389055818e-07,
+      "loss": 0.0199,
+      "step": 9453
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8524581325660707,
+      "learning_rate": 3.5960221694867516e-07,
+      "loss": 0.3448,
+      "step": 9454
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5628398563066372,
+      "learning_rate": 3.5950310098142016e-07,
+      "loss": 0.2538,
+      "step": 9455
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1186249288055055,
+      "learning_rate": 3.594039910080452e-07,
+      "loss": 0.0979,
+      "step": 9456
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5744758314234945,
+      "learning_rate": 3.5930488703277884e-07,
+      "loss": 0.3416,
+      "step": 9457
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.990597953243976,
+      "learning_rate": 3.5920578905984867e-07,
+      "loss": 0.284,
+      "step": 9458
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5819561244282796,
+      "learning_rate": 3.591066970934827e-07,
+      "loss": 0.1504,
+      "step": 9459
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3169886141667242,
+      "learning_rate": 3.5900761113790813e-07,
+      "loss": 0.255,
+      "step": 9460
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.165269825801838,
+      "learning_rate": 3.589085311973524e-07,
+      "loss": 0.3182,
+      "step": 9461
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.824506802886921,
+      "learning_rate": 3.588094572760423e-07,
+      "loss": 0.2083,
+      "step": 9462
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6393293552065233,
+      "learning_rate": 3.587103893782046e-07,
+      "loss": 0.0955,
+      "step": 9463
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1996395451813144,
+      "learning_rate": 3.5861132750806566e-07,
+      "loss": 0.4389,
+      "step": 9464
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.61484900781638,
+      "learning_rate": 3.5851227166985184e-07,
+      "loss": 0.2825,
+      "step": 9465
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.172490924953922,
+      "learning_rate": 3.5841322186778894e-07,
+      "loss": 0.1773,
+      "step": 9466
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.638159471201789,
+      "learning_rate": 3.583141781061024e-07,
+      "loss": 0.1132,
+      "step": 9467
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9351782254085814,
+      "learning_rate": 3.582151403890182e-07,
+      "loss": 0.2241,
+      "step": 9468
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9256261683344589,
+      "learning_rate": 3.581161087207608e-07,
+      "loss": 0.2268,
+      "step": 9469
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3051866291544902,
+      "learning_rate": 3.580170831055557e-07,
+      "loss": 0.2202,
+      "step": 9470
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2190511644389843,
+      "learning_rate": 3.57918063547627e-07,
+      "loss": 0.0569,
+      "step": 9471
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.442025476340937,
+      "learning_rate": 3.578190500511994e-07,
+      "loss": 0.1541,
+      "step": 9472
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8236563443916848,
+      "learning_rate": 3.5772004262049705e-07,
+      "loss": 0.3391,
+      "step": 9473
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.006509833273857,
+      "learning_rate": 3.5762104125974357e-07,
+      "loss": 0.1443,
+      "step": 9474
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.159114976755673,
+      "learning_rate": 3.575220459731627e-07,
+      "loss": 0.1633,
+      "step": 9475
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4073080062571323,
+      "learning_rate": 3.5742305676497785e-07,
+      "loss": 0.0046,
+      "step": 9476
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9576682490917284,
+      "learning_rate": 3.573240736394119e-07,
+      "loss": 0.3445,
+      "step": 9477
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.39686532945091335,
+      "learning_rate": 3.57225096600688e-07,
+      "loss": 0.0589,
+      "step": 9478
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8499172182669122,
+      "learning_rate": 3.5712612565302846e-07,
+      "loss": 0.183,
+      "step": 9479
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0256610774157449,
+      "learning_rate": 3.5702716080065544e-07,
+      "loss": 0.4106,
+      "step": 9480
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6932690415855755,
+      "learning_rate": 3.569282020477912e-07,
+      "loss": 0.115,
+      "step": 9481
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6391868470132124,
+      "learning_rate": 3.568292493986574e-07,
+      "loss": 0.2923,
+      "step": 9482
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0553878712678806,
+      "learning_rate": 3.5673030285747583e-07,
+      "loss": 0.126,
+      "step": 9483
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.21637608116406298,
+      "learning_rate": 3.566313624284674e-07,
+      "loss": 0.1065,
+      "step": 9484
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.875493412863062,
+      "learning_rate": 3.565324281158534e-07,
+      "loss": 0.158,
+      "step": 9485
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2088440849044081,
+      "learning_rate": 3.564334999238542e-07,
+      "loss": 0.0567,
+      "step": 9486
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6064417453901161,
+      "learning_rate": 3.563345778566907e-07,
+      "loss": 0.0139,
+      "step": 9487
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6574714732114035,
+      "learning_rate": 3.562356619185827e-07,
+      "loss": 0.2009,
+      "step": 9488
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.155984125520488,
+      "learning_rate": 3.5613675211375063e-07,
+      "loss": 0.1546,
+      "step": 9489
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5766695840302901,
+      "learning_rate": 3.560378484464137e-07,
+      "loss": 0.3639,
+      "step": 9490
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.48979977542295144,
+      "learning_rate": 3.559389509207916e-07,
+      "loss": 0.1652,
+      "step": 9491
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.493927423119219,
+      "learning_rate": 3.5584005954110343e-07,
+      "loss": 0.1472,
+      "step": 9492
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7184802566700499,
+      "learning_rate": 3.5574117431156826e-07,
+      "loss": 0.3289,
+      "step": 9493
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5233670046628419,
+      "learning_rate": 3.5564229523640466e-07,
+      "loss": 0.1573,
+      "step": 9494
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5534037970393029,
+      "learning_rate": 3.5554342231983063e-07,
+      "loss": 0.3675,
+      "step": 9495
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.866844066999668,
+      "learning_rate": 3.554445555660649e-07,
+      "loss": 0.0769,
+      "step": 9496
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7776888752887263,
+      "learning_rate": 3.5534569497932474e-07,
+      "loss": 0.2311,
+      "step": 9497
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7046551376374321,
+      "learning_rate": 3.552468405638282e-07,
+      "loss": 0.2205,
+      "step": 9498
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6692309773361224,
+      "learning_rate": 3.551479923237924e-07,
+      "loss": 0.1334,
+      "step": 9499
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.104945431067929,
+      "learning_rate": 3.5504915026343443e-07,
+      "loss": 0.071,
+      "step": 9500
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5064598508166142,
+      "learning_rate": 3.54950314386971e-07,
+      "loss": 0.3589,
+      "step": 9501
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8520594832192026,
+      "learning_rate": 3.548514846986189e-07,
+      "loss": 0.1262,
+      "step": 9502
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 15.362113891260808,
+      "learning_rate": 3.547526612025942e-07,
+      "loss": 0.3452,
+      "step": 9503
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9419430460431464,
+      "learning_rate": 3.5465384390311297e-07,
+      "loss": 0.447,
+      "step": 9504
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2961654803612481,
+      "learning_rate": 3.545550328043908e-07,
+      "loss": 0.1917,
+      "step": 9505
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5315887724165193,
+      "learning_rate": 3.5445622791064357e-07,
+      "loss": 0.2115,
+      "step": 9506
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4770144668206306,
+      "learning_rate": 3.543574292260861e-07,
+      "loss": 0.0086,
+      "step": 9507
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2355838871005962,
+      "learning_rate": 3.542586367549334e-07,
+      "loss": 0.1475,
+      "step": 9508
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2483751771922211,
+      "learning_rate": 3.541598505014004e-07,
+      "loss": 0.256,
+      "step": 9509
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4285539010868225,
+      "learning_rate": 3.540610704697011e-07,
+      "loss": 0.2119,
+      "step": 9510
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.734726231846914,
+      "learning_rate": 3.5396229666405026e-07,
+      "loss": 0.2798,
+      "step": 9511
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.482766810153079,
+      "learning_rate": 3.538635290886611e-07,
+      "loss": 0.2323,
+      "step": 9512
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0432071428183933,
+      "learning_rate": 3.5376476774774776e-07,
+      "loss": 0.3549,
+      "step": 9513
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7664898071153935,
+      "learning_rate": 3.536660126455233e-07,
+      "loss": 0.0092,
+      "step": 9514
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6304981059941067,
+      "learning_rate": 3.5356726378620103e-07,
+      "loss": 0.005,
+      "step": 9515
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.38579326887007837,
+      "learning_rate": 3.5346852117399347e-07,
+      "loss": 0.0814,
+      "step": 9516
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3293810669383335,
+      "learning_rate": 3.533697848131134e-07,
+      "loss": 0.1155,
+      "step": 9517
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 14.950871791178397,
+      "learning_rate": 3.5327105470777305e-07,
+      "loss": 0.1445,
+      "step": 9518
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7977325969398998,
+      "learning_rate": 3.531723308621847e-07,
+      "loss": 0.1063,
+      "step": 9519
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3291692439433445,
+      "learning_rate": 3.5307361328055976e-07,
+      "loss": 0.1668,
+      "step": 9520
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.029977035286095,
+      "learning_rate": 3.529749019671097e-07,
+      "loss": 0.1566,
+      "step": 9521
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.759593639531423,
+      "learning_rate": 3.5287619692604607e-07,
+      "loss": 0.1452,
+      "step": 9522
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.215541996792754,
+      "learning_rate": 3.527774981615794e-07,
+      "loss": 0.1516,
+      "step": 9523
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9919919160786849,
+      "learning_rate": 3.526788056779208e-07,
+      "loss": 0.1444,
+      "step": 9524
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6100328855461785,
+      "learning_rate": 3.5258011947928047e-07,
+      "loss": 0.1438,
+      "step": 9525
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6662172389324932,
+      "learning_rate": 3.524814395698686e-07,
+      "loss": 0.1544,
+      "step": 9526
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8541939750322884,
+      "learning_rate": 3.5238276595389495e-07,
+      "loss": 0.2717,
+      "step": 9527
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9632005762172113,
+      "learning_rate": 3.522840986355694e-07,
+      "loss": 0.2894,
+      "step": 9528
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.787435734934215,
+      "learning_rate": 3.5218543761910104e-07,
+      "loss": 0.0897,
+      "step": 9529
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33808649369160954,
+      "learning_rate": 3.520867829086992e-07,
+      "loss": 0.1316,
+      "step": 9530
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5051858975240087,
+      "learning_rate": 3.519881345085723e-07,
+      "loss": 0.0644,
+      "step": 9531
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5099164061138237,
+      "learning_rate": 3.5188949242292945e-07,
+      "loss": 0.2252,
+      "step": 9532
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8753734344434103,
+      "learning_rate": 3.517908566559783e-07,
+      "loss": 0.1211,
+      "step": 9533
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7426447517918439,
+      "learning_rate": 3.5169222721192735e-07,
+      "loss": 0.4824,
+      "step": 9534
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0894601162545856,
+      "learning_rate": 3.515936040949841e-07,
+      "loss": 0.0962,
+      "step": 9535
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8256599029554128,
+      "learning_rate": 3.514949873093559e-07,
+      "loss": 0.0358,
+      "step": 9536
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.405568848617357,
+      "learning_rate": 3.513963768592502e-07,
+      "loss": 0.1053,
+      "step": 9537
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7117838901206692,
+      "learning_rate": 3.5129777274887363e-07,
+      "loss": 0.0553,
+      "step": 9538
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8556497796650047,
+      "learning_rate": 3.511991749824332e-07,
+      "loss": 0.1805,
+      "step": 9539
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9670599418896123,
+      "learning_rate": 3.5110058356413497e-07,
+      "loss": 0.3328,
+      "step": 9540
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2557616840936097,
+      "learning_rate": 3.510019984981853e-07,
+      "loss": 0.016,
+      "step": 9541
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.192287509787701,
+      "learning_rate": 3.509034197887897e-07,
+      "loss": 0.3793,
+      "step": 9542
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5618701902969143,
+      "learning_rate": 3.5080484744015405e-07,
+      "loss": 0.1652,
+      "step": 9543
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4294697667042685,
+      "learning_rate": 3.5070628145648353e-07,
+      "loss": 0.1548,
+      "step": 9544
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2650894762362481,
+      "learning_rate": 3.5060772184198313e-07,
+      "loss": 0.0619,
+      "step": 9545
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2826780412617805,
+      "learning_rate": 3.5050916860085757e-07,
+      "loss": 0.1783,
+      "step": 9546
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.48055308768303967,
+      "learning_rate": 3.5041062173731153e-07,
+      "loss": 0.0374,
+      "step": 9547
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.26875555863780737,
+      "learning_rate": 3.503120812555491e-07,
+      "loss": 0.1087,
+      "step": 9548
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7210446372994427,
+      "learning_rate": 3.5021354715977404e-07,
+      "loss": 0.2469,
+      "step": 9549
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9612550213102113,
+      "learning_rate": 3.5011501945419034e-07,
+      "loss": 0.2909,
+      "step": 9550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3144814339545225,
+      "learning_rate": 3.5001649814300103e-07,
+      "loss": 0.3016,
+      "step": 9551
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5657734398150366,
+      "learning_rate": 3.4991798323040957e-07,
+      "loss": 0.0682,
+      "step": 9552
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7252407911586491,
+      "learning_rate": 3.4981947472061846e-07,
+      "loss": 0.2639,
+      "step": 9553
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6242112103120891,
+      "learning_rate": 3.497209726178306e-07,
+      "loss": 0.0052,
+      "step": 9554
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2664016430341551,
+      "learning_rate": 3.4962247692624806e-07,
+      "loss": 0.1158,
+      "step": 9555
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3797610985119093,
+      "learning_rate": 3.495239876500732e-07,
+      "loss": 0.1901,
+      "step": 9556
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0646627807982256,
+      "learning_rate": 3.494255047935072e-07,
+      "loss": 0.264,
+      "step": 9557
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8334236528510394,
+      "learning_rate": 3.493270283607521e-07,
+      "loss": 0.3143,
+      "step": 9558
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6066992204145405,
+      "learning_rate": 3.492285583560086e-07,
+      "loss": 0.2633,
+      "step": 9559
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5321456602640795,
+      "learning_rate": 3.4913009478347824e-07,
+      "loss": 0.1904,
+      "step": 9560
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7261091688805966,
+      "learning_rate": 3.49031637647361e-07,
+      "loss": 0.4801,
+      "step": 9561
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.033724474706224,
+      "learning_rate": 3.4893318695185767e-07,
+      "loss": 0.4179,
+      "step": 9562
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.075835070526746,
+      "learning_rate": 3.488347427011684e-07,
+      "loss": 0.3301,
+      "step": 9563
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.502068478136284,
+      "learning_rate": 3.4873630489949274e-07,
+      "loss": 0.0192,
+      "step": 9564
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.395113070811715,
+      "learning_rate": 3.486378735510306e-07,
+      "loss": 0.0063,
+      "step": 9565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7232580462657048,
+      "learning_rate": 3.4853944865998073e-07,
+      "loss": 0.2881,
+      "step": 9566
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9599165312685054,
+      "learning_rate": 3.484410302305427e-07,
+      "loss": 0.3233,
+      "step": 9567
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4703181695866034,
+      "learning_rate": 3.483426182669149e-07,
+      "loss": 0.1889,
+      "step": 9568
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.146495240531375,
+      "learning_rate": 3.4824421277329585e-07,
+      "loss": 0.4046,
+      "step": 9569
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4314771979829906,
+      "learning_rate": 3.481458137538838e-07,
+      "loss": 0.1848,
+      "step": 9570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9250065699509874,
+      "learning_rate": 3.4804742121287654e-07,
+      "loss": 0.4214,
+      "step": 9571
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3293318691339513,
+      "learning_rate": 3.4794903515447174e-07,
+      "loss": 0.188,
+      "step": 9572
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4910563428076659,
+      "learning_rate": 3.4785065558286696e-07,
+      "loss": 0.1863,
+      "step": 9573
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5969956253129487,
+      "learning_rate": 3.477522825022588e-07,
+      "loss": 0.0602,
+      "step": 9574
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3744545662743759,
+      "learning_rate": 3.476539159168446e-07,
+      "loss": 0.0669,
+      "step": 9575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.734299639347295,
+      "learning_rate": 3.475555558308205e-07,
+      "loss": 0.4074,
+      "step": 9576
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3504725250091854,
+      "learning_rate": 3.4745720224838275e-07,
+      "loss": 0.424,
+      "step": 9577
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4331875068739966,
+      "learning_rate": 3.4735885517372745e-07,
+      "loss": 0.0128,
+      "step": 9578
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.42824816220074896,
+      "learning_rate": 3.472605146110501e-07,
+      "loss": 0.153,
+      "step": 9579
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4351001430071234,
+      "learning_rate": 3.4716218056454647e-07,
+      "loss": 0.1793,
+      "step": 9580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4280842403548806,
+      "learning_rate": 3.4706385303841134e-07,
+      "loss": 0.1541,
+      "step": 9581
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5926766013134903,
+      "learning_rate": 3.4696553203683983e-07,
+      "loss": 0.1084,
+      "step": 9582
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0985680596498748,
+      "learning_rate": 3.4686721756402616e-07,
+      "loss": 0.2146,
+      "step": 9583
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8880529549801257,
+      "learning_rate": 3.4676890962416496e-07,
+      "loss": 0.017,
+      "step": 9584
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6677483589894967,
+      "learning_rate": 3.4667060822145e-07,
+      "loss": 0.2069,
+      "step": 9585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0592907579725246,
+      "learning_rate": 3.4657231336007523e-07,
+      "loss": 0.13,
+      "step": 9586
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6592791785561625,
+      "learning_rate": 3.4647402504423385e-07,
+      "loss": 0.1032,
+      "step": 9587
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5331203369768155,
+      "learning_rate": 3.463757432781193e-07,
+      "loss": 0.2472,
+      "step": 9588
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.921419468819464,
+      "learning_rate": 3.4627746806592444e-07,
+      "loss": 0.3503,
+      "step": 9589
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5934001506174543,
+      "learning_rate": 3.4617919941184166e-07,
+      "loss": 0.4153,
+      "step": 9590
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9025315697185343,
+      "learning_rate": 3.460809373200636e-07,
+      "loss": 0.051,
+      "step": 9591
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9931361896209535,
+      "learning_rate": 3.4598268179478194e-07,
+      "loss": 0.279,
+      "step": 9592
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.088044584822752,
+      "learning_rate": 3.45884432840189e-07,
+      "loss": 0.059,
+      "step": 9593
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2314237640656722,
+      "learning_rate": 3.4578619046047567e-07,
+      "loss": 0.2023,
+      "step": 9594
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5092112323077743,
+      "learning_rate": 3.456879546598337e-07,
+      "loss": 0.093,
+      "step": 9595
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.128925925754039,
+      "learning_rate": 3.4558972544245356e-07,
+      "loss": 0.1333,
+      "step": 9596
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8043015206803905,
+      "learning_rate": 3.454915028125263e-07,
+      "loss": 0.2976,
+      "step": 9597
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3068952477633422,
+      "learning_rate": 3.45393286774242e-07,
+      "loss": 0.2758,
+      "step": 9598
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9730817044959121,
+      "learning_rate": 3.452950773317911e-07,
+      "loss": 0.3222,
+      "step": 9599
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.272755291092515,
+      "learning_rate": 3.4519687448936295e-07,
+      "loss": 0.2205,
+      "step": 9600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5868444706628881,
+      "learning_rate": 3.4509867825114755e-07,
+      "loss": 0.1644,
+      "step": 9601
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8330587807219971,
+      "learning_rate": 3.450004886213337e-07,
+      "loss": 0.0324,
+      "step": 9602
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0068560904339283,
+      "learning_rate": 3.4490230560411085e-07,
+      "loss": 0.1586,
+      "step": 9603
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0001005538099705,
+      "learning_rate": 3.448041292036673e-07,
+      "loss": 0.2305,
+      "step": 9604
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4916345275421797,
+      "learning_rate": 3.447059594241916e-07,
+      "loss": 0.0538,
+      "step": 9605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.618687922918267,
+      "learning_rate": 3.446077962698718e-07,
+      "loss": 0.2956,
+      "step": 9606
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5696702585567011,
+      "learning_rate": 3.445096397448958e-07,
+      "loss": 0.2709,
+      "step": 9607
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1884662622061173,
+      "learning_rate": 3.4441148985345123e-07,
+      "loss": 0.0745,
+      "step": 9608
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1563562383245019,
+      "learning_rate": 3.443133465997251e-07,
+      "loss": 0.3654,
+      "step": 9609
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2414528424868526,
+      "learning_rate": 3.442152099879048e-07,
+      "loss": 0.0804,
+      "step": 9610
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9512809743923842,
+      "learning_rate": 3.441170800221765e-07,
+      "loss": 0.1162,
+      "step": 9611
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0785385980380031,
+      "learning_rate": 3.440189567067273e-07,
+      "loss": 0.0659,
+      "step": 9612
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5797862836251841,
+      "learning_rate": 3.4392084004574275e-07,
+      "loss": 0.3305,
+      "step": 9613
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6063870396351476,
+      "learning_rate": 3.4382273004340887e-07,
+      "loss": 0.2354,
+      "step": 9614
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5974500860148275,
+      "learning_rate": 3.4372462670391144e-07,
+      "loss": 0.1743,
+      "step": 9615
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3077113945833219,
+      "learning_rate": 3.436265300314355e-07,
+      "loss": 0.1657,
+      "step": 9616
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.603166167192824,
+      "learning_rate": 3.4352844003016624e-07,
+      "loss": 0.2456,
+      "step": 9617
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2856105969290681,
+      "learning_rate": 3.434303567042881e-07,
+      "loss": 0.2376,
+      "step": 9618
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9429974027176391,
+      "learning_rate": 3.433322800579859e-07,
+      "loss": 0.2456,
+      "step": 9619
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.880197302547717,
+      "learning_rate": 3.432342100954434e-07,
+      "loss": 0.1117,
+      "step": 9620
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2472669673392983,
+      "learning_rate": 3.431361468208448e-07,
+      "loss": 0.0912,
+      "step": 9621
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4222150174216388,
+      "learning_rate": 3.4303809023837327e-07,
+      "loss": 0.2073,
+      "step": 9622
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1100839309970587,
+      "learning_rate": 3.4294004035221246e-07,
+      "loss": 0.0011,
+      "step": 9623
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.43061421143584283,
+      "learning_rate": 3.428419971665452e-07,
+      "loss": 0.1153,
+      "step": 9624
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1921919757851094,
+      "learning_rate": 3.4274396068555446e-07,
+      "loss": 0.3812,
+      "step": 9625
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4931510013296186,
+      "learning_rate": 3.4264593091342225e-07,
+      "loss": 0.144,
+      "step": 9626
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.841924744830834,
+      "learning_rate": 3.425479078543312e-07,
+      "loss": 0.0298,
+      "step": 9627
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2763868997235948,
+      "learning_rate": 3.424498915124627e-07,
+      "loss": 0.2633,
+      "step": 9628
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4627046317141414,
+      "learning_rate": 3.423518818919987e-07,
+      "loss": 0.0877,
+      "step": 9629
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1415060188542492,
+      "learning_rate": 3.4225387899712017e-07,
+      "loss": 0.2889,
+      "step": 9630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8697188070980397,
+      "learning_rate": 3.4215588283200847e-07,
+      "loss": 0.4644,
+      "step": 9631
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7868394702210959,
+      "learning_rate": 3.42057893400844e-07,
+      "loss": 0.1948,
+      "step": 9632
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7210343365033178,
+      "learning_rate": 3.419599107078072e-07,
+      "loss": 0.0843,
+      "step": 9633
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9195153597489307,
+      "learning_rate": 3.418619347570785e-07,
+      "loss": 0.0828,
+      "step": 9634
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.31388945391280904,
+      "learning_rate": 3.4176396555283744e-07,
+      "loss": 0.2229,
+      "step": 9635
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 16.59159076340234,
+      "learning_rate": 3.4166600309926387e-07,
+      "loss": 0.2786,
+      "step": 9636
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5840827296823353,
+      "learning_rate": 3.4156804740053665e-07,
+      "loss": 0.1857,
+      "step": 9637
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.304655110458904,
+      "learning_rate": 3.414700984608352e-07,
+      "loss": 0.0982,
+      "step": 9638
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.23187899125055195,
+      "learning_rate": 3.4137215628433793e-07,
+      "loss": 0.0079,
+      "step": 9639
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.549274691583505,
+      "learning_rate": 3.412742208752234e-07,
+      "loss": 0.2203,
+      "step": 9640
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5426514275423491,
+      "learning_rate": 3.4117629223766965e-07,
+      "loss": 0.1101,
+      "step": 9641
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8701990932823046,
+      "learning_rate": 3.4107837037585456e-07,
+      "loss": 0.2096,
+      "step": 9642
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.31101715403328684,
+      "learning_rate": 3.4098045529395557e-07,
+      "loss": 0.09,
+      "step": 9643
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.49925692395483745,
+      "learning_rate": 3.408825469961503e-07,
+      "loss": 0.009,
+      "step": 9644
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0480465009927071,
+      "learning_rate": 3.407846454866153e-07,
+      "loss": 0.2926,
+      "step": 9645
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1996004083321903,
+      "learning_rate": 3.4068675076952735e-07,
+      "loss": 0.3273,
+      "step": 9646
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6171935971223315,
+      "learning_rate": 3.40588862849063e-07,
+      "loss": 0.1699,
+      "step": 9647
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4455311050832932,
+      "learning_rate": 3.40490981729398e-07,
+      "loss": 0.1556,
+      "step": 9648
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 15.928084012225668,
+      "learning_rate": 3.403931074147085e-07,
+      "loss": 0.1599,
+      "step": 9649
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4388893883048158,
+      "learning_rate": 3.4029523990916984e-07,
+      "loss": 0.0156,
+      "step": 9650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8493084512380781,
+      "learning_rate": 3.4019737921695734e-07,
+      "loss": 0.0307,
+      "step": 9651
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2978810644829313,
+      "learning_rate": 3.4009952534224573e-07,
+      "loss": 0.025,
+      "step": 9652
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0884515437922089,
+      "learning_rate": 3.400016782892101e-07,
+      "loss": 0.0912,
+      "step": 9653
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1045847972889378,
+      "learning_rate": 3.3990383806202427e-07,
+      "loss": 0.218,
+      "step": 9654
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3820573701876342,
+      "learning_rate": 3.398060046648627e-07,
+      "loss": 0.1362,
+      "step": 9655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2204567299780504,
+      "learning_rate": 3.3970817810189883e-07,
+      "loss": 0.2979,
+      "step": 9656
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3358705612120792,
+      "learning_rate": 3.396103583773066e-07,
+      "loss": 0.0846,
+      "step": 9657
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7443790564141608,
+      "learning_rate": 3.3951254549525865e-07,
+      "loss": 0.2865,
+      "step": 9658
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.864869465123004,
+      "learning_rate": 3.394147394599281e-07,
+      "loss": 0.1399,
+      "step": 9659
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0788401377321628,
+      "learning_rate": 3.3931694027548774e-07,
+      "loss": 0.1065,
+      "step": 9660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8862960168875686,
+      "learning_rate": 3.392191479461096e-07,
+      "loss": 0.1619,
+      "step": 9661
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.45105233462744826,
+      "learning_rate": 3.39121362475966e-07,
+      "loss": 0.011,
+      "step": 9662
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.40562950885699883,
+      "learning_rate": 3.3902358386922823e-07,
+      "loss": 0.1351,
+      "step": 9663
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0528420955859363,
+      "learning_rate": 3.389258121300682e-07,
+      "loss": 0.142,
+      "step": 9664
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.322898913995691,
+      "learning_rate": 3.388280472626567e-07,
+      "loss": 0.0512,
+      "step": 9665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9924111999275667,
+      "learning_rate": 3.3873028927116474e-07,
+      "loss": 0.1011,
+      "step": 9666
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6810908359051329,
+      "learning_rate": 3.386325381597628e-07,
+      "loss": 0.1435,
+      "step": 9667
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.973501401654767,
+      "learning_rate": 3.385347939326212e-07,
+      "loss": 0.161,
+      "step": 9668
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4225090616841344,
+      "learning_rate": 3.3843705659390977e-07,
+      "loss": 0.0649,
+      "step": 9669
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9910227797615438,
+      "learning_rate": 3.383393261477985e-07,
+      "loss": 0.1963,
+      "step": 9670
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8026710605913076,
+      "learning_rate": 3.382416025984563e-07,
+      "loss": 0.3476,
+      "step": 9671
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.43782863834267566,
+      "learning_rate": 3.3814388595005274e-07,
+      "loss": 0.0489,
+      "step": 9672
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7834382778568777,
+      "learning_rate": 3.380461762067564e-07,
+      "loss": 0.3936,
+      "step": 9673
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.140878511252451,
+      "learning_rate": 3.379484733727356e-07,
+      "loss": 0.2009,
+      "step": 9674
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.145944887695228,
+      "learning_rate": 3.3785077745215867e-07,
+      "loss": 0.1061,
+      "step": 9675
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7828966360437236,
+      "learning_rate": 3.377530884491936e-07,
+      "loss": 0.133,
+      "step": 9676
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9790514628359507,
+      "learning_rate": 3.3765540636800795e-07,
+      "loss": 0.2241,
+      "step": 9677
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6244034945981407,
+      "learning_rate": 3.375577312127689e-07,
+      "loss": 0.0923,
+      "step": 9678
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1176109283196176,
+      "learning_rate": 3.374600629876437e-07,
+      "loss": 0.2063,
+      "step": 9679
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1400215488775414,
+      "learning_rate": 3.373624016967989e-07,
+      "loss": 0.339,
+      "step": 9680
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7266091394368728,
+      "learning_rate": 3.372647473444011e-07,
+      "loss": 0.1516,
+      "step": 9681
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.998266862212202,
+      "learning_rate": 3.371670999346162e-07,
+      "loss": 0.226,
+      "step": 9682
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1303660680185235,
+      "learning_rate": 3.370694594716103e-07,
+      "loss": 0.3154,
+      "step": 9683
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5290700509410663,
+      "learning_rate": 3.369718259595486e-07,
+      "loss": 0.1477,
+      "step": 9684
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0769691318048444,
+      "learning_rate": 3.368741994025966e-07,
+      "loss": 0.1802,
+      "step": 9685
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1874478464550087,
+      "learning_rate": 3.367765798049193e-07,
+      "loss": 0.2394,
+      "step": 9686
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9981652510796589,
+      "learning_rate": 3.36678967170681e-07,
+      "loss": 0.2965,
+      "step": 9687
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.38936891349843017,
+      "learning_rate": 3.365813615040465e-07,
+      "loss": 0.0568,
+      "step": 9688
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7970928639037027,
+      "learning_rate": 3.3648376280917946e-07,
+      "loss": 0.0278,
+      "step": 9689
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.44544585074375437,
+      "learning_rate": 3.3638617109024405e-07,
+      "loss": 0.0273,
+      "step": 9690
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.623778611811336,
+      "learning_rate": 3.3628858635140317e-07,
+      "loss": 0.3168,
+      "step": 9691
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.838098790065471,
+      "learning_rate": 3.361910085968205e-07,
+      "loss": 0.1211,
+      "step": 9692
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5313964576854305,
+      "learning_rate": 3.360934378306586e-07,
+      "loss": 0.1963,
+      "step": 9693
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5870650851985146,
+      "learning_rate": 3.359958740570802e-07,
+      "loss": 0.2638,
+      "step": 9694
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7881511462464323,
+      "learning_rate": 3.3589831728024733e-07,
+      "loss": 0.2737,
+      "step": 9695
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0385252321720932,
+      "learning_rate": 3.358007675043224e-07,
+      "loss": 0.3148,
+      "step": 9696
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.494414612317763,
+      "learning_rate": 3.357032247334666e-07,
+      "loss": 0.0761,
+      "step": 9697
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4388717112905187,
+      "learning_rate": 3.356056889718417e-07,
+      "loss": 0.2136,
+      "step": 9698
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1491435647868617,
+      "learning_rate": 3.355081602236086e-07,
+      "loss": 0.2962,
+      "step": 9699
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0983749076551765,
+      "learning_rate": 3.3541063849292785e-07,
+      "loss": 0.1626,
+      "step": 9700
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 19.719457677317227,
+      "learning_rate": 3.3531312378396023e-07,
+      "loss": 0.1695,
+      "step": 9701
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5293311824003611,
+      "learning_rate": 3.352156161008658e-07,
+      "loss": 0.1906,
+      "step": 9702
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7800234034844945,
+      "learning_rate": 3.3511811544780445e-07,
+      "loss": 0.0977,
+      "step": 9703
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2912943653128102,
+      "learning_rate": 3.3502062182893563e-07,
+      "loss": 0.0991,
+      "step": 9704
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9972025373690413,
+      "learning_rate": 3.3492313524841896e-07,
+      "loss": 0.4256,
+      "step": 9705
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.394740195822023,
+      "learning_rate": 3.34825655710413e-07,
+      "loss": 0.0541,
+      "step": 9706
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8023099451970512,
+      "learning_rate": 3.3472818321907677e-07,
+      "loss": 0.1039,
+      "step": 9707
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5450040172144064,
+      "learning_rate": 3.3463071777856826e-07,
+      "loss": 0.0747,
+      "step": 9708
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6629103130157872,
+      "learning_rate": 3.34533259393046e-07,
+      "loss": 0.2211,
+      "step": 9709
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0004173541174128,
+      "learning_rate": 3.344358080666674e-07,
+      "loss": 0.2864,
+      "step": 9710
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5784660571422798,
+      "learning_rate": 3.3433836380359017e-07,
+      "loss": 0.1308,
+      "step": 9711
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.429189676289002,
+      "learning_rate": 3.342409266079711e-07,
+      "loss": 0.202,
+      "step": 9712
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.455297020664025,
+      "learning_rate": 3.3414349648396755e-07,
+      "loss": 0.1935,
+      "step": 9713
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3197767456611609,
+      "learning_rate": 3.340460734357359e-07,
+      "loss": 0.015,
+      "step": 9714
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6251486683180227,
+      "learning_rate": 3.339486574674321e-07,
+      "loss": 0.105,
+      "step": 9715
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5383432674816208,
+      "learning_rate": 3.338512485832127e-07,
+      "loss": 0.2392,
+      "step": 9716
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5975307515061975,
+      "learning_rate": 3.3375384678723275e-07,
+      "loss": 0.2254,
+      "step": 9717
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2249868304715692,
+      "learning_rate": 3.3365645208364814e-07,
+      "loss": 0.206,
+      "step": 9718
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5727470827042784,
+      "learning_rate": 3.335590644766134e-07,
+      "loss": 0.1096,
+      "step": 9719
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26983518233568515,
+      "learning_rate": 3.3346168397028375e-07,
+      "loss": 0.1458,
+      "step": 9720
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7253371494377086,
+      "learning_rate": 3.333643105688134e-07,
+      "loss": 0.3655,
+      "step": 9721
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.25924859035933084,
+      "learning_rate": 3.3326694427635657e-07,
+      "loss": 0.0089,
+      "step": 9722
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4638056171243157,
+      "learning_rate": 3.3316958509706695e-07,
+      "loss": 0.1153,
+      "step": 9723
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7050846533927227,
+      "learning_rate": 3.3307223303509835e-07,
+      "loss": 0.1181,
+      "step": 9724
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.51324220020139,
+      "learning_rate": 3.329748880946037e-07,
+      "loss": 0.3162,
+      "step": 9725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7059254358554043,
+      "learning_rate": 3.328775502797363e-07,
+      "loss": 0.1879,
+      "step": 9726
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.454404948998227,
+      "learning_rate": 3.3278021959464857e-07,
+      "loss": 0.1415,
+      "step": 9727
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.855437271924765,
+      "learning_rate": 3.3268289604349266e-07,
+      "loss": 0.2297,
+      "step": 9728
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8428411441037661,
+      "learning_rate": 3.325855796304208e-07,
+      "loss": 0.1808,
+      "step": 9729
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9022351116652456,
+      "learning_rate": 3.324882703595845e-07,
+      "loss": 0.3482,
+      "step": 9730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5492724358666593,
+      "learning_rate": 3.3239096823513565e-07,
+      "loss": 0.1431,
+      "step": 9731
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.918600336029918,
+      "learning_rate": 3.322936732612247e-07,
+      "loss": 0.1724,
+      "step": 9732
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7260838766547156,
+      "learning_rate": 3.321963854420031e-07,
+      "loss": 0.0123,
+      "step": 9733
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.844162753335101,
+      "learning_rate": 3.3209910478162077e-07,
+      "loss": 0.2646,
+      "step": 9734
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2387889557100995,
+      "learning_rate": 3.320018312842284e-07,
+      "loss": 0.1458,
+      "step": 9735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.224096149135845,
+      "learning_rate": 3.3190456495397534e-07,
+      "loss": 0.2795,
+      "step": 9736
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9895912364231522,
+      "learning_rate": 3.318073057950117e-07,
+      "loss": 0.089,
+      "step": 9737
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.013396254623336,
+      "learning_rate": 3.317100538114863e-07,
+      "loss": 0.1287,
+      "step": 9738
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.944640453005538,
+      "learning_rate": 3.3161280900754845e-07,
+      "loss": 0.0451,
+      "step": 9739
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4478979865196622,
+      "learning_rate": 3.3151557138734655e-07,
+      "loss": 0.3024,
+      "step": 9740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.140879600682289,
+      "learning_rate": 3.314183409550292e-07,
+      "loss": 0.2722,
+      "step": 9741
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2735408774536647,
+      "learning_rate": 3.313211177147444e-07,
+      "loss": 0.0224,
+      "step": 9742
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2281358995379472,
+      "learning_rate": 3.3122390167063965e-07,
+      "loss": 0.2272,
+      "step": 9743
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4739552998196937,
+      "learning_rate": 3.311266928268626e-07,
+      "loss": 0.1578,
+      "step": 9744
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8496568020715792,
+      "learning_rate": 3.3102949118756017e-07,
+      "loss": 0.1799,
+      "step": 9745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1675489398324537,
+      "learning_rate": 3.3093229675687945e-07,
+      "loss": 0.0326,
+      "step": 9746
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9339951663613293,
+      "learning_rate": 3.3083510953896676e-07,
+      "loss": 0.2022,
+      "step": 9747
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6903485604455774,
+      "learning_rate": 3.307379295379684e-07,
+      "loss": 0.2035,
+      "step": 9748
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0305568072894793,
+      "learning_rate": 3.3064075675803016e-07,
+      "loss": 0.2289,
+      "step": 9749
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.941810575292261,
+      "learning_rate": 3.3054359120329786e-07,
+      "loss": 0.1156,
+      "step": 9750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5583045424424432,
+      "learning_rate": 3.3044643287791643e-07,
+      "loss": 0.1339,
+      "step": 9751
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9404728184231548,
+      "learning_rate": 3.3034928178603115e-07,
+      "loss": 0.0288,
+      "step": 9752
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8899062681751952,
+      "learning_rate": 3.3025213793178643e-07,
+      "loss": 0.1342,
+      "step": 9753
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3018027126828369,
+      "learning_rate": 3.301550013193268e-07,
+      "loss": 0.2209,
+      "step": 9754
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4847358677401563,
+      "learning_rate": 3.300578719527963e-07,
+      "loss": 0.0411,
+      "step": 9755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7310701196391066,
+      "learning_rate": 3.2996074983633846e-07,
+      "loss": 0.3191,
+      "step": 9756
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5122777123842382,
+      "learning_rate": 3.2986363497409707e-07,
+      "loss": 0.1452,
+      "step": 9757
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.29930004610316,
+      "learning_rate": 3.2976652737021483e-07,
+      "loss": 0.1408,
+      "step": 9758
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7674445115922284,
+      "learning_rate": 3.2966942702883485e-07,
+      "loss": 0.1664,
+      "step": 9759
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2529105157325676,
+      "learning_rate": 3.2957233395409945e-07,
+      "loss": 0.0529,
+      "step": 9760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7890271333132229,
+      "learning_rate": 3.29475248150151e-07,
+      "loss": 0.1605,
+      "step": 9761
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7592650520725693,
+      "learning_rate": 3.293781696211311e-07,
+      "loss": 0.2552,
+      "step": 9762
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1271683139864541,
+      "learning_rate": 3.2928109837118165e-07,
+      "loss": 0.1316,
+      "step": 9763
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.793001244147618,
+      "learning_rate": 3.291840344044435e-07,
+      "loss": 0.0628,
+      "step": 9764
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2535944043000978,
+      "learning_rate": 3.2908697772505807e-07,
+      "loss": 0.0689,
+      "step": 9765
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.561272604758733,
+      "learning_rate": 3.2898992833716563e-07,
+      "loss": 0.289,
+      "step": 9766
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3481681415578158,
+      "learning_rate": 3.288928862449066e-07,
+      "loss": 0.193,
+      "step": 9767
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1000027488492163,
+      "learning_rate": 3.2879585145242117e-07,
+      "loss": 0.1165,
+      "step": 9768
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7352536233546072,
+      "learning_rate": 3.2869882396384873e-07,
+      "loss": 0.1636,
+      "step": 9769
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1403764319347593,
+      "learning_rate": 3.286018037833289e-07,
+      "loss": 0.414,
+      "step": 9770
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6444888092476602,
+      "learning_rate": 3.2850479091500053e-07,
+      "loss": 0.0698,
+      "step": 9771
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8511599200217639,
+      "learning_rate": 3.284077853630027e-07,
+      "loss": 0.246,
+      "step": 9772
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27623584953854535,
+      "learning_rate": 3.2831078713147354e-07,
+      "loss": 0.0084,
+      "step": 9773
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2594911388306558,
+      "learning_rate": 3.2821379622455146e-07,
+      "loss": 0.2543,
+      "step": 9774
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 18.190853615986125,
+      "learning_rate": 3.2811681264637405e-07,
+      "loss": 0.3407,
+      "step": 9775
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.4178107490553025,
+      "learning_rate": 3.280198364010791e-07,
+      "loss": 0.3291,
+      "step": 9776
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.980847201905259,
+      "learning_rate": 3.2792286749280346e-07,
+      "loss": 0.1604,
+      "step": 9777
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1848945969660434,
+      "learning_rate": 3.2782590592568436e-07,
+      "loss": 0.1019,
+      "step": 9778
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7675199347965304,
+      "learning_rate": 3.2772895170385816e-07,
+      "loss": 0.1948,
+      "step": 9779
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6438653739147412,
+      "learning_rate": 3.2763200483146125e-07,
+      "loss": 0.2464,
+      "step": 9780
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5752937854883382,
+      "learning_rate": 3.275350653126294e-07,
+      "loss": 0.2327,
+      "step": 9781
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3260922813825861,
+      "learning_rate": 3.274381331514986e-07,
+      "loss": 0.216,
+      "step": 9782
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6309515879234886,
+      "learning_rate": 3.2734120835220377e-07,
+      "loss": 0.1001,
+      "step": 9783
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9280792828751381,
+      "learning_rate": 3.2724429091887993e-07,
+      "loss": 0.2191,
+      "step": 9784
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1782041291181589,
+      "learning_rate": 3.271473808556622e-07,
+      "loss": 0.4337,
+      "step": 9785
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3325734765309915,
+      "learning_rate": 3.270504781666845e-07,
+      "loss": 0.1293,
+      "step": 9786
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.11298133267691497,
+      "learning_rate": 3.269535828560812e-07,
+      "loss": 0.0005,
+      "step": 9787
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6464759162177716,
+      "learning_rate": 3.268566949279857e-07,
+      "loss": 0.1142,
+      "step": 9788
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.900505378403551,
+      "learning_rate": 3.26759814386532e-07,
+      "loss": 0.1764,
+      "step": 9789
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.108662414175651,
+      "learning_rate": 3.2666294123585253e-07,
+      "loss": 0.1668,
+      "step": 9790
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.205937580649672,
+      "learning_rate": 3.2656607548008064e-07,
+      "loss": 0.2206,
+      "step": 9791
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1276834088401069,
+      "learning_rate": 3.2646921712334854e-07,
+      "loss": 0.1915,
+      "step": 9792
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.041790434933491,
+      "learning_rate": 3.263723661697885e-07,
+      "loss": 0.0513,
+      "step": 9793
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.45748085143230005,
+      "learning_rate": 3.2627552262353223e-07,
+      "loss": 0.1839,
+      "step": 9794
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.37755251386908534,
+      "learning_rate": 3.2617868648871167e-07,
+      "loss": 0.0076,
+      "step": 9795
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9007939493937487,
+      "learning_rate": 3.2608185776945773e-07,
+      "loss": 0.0692,
+      "step": 9796
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.41751782266332393,
+      "learning_rate": 3.259850364699012e-07,
+      "loss": 0.0901,
+      "step": 9797
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7244555180524256,
+      "learning_rate": 3.2588822259417294e-07,
+      "loss": 0.1607,
+      "step": 9798
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.639742382083237,
+      "learning_rate": 3.2579141614640293e-07,
+      "loss": 0.0075,
+      "step": 9799
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.83033785241535,
+      "learning_rate": 3.2569461713072145e-07,
+      "loss": 0.3367,
+      "step": 9800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8677793501382356,
+      "learning_rate": 3.255978255512579e-07,
+      "loss": 0.2052,
+      "step": 9801
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8380638557553156,
+      "learning_rate": 3.2550104141214184e-07,
+      "loss": 0.251,
+      "step": 9802
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.065585588617662,
+      "learning_rate": 3.2540426471750197e-07,
+      "loss": 0.0808,
+      "step": 9803
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1518702118880153,
+      "learning_rate": 3.253074954714674e-07,
+      "loss": 0.2295,
+      "step": 9804
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.898787117839744,
+      "learning_rate": 3.2521073367816597e-07,
+      "loss": 0.0493,
+      "step": 9805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6507183401082992,
+      "learning_rate": 3.251139793417263e-07,
+      "loss": 0.013,
+      "step": 9806
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3547308846636823,
+      "learning_rate": 3.250172324662756e-07,
+      "loss": 0.1052,
+      "step": 9807
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 20.268166422279045,
+      "learning_rate": 3.249204930559417e-07,
+      "loss": 0.1381,
+      "step": 9808
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5414980160997245,
+      "learning_rate": 3.248237611148514e-07,
+      "loss": 0.2751,
+      "step": 9809
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.27660086068766,
+      "learning_rate": 3.247270366471316e-07,
+      "loss": 0.1243,
+      "step": 9810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5296173960649535,
+      "learning_rate": 3.246303196569089e-07,
+      "loss": 0.1863,
+      "step": 9811
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.590121154826465,
+      "learning_rate": 3.2453361014830904e-07,
+      "loss": 0.023,
+      "step": 9812
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8378317787950598,
+      "learning_rate": 3.244369081254584e-07,
+      "loss": 0.1815,
+      "step": 9813
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.678106466930295,
+      "learning_rate": 3.24340213592482e-07,
+      "loss": 0.2562,
+      "step": 9814
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.061081552019047,
+      "learning_rate": 3.242435265535053e-07,
+      "loss": 0.2101,
+      "step": 9815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4553365836982701,
+      "learning_rate": 3.241468470126529e-07,
+      "loss": 0.3213,
+      "step": 9816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6583729938209066,
+      "learning_rate": 3.240501749740496e-07,
+      "loss": 0.1729,
+      "step": 9817
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5780146377092252,
+      "learning_rate": 3.239535104418196e-07,
+      "loss": 0.3041,
+      "step": 9818
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.501996112418068,
+      "learning_rate": 3.2385685342008674e-07,
+      "loss": 0.2809,
+      "step": 9819
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6947544197855776,
+      "learning_rate": 3.2376020391297455e-07,
+      "loss": 0.1819,
+      "step": 9820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5802086760688637,
+      "learning_rate": 3.236635619246064e-07,
+      "loss": 0.1085,
+      "step": 9821
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6207744953822706,
+      "learning_rate": 3.235669274591051e-07,
+      "loss": 0.074,
+      "step": 9822
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6310128693542639,
+      "learning_rate": 3.2347030052059354e-07,
+      "loss": 0.0043,
+      "step": 9823
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5035220755686219,
+      "learning_rate": 3.233736811131938e-07,
+      "loss": 0.3291,
+      "step": 9824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3402357618229086,
+      "learning_rate": 3.232770692410277e-07,
+      "loss": 0.1443,
+      "step": 9825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7722676121731585,
+      "learning_rate": 3.2318046490821716e-07,
+      "loss": 0.2044,
+      "step": 9826
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5447117817685545,
+      "learning_rate": 3.2308386811888346e-07,
+      "loss": 0.0425,
+      "step": 9827
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8505989702153116,
+      "learning_rate": 3.229872788771476e-07,
+      "loss": 0.4057,
+      "step": 9828
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6823708881589299,
+      "learning_rate": 3.2289069718713016e-07,
+      "loss": 0.1218,
+      "step": 9829
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0516683346630333,
+      "learning_rate": 3.227941230529517e-07,
+      "loss": 0.1493,
+      "step": 9830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5065633695756446,
+      "learning_rate": 3.2269755647873214e-07,
+      "loss": 0.2123,
+      "step": 9831
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 14.212562342131438,
+      "learning_rate": 3.226009974685914e-07,
+      "loss": 0.0154,
+      "step": 9832
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.42259660204897154,
+      "learning_rate": 3.225044460266485e-07,
+      "loss": 0.1057,
+      "step": 9833
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.89121376067872,
+      "learning_rate": 3.2240790215702297e-07,
+      "loss": 0.4477,
+      "step": 9834
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.016530243754151,
+      "learning_rate": 3.223113658638331e-07,
+      "loss": 0.0705,
+      "step": 9835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.19237502739439066,
+      "learning_rate": 3.222148371511977e-07,
+      "loss": 0.1574,
+      "step": 9836
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9736830476847157,
+      "learning_rate": 3.221183160232348e-07,
+      "loss": 0.2513,
+      "step": 9837
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5915501474496268,
+      "learning_rate": 3.2202180248406196e-07,
+      "loss": 0.3128,
+      "step": 9838
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4582903970709344,
+      "learning_rate": 3.21925296537797e-07,
+      "loss": 0.0935,
+      "step": 9839
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6102361646494264,
+      "learning_rate": 3.218287981885567e-07,
+      "loss": 0.2526,
+      "step": 9840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.70409353789978,
+      "learning_rate": 3.2173230744045815e-07,
+      "loss": 0.3165,
+      "step": 9841
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3060056987874527,
+      "learning_rate": 3.216358242976176e-07,
+      "loss": 0.0452,
+      "step": 9842
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.359948758444199,
+      "learning_rate": 3.2153934876415143e-07,
+      "loss": 0.1265,
+      "step": 9843
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0029068397163248,
+      "learning_rate": 3.214428808441754e-07,
+      "loss": 0.2236,
+      "step": 9844
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6554839420953738,
+      "learning_rate": 3.2134642054180493e-07,
+      "loss": 0.1272,
+      "step": 9845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1511883519595218,
+      "learning_rate": 3.2124996786115524e-07,
+      "loss": 0.1783,
+      "step": 9846
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3290639556624124,
+      "learning_rate": 3.211535228063415e-07,
+      "loss": 0.3526,
+      "step": 9847
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.436707333221691,
+      "learning_rate": 3.210570853814778e-07,
+      "loss": 0.0636,
+      "step": 9848
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9722318766730552,
+      "learning_rate": 3.209606555906788e-07,
+      "loss": 0.2851,
+      "step": 9849
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6854531809469963,
+      "learning_rate": 3.208642334380578e-07,
+      "loss": 0.2056,
+      "step": 9850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1512496298708337,
+      "learning_rate": 3.2076781892772904e-07,
+      "loss": 0.1194,
+      "step": 9851
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9946664661922305,
+      "learning_rate": 3.2067141206380526e-07,
+      "loss": 0.0877,
+      "step": 9852
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7769647861589055,
+      "learning_rate": 3.2057501285039957e-07,
+      "loss": 0.2582,
+      "step": 9853
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6634267476529329,
+      "learning_rate": 3.204786212916245e-07,
+      "loss": 0.2364,
+      "step": 9854
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6357232327354747,
+      "learning_rate": 3.2038223739159225e-07,
+      "loss": 0.1074,
+      "step": 9855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6956484779394868,
+      "learning_rate": 3.2028586115441504e-07,
+      "loss": 0.3608,
+      "step": 9856
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4483626126431408,
+      "learning_rate": 3.2018949258420404e-07,
+      "loss": 0.2857,
+      "step": 9857
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.892691276179684,
+      "learning_rate": 3.2009313168507096e-07,
+      "loss": 0.1045,
+      "step": 9858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.604541284668023,
+      "learning_rate": 3.1999677846112634e-07,
+      "loss": 0.3529,
+      "step": 9859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8117811118884894,
+      "learning_rate": 3.1990043291648116e-07,
+      "loss": 0.2828,
+      "step": 9860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8544328729224833,
+      "learning_rate": 3.198040950552454e-07,
+      "loss": 0.2184,
+      "step": 9861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.107901832492373,
+      "learning_rate": 3.197077648815293e-07,
+      "loss": 0.211,
+      "step": 9862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2094327134884784,
+      "learning_rate": 3.196114423994424e-07,
+      "loss": 0.1163,
+      "step": 9863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2872753391107936,
+      "learning_rate": 3.19515127613094e-07,
+      "loss": 0.1766,
+      "step": 9864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.380807332028745,
+      "learning_rate": 3.1941882052659307e-07,
+      "loss": 0.2257,
+      "step": 9865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1074461411061158,
+      "learning_rate": 3.1932252114404813e-07,
+      "loss": 0.3668,
+      "step": 9866
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.38756982710411836,
+      "learning_rate": 3.1922622946956787e-07,
+      "loss": 0.1879,
+      "step": 9867
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1471559271998129,
+      "learning_rate": 3.191299455072598e-07,
+      "loss": 0.1824,
+      "step": 9868
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.467852571762619,
+      "learning_rate": 3.190336692612321e-07,
+      "loss": 0.497,
+      "step": 9869
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8289119305925776,
+      "learning_rate": 3.1893740073559164e-07,
+      "loss": 0.0471,
+      "step": 9870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2894254871999359,
+      "learning_rate": 3.1884113993444576e-07,
+      "loss": 0.1185,
+      "step": 9871
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2244586613989046,
+      "learning_rate": 3.18744886861901e-07,
+      "loss": 0.3413,
+      "step": 9872
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7645184562699384,
+      "learning_rate": 3.186486415220638e-07,
+      "loss": 0.3322,
+      "step": 9873
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7867369180073533,
+      "learning_rate": 3.1855240391903994e-07,
+      "loss": 0.2233,
+      "step": 9874
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.21979088857614315,
+      "learning_rate": 3.1845617405693554e-07,
+      "loss": 0.0017,
+      "step": 9875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.412127161499392,
+      "learning_rate": 3.1835995193985546e-07,
+      "loss": 0.1026,
+      "step": 9876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.150974680950936,
+      "learning_rate": 3.1826373757190515e-07,
+      "loss": 0.0228,
+      "step": 9877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5825104628199708,
+      "learning_rate": 3.1816753095718914e-07,
+      "loss": 0.2368,
+      "step": 9878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.99293123207642,
+      "learning_rate": 3.1807133209981167e-07,
+      "loss": 0.1077,
+      "step": 9879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9097358517491003,
+      "learning_rate": 3.17975141003877e-07,
+      "loss": 0.3015,
+      "step": 9880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.226154462304313,
+      "learning_rate": 3.1787895767348856e-07,
+      "loss": 0.1915,
+      "step": 9881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0056129689625313,
+      "learning_rate": 3.177827821127501e-07,
+      "loss": 0.1646,
+      "step": 9882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5020985357185783,
+      "learning_rate": 3.1768661432576425e-07,
+      "loss": 0.4339,
+      "step": 9883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.327293405068815,
+      "learning_rate": 3.175904543166342e-07,
+      "loss": 0.238,
+      "step": 9884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.45995849192979227,
+      "learning_rate": 3.174943020894618e-07,
+      "loss": 0.2684,
+      "step": 9885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.683662106369194,
+      "learning_rate": 3.173981576483495e-07,
+      "loss": 0.3656,
+      "step": 9886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.639972937803985,
+      "learning_rate": 3.173020209973988e-07,
+      "loss": 0.0426,
+      "step": 9887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6160806936059422,
+      "learning_rate": 3.172058921407112e-07,
+      "loss": 0.1846,
+      "step": 9888
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7642852608321751,
+      "learning_rate": 3.1710977108238764e-07,
+      "loss": 0.2798,
+      "step": 9889
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.56581276208378,
+      "learning_rate": 3.170136578265289e-07,
+      "loss": 0.1302,
+      "step": 9890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 15.594260556517076,
+      "learning_rate": 3.1691755237723536e-07,
+      "loss": 0.1422,
+      "step": 9891
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.893153347593312,
+      "learning_rate": 3.168214547386072e-07,
+      "loss": 0.2696,
+      "step": 9892
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4365150509337516,
+      "learning_rate": 3.1672536491474404e-07,
+      "loss": 0.1797,
+      "step": 9893
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.122466333845425,
+      "learning_rate": 3.1662928290974514e-07,
+      "loss": 0.0621,
+      "step": 9894
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 15.115066480424131,
+      "learning_rate": 3.165332087277097e-07,
+      "loss": 0.1453,
+      "step": 9895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.808008933131503,
+      "learning_rate": 3.164371423727362e-07,
+      "loss": 0.2404,
+      "step": 9896
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4273001064281147,
+      "learning_rate": 3.163410838489234e-07,
+      "loss": 0.1283,
+      "step": 9897
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0126728978340207,
+      "learning_rate": 3.162450331603691e-07,
+      "loss": 0.2811,
+      "step": 9898
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6151956085625986,
+      "learning_rate": 3.1614899031117113e-07,
+      "loss": 0.2404,
+      "step": 9899
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.918530267726266,
+      "learning_rate": 3.160529553054267e-07,
+      "loss": 0.1715,
+      "step": 9900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7940771940834439,
+      "learning_rate": 3.159569281472332e-07,
+      "loss": 0.0646,
+      "step": 9901
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.132041643094976,
+      "learning_rate": 3.1586090884068685e-07,
+      "loss": 0.1998,
+      "step": 9902
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.09350245611442531,
+      "learning_rate": 3.1576489738988455e-07,
+      "loss": 0.0017,
+      "step": 9903
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7649371838661816,
+      "learning_rate": 3.1566889379892193e-07,
+      "loss": 0.0868,
+      "step": 9904
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9467610536431719,
+      "learning_rate": 3.15572898071895e-07,
+      "loss": 0.2727,
+      "step": 9905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8475691244522232,
+      "learning_rate": 3.1547691021289885e-07,
+      "loss": 0.1743,
+      "step": 9906
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3443031278498072,
+      "learning_rate": 3.1538093022602857e-07,
+      "loss": 0.0531,
+      "step": 9907
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4824038418934789,
+      "learning_rate": 3.1528495811537916e-07,
+      "loss": 0.251,
+      "step": 9908
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9748168955001419,
+      "learning_rate": 3.151889938850445e-07,
+      "loss": 0.074,
+      "step": 9909
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7990677806453492,
+      "learning_rate": 3.1509303753911916e-07,
+      "loss": 0.1646,
+      "step": 9910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.059348594169725,
+      "learning_rate": 3.149970890816963e-07,
+      "loss": 0.2524,
+      "step": 9911
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3788194423194713,
+      "learning_rate": 3.1490114851686975e-07,
+      "loss": 0.2684,
+      "step": 9912
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7774822863336062,
+      "learning_rate": 3.148052158487321e-07,
+      "loss": 0.0269,
+      "step": 9913
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3721846056736207,
+      "learning_rate": 3.147092910813764e-07,
+      "loss": 0.0962,
+      "step": 9914
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9218045254917826,
+      "learning_rate": 3.146133742188946e-07,
+      "loss": 0.1785,
+      "step": 9915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2376069877986604,
+      "learning_rate": 3.145174652653791e-07,
+      "loss": 0.0841,
+      "step": 9916
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5314508960948032,
+      "learning_rate": 3.144215642249213e-07,
+      "loss": 0.129,
+      "step": 9917
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6824153588707195,
+      "learning_rate": 3.143256711016128e-07,
+      "loss": 0.2261,
+      "step": 9918
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1063518275775739,
+      "learning_rate": 3.1422978589954443e-07,
+      "loss": 0.1102,
+      "step": 9919
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5319474605010603,
+      "learning_rate": 3.1413390862280655e-07,
+      "loss": 0.2272,
+      "step": 9920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5168683468987119,
+      "learning_rate": 3.1403803927549006e-07,
+      "loss": 0.595,
+      "step": 9921
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0728593540967328,
+      "learning_rate": 3.139421778616844e-07,
+      "loss": 0.1412,
+      "step": 9922
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1104532322306133,
+      "learning_rate": 3.1384632438547964e-07,
+      "loss": 0.2981,
+      "step": 9923
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8048437881054166,
+      "learning_rate": 3.137504788509648e-07,
+      "loss": 0.2572,
+      "step": 9924
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.525706281740437,
+      "learning_rate": 3.1365464126222897e-07,
+      "loss": 0.1974,
+      "step": 9925
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3086253821196317,
+      "learning_rate": 3.135588116233607e-07,
+      "loss": 0.0059,
+      "step": 9926
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3348227131092952,
+      "learning_rate": 3.1346298993844844e-07,
+      "loss": 0.2634,
+      "step": 9927
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4285186408888528,
+      "learning_rate": 3.1336717621157986e-07,
+      "loss": 0.0997,
+      "step": 9928
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7934943600101988,
+      "learning_rate": 3.132713704468429e-07,
+      "loss": 0.2026,
+      "step": 9929
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.100488292872332,
+      "learning_rate": 3.131755726483245e-07,
+      "loss": 0.1602,
+      "step": 9930
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5042898575656839,
+      "learning_rate": 3.1307978282011186e-07,
+      "loss": 0.089,
+      "step": 9931
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.345181178870036,
+      "learning_rate": 3.129840009662913e-07,
+      "loss": 0.1918,
+      "step": 9932
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9917433650929446,
+      "learning_rate": 3.1288822709094933e-07,
+      "loss": 0.1131,
+      "step": 9933
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.45394084261043055,
+      "learning_rate": 3.1279246119817174e-07,
+      "loss": 0.2909,
+      "step": 9934
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5376000378997712,
+      "learning_rate": 3.1269670329204393e-07,
+      "loss": 0.1752,
+      "step": 9935
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.782579958651778,
+      "learning_rate": 3.126009533766515e-07,
+      "loss": 0.174,
+      "step": 9936
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1027129316539956,
+      "learning_rate": 3.1250521145607894e-07,
+      "loss": 0.1202,
+      "step": 9937
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6921228097307655,
+      "learning_rate": 3.124094775344112e-07,
+      "loss": 0.216,
+      "step": 9938
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8306429099856422,
+      "learning_rate": 3.1231375161573194e-07,
+      "loss": 0.1582,
+      "step": 9939
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0218445145187087,
+      "learning_rate": 3.122180337041256e-07,
+      "loss": 0.0916,
+      "step": 9940
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3578911721646219,
+      "learning_rate": 3.121223238036752e-07,
+      "loss": 0.3942,
+      "step": 9941
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.109048368111545,
+      "learning_rate": 3.120266219184643e-07,
+      "loss": 0.1368,
+      "step": 9942
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5985847161275715,
+      "learning_rate": 3.1193092805257554e-07,
+      "loss": 0.22,
+      "step": 9943
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30522493991220756,
+      "learning_rate": 3.118352422100915e-07,
+      "loss": 0.0889,
+      "step": 9944
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5032988524392258,
+      "learning_rate": 3.117395643950941e-07,
+      "loss": 0.1759,
+      "step": 9945
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.1292612129831685,
+      "learning_rate": 3.116438946116656e-07,
+      "loss": 0.294,
+      "step": 9946
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.48900880580396,
+      "learning_rate": 3.115482328638872e-07,
+      "loss": 0.0879,
+      "step": 9947
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.261090140424646,
+      "learning_rate": 3.1145257915583975e-07,
+      "loss": 0.1679,
+      "step": 9948
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0631672282215923,
+      "learning_rate": 3.1135693349160463e-07,
+      "loss": 0.1693,
+      "step": 9949
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1328929962914109,
+      "learning_rate": 3.112612958752617e-07,
+      "loss": 0.0741,
+      "step": 9950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.47584586979102117,
+      "learning_rate": 3.111656663108914e-07,
+      "loss": 0.0907,
+      "step": 9951
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8814139796048461,
+      "learning_rate": 3.110700448025732e-07,
+      "loss": 0.1669,
+      "step": 9952
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.525102249279778,
+      "learning_rate": 3.1097443135438696e-07,
+      "loss": 0.143,
+      "step": 9953
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.360034123319944,
+      "learning_rate": 3.1087882597041125e-07,
+      "loss": 0.2441,
+      "step": 9954
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8884149877409633,
+      "learning_rate": 3.1078322865472517e-07,
+      "loss": 0.2257,
+      "step": 9955
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4119783766819232,
+      "learning_rate": 3.1068763941140676e-07,
+      "loss": 0.456,
+      "step": 9956
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5584195161214445,
+      "learning_rate": 3.105920582445344e-07,
+      "loss": 0.2614,
+      "step": 9957
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6627540683760709,
+      "learning_rate": 3.104964851581855e-07,
+      "loss": 0.2102,
+      "step": 9958
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.623198031090522,
+      "learning_rate": 3.104009201564376e-07,
+      "loss": 0.301,
+      "step": 9959
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.718368521814724,
+      "learning_rate": 3.103053632433674e-07,
+      "loss": 0.0962,
+      "step": 9960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4246775664980575,
+      "learning_rate": 3.102098144230518e-07,
+      "loss": 0.1491,
+      "step": 9961
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3173459440798687,
+      "learning_rate": 3.101142736995672e-07,
+      "loss": 0.0318,
+      "step": 9962
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.97633954457249,
+      "learning_rate": 3.1001874107698916e-07,
+      "loss": 0.0127,
+      "step": 9963
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2916660812561684,
+      "learning_rate": 3.0992321655939377e-07,
+      "loss": 0.2368,
+      "step": 9964
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.851010525009808,
+      "learning_rate": 3.0982770015085586e-07,
+      "loss": 0.112,
+      "step": 9965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.920113350536923,
+      "learning_rate": 3.097321918554507e-07,
+      "loss": 0.2247,
+      "step": 9966
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6563612146386266,
+      "learning_rate": 3.096366916772526e-07,
+      "loss": 0.1539,
+      "step": 9967
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.80807293742905,
+      "learning_rate": 3.09541199620336e-07,
+      "loss": 0.0922,
+      "step": 9968
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.29216820415518563,
+      "learning_rate": 3.0944571568877466e-07,
+      "loss": 0.2328,
+      "step": 9969
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7729377265353073,
+      "learning_rate": 3.093502398866422e-07,
+      "loss": 0.0259,
+      "step": 9970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0381622148450527,
+      "learning_rate": 3.0925477221801156e-07,
+      "loss": 0.4978,
+      "step": 9971
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.22838959937936235,
+      "learning_rate": 3.0915931268695604e-07,
+      "loss": 0.0997,
+      "step": 9972
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.900846762164515,
+      "learning_rate": 3.090638612975477e-07,
+      "loss": 0.1945,
+      "step": 9973
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5443537803913264,
+      "learning_rate": 3.089684180538591e-07,
+      "loss": 0.1917,
+      "step": 9974
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2006464030075767,
+      "learning_rate": 3.0887298295996177e-07,
+      "loss": 0.2563,
+      "step": 9975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7391239363071468,
+      "learning_rate": 3.0877755601992694e-07,
+      "loss": 0.1639,
+      "step": 9976
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6390049803932814,
+      "learning_rate": 3.0868213723782616e-07,
+      "loss": 0.0482,
+      "step": 9977
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4805558010039064,
+      "learning_rate": 3.0858672661772987e-07,
+      "loss": 0.3875,
+      "step": 9978
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9976928758079943,
+      "learning_rate": 3.0849132416370883e-07,
+      "loss": 0.1156,
+      "step": 9979
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.45979684673637516,
+      "learning_rate": 3.0839592987983264e-07,
+      "loss": 0.1015,
+      "step": 9980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4793416610666731,
+      "learning_rate": 3.0830054377017147e-07,
+      "loss": 0.1903,
+      "step": 9981
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5359966058214101,
+      "learning_rate": 3.082051658387943e-07,
+      "loss": 0.087,
+      "step": 9982
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.385229362388537,
+      "learning_rate": 3.081097960897705e-07,
+      "loss": 0.27,
+      "step": 9983
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6884951815819341,
+      "learning_rate": 3.0801443452716827e-07,
+      "loss": 0.0878,
+      "step": 9984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.112456931004475,
+      "learning_rate": 3.079190811550565e-07,
+      "loss": 0.4376,
+      "step": 9985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.4666678772114836,
+      "learning_rate": 3.078237359775026e-07,
+      "loss": 0.2145,
+      "step": 9986
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.46319190130088983,
+      "learning_rate": 3.0772839899857463e-07,
+      "loss": 0.1443,
+      "step": 9987
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9449454412277086,
+      "learning_rate": 3.0763307022233967e-07,
+      "loss": 0.2502,
+      "step": 9988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5370357256102969,
+      "learning_rate": 3.075377496528645e-07,
+      "loss": 0.2487,
+      "step": 9989
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.35345407023433,
+      "learning_rate": 3.07442437294216e-07,
+      "loss": 0.0602,
+      "step": 9990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7936235863106439,
+      "learning_rate": 3.0734713315046004e-07,
+      "loss": 0.3272,
+      "step": 9991
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4594258122530349,
+      "learning_rate": 3.0725183722566286e-07,
+      "loss": 0.0157,
+      "step": 9992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9367334480193928,
+      "learning_rate": 3.0715654952388954e-07,
+      "loss": 0.242,
+      "step": 9993
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.361592574750528,
+      "learning_rate": 3.0706127004920557e-07,
+      "loss": 0.2584,
+      "step": 9994
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7880141354801808,
+      "learning_rate": 3.0696599880567576e-07,
+      "loss": 0.2367,
+      "step": 9995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7058275558300987,
+      "learning_rate": 3.0687073579736443e-07,
+      "loss": 0.4972,
+      "step": 9996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6251495941489278,
+      "learning_rate": 3.067754810283356e-07,
+      "loss": 0.2803,
+      "step": 9997
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5206111780187221,
+      "learning_rate": 3.0668023450265343e-07,
+      "loss": 0.075,
+      "step": 9998
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4054752988384419,
+      "learning_rate": 3.0658499622438093e-07,
+      "loss": 0.137,
+      "step": 9999
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7106065029504237,
+      "learning_rate": 3.0648976619758143e-07,
+      "loss": 0.1458,
+      "step": 10000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6995757523269095,
+      "learning_rate": 3.0639454442631735e-07,
+      "loss": 0.0273,
+      "step": 10001
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5310302256045529,
+      "learning_rate": 3.0629933091465136e-07,
+      "loss": 0.1255,
+      "step": 10002
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.745436216301801,
+      "learning_rate": 3.062041256666452e-07,
+      "loss": 0.2902,
+      "step": 10003
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6331744260702172,
+      "learning_rate": 3.0610892868636076e-07,
+      "loss": 0.3064,
+      "step": 10004
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9345158770750337,
+      "learning_rate": 3.06013739977859e-07,
+      "loss": 0.1613,
+      "step": 10005
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3799141040742937,
+      "learning_rate": 3.0591855954520106e-07,
+      "loss": 0.2232,
+      "step": 10006
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4388784706078745,
+      "learning_rate": 3.0582338739244765e-07,
+      "loss": 0.0198,
+      "step": 10007
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.048807729296082,
+      "learning_rate": 3.0572822352365873e-07,
+      "loss": 0.4561,
+      "step": 10008
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7215751437856766,
+      "learning_rate": 3.0563306794289455e-07,
+      "loss": 0.2908,
+      "step": 10009
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5288074995535735,
+      "learning_rate": 3.055379206542142e-07,
+      "loss": 0.0895,
+      "step": 10010
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.420119204931513,
+      "learning_rate": 3.0544278166167725e-07,
+      "loss": 0.2203,
+      "step": 10011
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9192937025417116,
+      "learning_rate": 3.0534765096934214e-07,
+      "loss": 0.324,
+      "step": 10012
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3216767944270263,
+      "learning_rate": 3.0525252858126765e-07,
+      "loss": 0.1034,
+      "step": 10013
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8025912749416866,
+      "learning_rate": 3.051574145015118e-07,
+      "loss": 0.419,
+      "step": 10014
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.38860547799048384,
+      "learning_rate": 3.050623087341323e-07,
+      "loss": 0.2668,
+      "step": 10015
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30647173013463946,
+      "learning_rate": 3.049672112831867e-07,
+      "loss": 0.2673,
+      "step": 10016
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1634362172846398,
+      "learning_rate": 3.048721221527317e-07,
+      "loss": 0.1805,
+      "step": 10017
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2710136400778729,
+      "learning_rate": 3.047770413468245e-07,
+      "loss": 0.1133,
+      "step": 10018
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8712449299084689,
+      "learning_rate": 3.046819688695209e-07,
+      "loss": 0.3285,
+      "step": 10019
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6709309625179394,
+      "learning_rate": 3.0458690472487735e-07,
+      "loss": 0.1468,
+      "step": 10020
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8985259676730797,
+      "learning_rate": 3.0449184891694914e-07,
+      "loss": 0.1437,
+      "step": 10021
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.161120386247777,
+      "learning_rate": 3.0439680144979174e-07,
+      "loss": 0.4697,
+      "step": 10022
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5839040079934381,
+      "learning_rate": 3.0430176232745984e-07,
+      "loss": 0.1546,
+      "step": 10023
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.46234028993005594,
+      "learning_rate": 3.0420673155400846e-07,
+      "loss": 0.0834,
+      "step": 10024
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.57650839058358,
+      "learning_rate": 3.041117091334913e-07,
+      "loss": 0.3336,
+      "step": 10025
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0261905910225066,
+      "learning_rate": 3.040166950699625e-07,
+      "loss": 0.3451,
+      "step": 10026
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9864322855566573,
+      "learning_rate": 3.039216893674753e-07,
+      "loss": 0.2225,
+      "step": 10027
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4240512723345102,
+      "learning_rate": 3.038266920300833e-07,
+      "loss": 0.2028,
+      "step": 10028
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4419891423604961,
+      "learning_rate": 3.037317030618388e-07,
+      "loss": 0.0258,
+      "step": 10029
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1329131507572303,
+      "learning_rate": 3.036367224667944e-07,
+      "loss": 0.3439,
+      "step": 10030
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5758442686255685,
+      "learning_rate": 3.0354175024900214e-07,
+      "loss": 0.1377,
+      "step": 10031
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2849863684374576,
+      "learning_rate": 3.0344678641251364e-07,
+      "loss": 0.4759,
+      "step": 10032
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2759617670586145,
+      "learning_rate": 3.0335183096138064e-07,
+      "loss": 0.351,
+      "step": 10033
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7502190509325263,
+      "learning_rate": 3.0325688389965355e-07,
+      "loss": 0.3014,
+      "step": 10034
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2910158057038357,
+      "learning_rate": 3.0316194523138355e-07,
+      "loss": 0.2676,
+      "step": 10035
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9492687373755502,
+      "learning_rate": 3.030670149606205e-07,
+      "loss": 0.2831,
+      "step": 10036
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.22668594642335435,
+      "learning_rate": 3.029720930914146e-07,
+      "loss": 0.0998,
+      "step": 10037
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.43713275265812845,
+      "learning_rate": 3.0287717962781506e-07,
+      "loss": 0.3113,
+      "step": 10038
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.089456274733409,
+      "learning_rate": 3.0278227457387144e-07,
+      "loss": 0.1938,
+      "step": 10039
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5208155172137892,
+      "learning_rate": 3.0268737793363246e-07,
+      "loss": 0.1167,
+      "step": 10040
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9715374167494162,
+      "learning_rate": 3.025924897111466e-07,
+      "loss": 0.2882,
+      "step": 10041
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9099912635361547,
+      "learning_rate": 3.024976099104618e-07,
+      "loss": 0.583,
+      "step": 10042
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.437331688200397,
+      "learning_rate": 3.0240273853562625e-07,
+      "loss": 0.1859,
+      "step": 10043
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2434999435163299,
+      "learning_rate": 3.023078755906871e-07,
+      "loss": 0.2225,
+      "step": 10044
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2345209468975693,
+      "learning_rate": 3.0221302107969114e-07,
+      "loss": 0.39,
+      "step": 10045
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.37678068180372903,
+      "learning_rate": 3.021181750066856e-07,
+      "loss": 0.3625,
+      "step": 10046
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7321066389547666,
+      "learning_rate": 3.0202333737571616e-07,
+      "loss": 0.1433,
+      "step": 10047
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3389152957541529,
+      "learning_rate": 3.0192850819082937e-07,
+      "loss": 0.2074,
+      "step": 10048
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9342221122330095,
+      "learning_rate": 3.018336874560705e-07,
+      "loss": 0.2716,
+      "step": 10049
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2871994912978746,
+      "learning_rate": 3.017388751754849e-07,
+      "loss": 0.4026,
+      "step": 10050
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8573237685064583,
+      "learning_rate": 3.016440713531174e-07,
+      "loss": 0.3745,
+      "step": 10051
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9915866499862802,
+      "learning_rate": 3.0154927599301274e-07,
+      "loss": 0.5495,
+      "step": 10052
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1618492345434936,
+      "learning_rate": 3.014544890992147e-07,
+      "loss": 0.0037,
+      "step": 10053
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.23930629093503197,
+      "learning_rate": 3.013597106757674e-07,
+      "loss": 0.0881,
+      "step": 10054
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7039139070001506,
+      "learning_rate": 3.0126494072671405e-07,
+      "loss": 0.0285,
+      "step": 10055
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.682928743757131,
+      "learning_rate": 3.01170179256098e-07,
+      "loss": 0.1217,
+      "step": 10056
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1586214481233266,
+      "learning_rate": 3.0107542626796165e-07,
+      "loss": 0.2061,
+      "step": 10057
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.33420648078880516,
+      "learning_rate": 3.009806817663475e-07,
+      "loss": 0.0271,
+      "step": 10058
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3803203391794372,
+      "learning_rate": 3.0088594575529774e-07,
+      "loss": 0.0549,
+      "step": 10059
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.44753180417879,
+      "learning_rate": 3.0079121823885356e-07,
+      "loss": 0.0903,
+      "step": 10060
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.38166879068954435,
+      "learning_rate": 3.006964992210567e-07,
+      "loss": 0.0024,
+      "step": 10061
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.17810504323650683,
+      "learning_rate": 3.006017887059476e-07,
+      "loss": 0.1087,
+      "step": 10062
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1141266485442696,
+      "learning_rate": 3.005070866975673e-07,
+      "loss": 0.3222,
+      "step": 10063
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9520269321580211,
+      "learning_rate": 3.0041239319995544e-07,
+      "loss": 0.2189,
+      "step": 10064
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.976190605386346,
+      "learning_rate": 3.003177082171523e-07,
+      "loss": 0.3072,
+      "step": 10065
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.053057840162801,
+      "learning_rate": 3.0022303175319695e-07,
+      "loss": 0.2027,
+      "step": 10066
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5750296983297821,
+      "learning_rate": 3.001283638121288e-07,
+      "loss": 0.0482,
+      "step": 10067
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2917914081263246,
+      "learning_rate": 3.000337043979864e-07,
+      "loss": 0.2143,
+      "step": 10068
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.191854693995644,
+      "learning_rate": 2.9993905351480823e-07,
+      "loss": 0.1763,
+      "step": 10069
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7433518076973855,
+      "learning_rate": 2.9984441116663206e-07,
+      "loss": 0.2507,
+      "step": 10070
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4834550371552282,
+      "learning_rate": 2.997497773574959e-07,
+      "loss": 0.3234,
+      "step": 10071
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.049990209930852,
+      "learning_rate": 2.9965515209143674e-07,
+      "loss": 0.2025,
+      "step": 10072
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6302614591140923,
+      "learning_rate": 2.9956053537249137e-07,
+      "loss": 0.3305,
+      "step": 10073
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8719787278468222,
+      "learning_rate": 2.994659272046966e-07,
+      "loss": 0.1301,
+      "step": 10074
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6982705460270011,
+      "learning_rate": 2.993713275920885e-07,
+      "loss": 0.1801,
+      "step": 10075
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6275087308085543,
+      "learning_rate": 2.992767365387029e-07,
+      "loss": 0.0436,
+      "step": 10076
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2712916019561169,
+      "learning_rate": 2.9918215404857505e-07,
+      "loss": 0.0128,
+      "step": 10077
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1491099588932046,
+      "learning_rate": 2.990875801257404e-07,
+      "loss": 0.1069,
+      "step": 10078
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7451719885124801,
+      "learning_rate": 2.9899301477423336e-07,
+      "loss": 0.2882,
+      "step": 10079
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4185935208862142,
+      "learning_rate": 2.9889845799808854e-07,
+      "loss": 0.1663,
+      "step": 10080
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9512447606296824,
+      "learning_rate": 2.9880390980133954e-07,
+      "loss": 0.22,
+      "step": 10081
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7203886033678134,
+      "learning_rate": 2.9870937018802045e-07,
+      "loss": 0.1875,
+      "step": 10082
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.80985039946878,
+      "learning_rate": 2.98614839162164e-07,
+      "loss": 0.3262,
+      "step": 10083
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3829490125097748,
+      "learning_rate": 2.985203167278035e-07,
+      "loss": 0.1044,
+      "step": 10084
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6240450139913647,
+      "learning_rate": 2.984258028889715e-07,
+      "loss": 0.0988,
+      "step": 10085
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.005710200407289,
+      "learning_rate": 2.983312976496996e-07,
+      "loss": 0.3275,
+      "step": 10086
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9079486894107951,
+      "learning_rate": 2.982368010140203e-07,
+      "loss": 0.0682,
+      "step": 10087
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9466190574800312,
+      "learning_rate": 2.981423129859643e-07,
+      "loss": 0.385,
+      "step": 10088
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7014769328021367,
+      "learning_rate": 2.980478335695633e-07,
+      "loss": 0.1394,
+      "step": 10089
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0209469875880932,
+      "learning_rate": 2.9795336276884753e-07,
+      "loss": 0.2723,
+      "step": 10090
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6572692611290462,
+      "learning_rate": 2.9785890058784756e-07,
+      "loss": 0.1291,
+      "step": 10091
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7677575212627107,
+      "learning_rate": 2.977644470305931e-07,
+      "loss": 0.1856,
+      "step": 10092
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.451153090437871,
+      "learning_rate": 2.9767000210111403e-07,
+      "loss": 0.2387,
+      "step": 10093
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.898565856327848,
+      "learning_rate": 2.9757556580343923e-07,
+      "loss": 0.3278,
+      "step": 10094
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8480122659879772,
+      "learning_rate": 2.9748113814159795e-07,
+      "loss": 0.2639,
+      "step": 10095
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3145698739956969,
+      "learning_rate": 2.9738671911961826e-07,
+      "loss": 0.0033,
+      "step": 10096
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0598471603856254,
+      "learning_rate": 2.9729230874152863e-07,
+      "loss": 0.0895,
+      "step": 10097
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3706323597084116,
+      "learning_rate": 2.9719790701135656e-07,
+      "loss": 0.0065,
+      "step": 10098
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 14.908254250704342,
+      "learning_rate": 2.9710351393312926e-07,
+      "loss": 0.2405,
+      "step": 10099
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7518823938104033,
+      "learning_rate": 2.970091295108741e-07,
+      "loss": 0.2358,
+      "step": 10100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7617448569859814,
+      "learning_rate": 2.9691475374861747e-07,
+      "loss": 0.4019,
+      "step": 10101
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.40944923641609,
+      "learning_rate": 2.968203866503857e-07,
+      "loss": 0.1179,
+      "step": 10102
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5376033138473534,
+      "learning_rate": 2.967260282202046e-07,
+      "loss": 0.1083,
+      "step": 10103
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6607908824701466,
+      "learning_rate": 2.9663167846209996e-07,
+      "loss": 0.1846,
+      "step": 10104
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5919673760987662,
+      "learning_rate": 2.9653733738009657e-07,
+      "loss": 0.1148,
+      "step": 10105
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1449331951659232,
+      "learning_rate": 2.964430049782195e-07,
+      "loss": 0.2059,
+      "step": 10106
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.113220674165706,
+      "learning_rate": 2.9634868126049287e-07,
+      "loss": 0.2069,
+      "step": 10107
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.432933815515144,
+      "learning_rate": 2.9625436623094113e-07,
+      "loss": 0.0945,
+      "step": 10108
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.412151247822407,
+      "learning_rate": 2.961600598935875e-07,
+      "loss": 0.1954,
+      "step": 10109
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.692059780788752,
+      "learning_rate": 2.960657622524556e-07,
+      "loss": 0.2773,
+      "step": 10110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4465622939612175,
+      "learning_rate": 2.959714733115681e-07,
+      "loss": 0.1709,
+      "step": 10111
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.035136004114922,
+      "learning_rate": 2.9587719307494787e-07,
+      "loss": 0.2606,
+      "step": 10112
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8765561261070781,
+      "learning_rate": 2.9578292154661696e-07,
+      "loss": 0.2269,
+      "step": 10113
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4644450559399976,
+      "learning_rate": 2.956886587305969e-07,
+      "loss": 0.1751,
+      "step": 10114
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8967843536064332,
+      "learning_rate": 2.9559440463090963e-07,
+      "loss": 0.149,
+      "step": 10115
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7431122814227695,
+      "learning_rate": 2.955001592515758e-07,
+      "loss": 0.1275,
+      "step": 10116
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1763049705796447,
+      "learning_rate": 2.9540592259661634e-07,
+      "loss": 0.1584,
+      "step": 10117
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8440220796152866,
+      "learning_rate": 2.9531169467005147e-07,
+      "loss": 0.109,
+      "step": 10118
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.238583783068823,
+      "learning_rate": 2.9521747547590114e-07,
+      "loss": 0.3639,
+      "step": 10119
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4512614817524705,
+      "learning_rate": 2.9512326501818505e-07,
+      "loss": 0.116,
+      "step": 10120
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8378187597679194,
+      "learning_rate": 2.9502906330092233e-07,
+      "loss": 0.0733,
+      "step": 10121
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2417661979508154,
+      "learning_rate": 2.949348703281317e-07,
+      "loss": 0.2177,
+      "step": 10122
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.989954464847215,
+      "learning_rate": 2.948406861038319e-07,
+      "loss": 0.3178,
+      "step": 10123
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.924334248056165,
+      "learning_rate": 2.947465106320407e-07,
+      "loss": 0.197,
+      "step": 10124
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9169590362837634,
+      "learning_rate": 2.9465234391677614e-07,
+      "loss": 0.2571,
+      "step": 10125
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.612481670372759,
+      "learning_rate": 2.945581859620554e-07,
+      "loss": 0.1421,
+      "step": 10126
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7364090504313137,
+      "learning_rate": 2.9446403677189523e-07,
+      "loss": 0.1013,
+      "step": 10127
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.755564132742967,
+      "learning_rate": 2.943698963503125e-07,
+      "loss": 0.1658,
+      "step": 10128
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1002501104171318,
+      "learning_rate": 2.942757647013233e-07,
+      "loss": 0.2312,
+      "step": 10129
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.8455882466354714,
+      "learning_rate": 2.941816418289438e-07,
+      "loss": 0.0073,
+      "step": 10130
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6175502859939063,
+      "learning_rate": 2.940875277371889e-07,
+      "loss": 0.0127,
+      "step": 10131
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3002325125406036,
+      "learning_rate": 2.9399342243007423e-07,
+      "loss": 0.2075,
+      "step": 10132
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6189822446095793,
+      "learning_rate": 2.938993259116141e-07,
+      "loss": 0.1336,
+      "step": 10133
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7771168105077464,
+      "learning_rate": 2.9380523818582325e-07,
+      "loss": 0.204,
+      "step": 10134
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.246312885201733,
+      "learning_rate": 2.9371115925671517e-07,
+      "loss": 0.2103,
+      "step": 10135
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9666412012084128,
+      "learning_rate": 2.9361708912830403e-07,
+      "loss": 0.0553,
+      "step": 10136
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0650677233362091,
+      "learning_rate": 2.935230278046025e-07,
+      "loss": 0.4619,
+      "step": 10137
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4880920623331828,
+      "learning_rate": 2.934289752896238e-07,
+      "loss": 0.1457,
+      "step": 10138
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7050596125092834,
+      "learning_rate": 2.9333493158738033e-07,
+      "loss": 0.2861,
+      "step": 10139
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.598019207815051,
+      "learning_rate": 2.9324089670188397e-07,
+      "loss": 0.2154,
+      "step": 10140
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3899083727109987,
+      "learning_rate": 2.931468706371468e-07,
+      "loss": 0.1716,
+      "step": 10141
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7443308974448324,
+      "learning_rate": 2.9305285339717964e-07,
+      "loss": 0.2573,
+      "step": 10142
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6111182814040098,
+      "learning_rate": 2.929588449859941e-07,
+      "loss": 0.0775,
+      "step": 10143
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6031787105155585,
+      "learning_rate": 2.9286484540760024e-07,
+      "loss": 0.174,
+      "step": 10144
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.059288171325278,
+      "learning_rate": 2.927708546660085e-07,
+      "loss": 0.156,
+      "step": 10145
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6155850848380845,
+      "learning_rate": 2.926768727652287e-07,
+      "loss": 0.2203,
+      "step": 10146
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 17.661522486178317,
+      "learning_rate": 2.925828997092703e-07,
+      "loss": 0.2288,
+      "step": 10147
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5909122067725574,
+      "learning_rate": 2.9248893550214225e-07,
+      "loss": 0.3839,
+      "step": 10148
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.663937159106112,
+      "learning_rate": 2.9239498014785357e-07,
+      "loss": 0.1143,
+      "step": 10149
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6793887715076052,
+      "learning_rate": 2.923010336504121e-07,
+      "loss": 0.3644,
+      "step": 10150
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7528713435368135,
+      "learning_rate": 2.9220709601382643e-07,
+      "loss": 0.1757,
+      "step": 10151
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.125473558541575,
+      "learning_rate": 2.9211316724210344e-07,
+      "loss": 0.0956,
+      "step": 10152
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3127818765202288,
+      "learning_rate": 2.920192473392509e-07,
+      "loss": 0.1067,
+      "step": 10153
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.837624615422415,
+      "learning_rate": 2.919253363092753e-07,
+      "loss": 0.231,
+      "step": 10154
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2875846712266474,
+      "learning_rate": 2.918314341561829e-07,
+      "loss": 0.1165,
+      "step": 10155
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4005460474096294,
+      "learning_rate": 2.9173754088398027e-07,
+      "loss": 0.1164,
+      "step": 10156
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.84564787952284,
+      "learning_rate": 2.9164365649667255e-07,
+      "loss": 0.3742,
+      "step": 10157
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7321126275754902,
+      "learning_rate": 2.915497809982653e-07,
+      "loss": 0.0675,
+      "step": 10158
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6524438418410831,
+      "learning_rate": 2.914559143927637e-07,
+      "loss": 0.1916,
+      "step": 10159
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7163088981075492,
+      "learning_rate": 2.913620566841718e-07,
+      "loss": 0.3322,
+      "step": 10160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5007082039915902,
+      "learning_rate": 2.9126820787649397e-07,
+      "loss": 0.199,
+      "step": 10161
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6585978086439676,
+      "learning_rate": 2.911743679737342e-07,
+      "loss": 0.0193,
+      "step": 10162
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1463803896353832,
+      "learning_rate": 2.9108053697989543e-07,
+      "loss": 0.2195,
+      "step": 10163
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0451761528139054,
+      "learning_rate": 2.9098671489898114e-07,
+      "loss": 0.0989,
+      "step": 10164
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.89734534298001,
+      "learning_rate": 2.908929017349936e-07,
+      "loss": 0.4223,
+      "step": 10165
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8270052705829065,
+      "learning_rate": 2.9079909749193544e-07,
+      "loss": 0.2895,
+      "step": 10166
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.364870594769816,
+      "learning_rate": 2.907053021738083e-07,
+      "loss": 0.1091,
+      "step": 10167
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.423504499918296,
+      "learning_rate": 2.906115157846135e-07,
+      "loss": 0.0939,
+      "step": 10168
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9835994607264935,
+      "learning_rate": 2.9051773832835257e-07,
+      "loss": 0.1069,
+      "step": 10169
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6078985931798069,
+      "learning_rate": 2.904239698090258e-07,
+      "loss": 0.2006,
+      "step": 10170
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8585419408299761,
+      "learning_rate": 2.9033021023063405e-07,
+      "loss": 0.0775,
+      "step": 10171
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9431968041486721,
+      "learning_rate": 2.9023645959717676e-07,
+      "loss": 0.1079,
+      "step": 10172
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2828754239310904,
+      "learning_rate": 2.90142717912654e-07,
+      "loss": 0.0871,
+      "step": 10173
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.38468142708540887,
+      "learning_rate": 2.9004898518106457e-07,
+      "loss": 0.1511,
+      "step": 10174
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8713158391269873,
+      "learning_rate": 2.899552614064077e-07,
+      "loss": 0.115,
+      "step": 10175
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.176435073014773,
+      "learning_rate": 2.8986154659268137e-07,
+      "loss": 0.1554,
+      "step": 10176
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0177193142856265,
+      "learning_rate": 2.8976784074388395e-07,
+      "loss": 0.4084,
+      "step": 10177
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.369229236259131,
+      "learning_rate": 2.896741438640132e-07,
+      "loss": 0.2152,
+      "step": 10178
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4581195734418044,
+      "learning_rate": 2.8958045595706617e-07,
+      "loss": 0.1189,
+      "step": 10179
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9939602725598458,
+      "learning_rate": 2.894867770270398e-07,
+      "loss": 0.111,
+      "step": 10180
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5609244693542784,
+      "learning_rate": 2.8939310707793097e-07,
+      "loss": 0.1326,
+      "step": 10181
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1259466010058046,
+      "learning_rate": 2.8929944611373554e-07,
+      "loss": 0.2838,
+      "step": 10182
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.833719703385876,
+      "learning_rate": 2.8920579413844904e-07,
+      "loss": 0.0472,
+      "step": 10183
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6790088170753138,
+      "learning_rate": 2.891121511560674e-07,
+      "loss": 0.1817,
+      "step": 10184
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0421877559803852,
+      "learning_rate": 2.8901851717058513e-07,
+      "loss": 0.1087,
+      "step": 10185
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5242384686078717,
+      "learning_rate": 2.889248921859972e-07,
+      "loss": 0.3666,
+      "step": 10186
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4096621419743181,
+      "learning_rate": 2.888312762062974e-07,
+      "loss": 0.0943,
+      "step": 10187
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1403917586988235,
+      "learning_rate": 2.887376692354803e-07,
+      "loss": 0.4638,
+      "step": 10188
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1251275356077421,
+      "learning_rate": 2.886440712775385e-07,
+      "loss": 0.2976,
+      "step": 10189
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.40686203098349166,
+      "learning_rate": 2.8855048233646576e-07,
+      "loss": 0.3208,
+      "step": 10190
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.81965736341543,
+      "learning_rate": 2.884569024162543e-07,
+      "loss": 0.3933,
+      "step": 10191
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6045867870038998,
+      "learning_rate": 2.8836333152089687e-07,
+      "loss": 0.1915,
+      "step": 10192
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6880109823058511,
+      "learning_rate": 2.88269769654385e-07,
+      "loss": 0.3868,
+      "step": 10193
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0855868710341854,
+      "learning_rate": 2.881762168207105e-07,
+      "loss": 0.1405,
+      "step": 10194
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7032886920100117,
+      "learning_rate": 2.8808267302386423e-07,
+      "loss": 0.2338,
+      "step": 10195
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7592264969634548,
+      "learning_rate": 2.8798913826783724e-07,
+      "loss": 0.2576,
+      "step": 10196
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3478853546872511,
+      "learning_rate": 2.8789561255661986e-07,
+      "loss": 0.1196,
+      "step": 10197
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6906784802526151,
+      "learning_rate": 2.878020958942019e-07,
+      "loss": 0.2686,
+      "step": 10198
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7904765996454155,
+      "learning_rate": 2.8770858828457336e-07,
+      "loss": 0.1888,
+      "step": 10199
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5154334002839001,
+      "learning_rate": 2.8761508973172286e-07,
+      "loss": 0.1782,
+      "step": 10200
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3354066006994746,
+      "learning_rate": 2.875216002396399e-07,
+      "loss": 0.1028,
+      "step": 10201
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.444505493427869,
+      "learning_rate": 2.8742811981231236e-07,
+      "loss": 0.0521,
+      "step": 10202
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.47817889508201,
+      "learning_rate": 2.873346484537288e-07,
+      "loss": 0.298,
+      "step": 10203
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6336947599145452,
+      "learning_rate": 2.8724118616787636e-07,
+      "loss": 0.5221,
+      "step": 10204
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.928730037937413,
+      "learning_rate": 2.8714773295874283e-07,
+      "loss": 0.3397,
+      "step": 10205
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.858571369987186,
+      "learning_rate": 2.870542888303148e-07,
+      "loss": 0.0332,
+      "step": 10206
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.078718939124812,
+      "learning_rate": 2.86960853786579e-07,
+      "loss": 0.286,
+      "step": 10207
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1468493477250459,
+      "learning_rate": 2.868674278315214e-07,
+      "loss": 0.139,
+      "step": 10208
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4781322692514693,
+      "learning_rate": 2.8677401096912767e-07,
+      "loss": 0.268,
+      "step": 10209
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.9173659262253786,
+      "learning_rate": 2.8668060320338345e-07,
+      "loss": 0.2243,
+      "step": 10210
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2222157176630717,
+      "learning_rate": 2.865872045382733e-07,
+      "loss": 0.3454,
+      "step": 10211
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3029370794229205,
+      "learning_rate": 2.86493814977782e-07,
+      "loss": 0.0399,
+      "step": 10212
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.468774882671728,
+      "learning_rate": 2.864004345258938e-07,
+      "loss": 0.294,
+      "step": 10213
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6157720857306135,
+      "learning_rate": 2.863070631865926e-07,
+      "loss": 0.3866,
+      "step": 10214
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.782503562281586,
+      "learning_rate": 2.8621370096386155e-07,
+      "loss": 0.3746,
+      "step": 10215
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.248339048644885,
+      "learning_rate": 2.861203478616839e-07,
+      "loss": 0.2226,
+      "step": 10216
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3511410845734303,
+      "learning_rate": 2.86027003884042e-07,
+      "loss": 0.2554,
+      "step": 10217
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1966562512631873,
+      "learning_rate": 2.8593366903491845e-07,
+      "loss": 0.2706,
+      "step": 10218
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.47867191486930216,
+      "learning_rate": 2.8584034331829465e-07,
+      "loss": 0.066,
+      "step": 10219
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5339949316047048,
+      "learning_rate": 2.8574702673815257e-07,
+      "loss": 0.12,
+      "step": 10220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5611385857681584,
+      "learning_rate": 2.856537192984728e-07,
+      "loss": 0.1945,
+      "step": 10221
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7098057171282469,
+      "learning_rate": 2.8556042100323653e-07,
+      "loss": 0.2918,
+      "step": 10222
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.142105614411486,
+      "learning_rate": 2.854671318564237e-07,
+      "loss": 0.1315,
+      "step": 10223
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.35340259488248843,
+      "learning_rate": 2.853738518620141e-07,
+      "loss": 0.0505,
+      "step": 10224
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6201995876999788,
+      "learning_rate": 2.8528058102398767e-07,
+      "loss": 0.2407,
+      "step": 10225
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2426319278333497,
+      "learning_rate": 2.851873193463231e-07,
+      "loss": 0.3811,
+      "step": 10226
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9231925775715792,
+      "learning_rate": 2.850940668329995e-07,
+      "loss": 0.1911,
+      "step": 10227
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.683992341126847,
+      "learning_rate": 2.8500082348799484e-07,
+      "loss": 0.1405,
+      "step": 10228
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.030114905525364,
+      "learning_rate": 2.8490758931528733e-07,
+      "loss": 0.3582,
+      "step": 10229
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5730900085900599,
+      "learning_rate": 2.848143643188544e-07,
+      "loss": 0.2923,
+      "step": 10230
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3130173591499996,
+      "learning_rate": 2.847211485026732e-07,
+      "loss": 0.1366,
+      "step": 10231
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6563485023553393,
+      "learning_rate": 2.8462794187072056e-07,
+      "loss": 0.1694,
+      "step": 10232
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.513843404179402,
+      "learning_rate": 2.8453474442697313e-07,
+      "loss": 0.0761,
+      "step": 10233
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4821740985558005,
+      "learning_rate": 2.8444155617540645e-07,
+      "loss": 0.3699,
+      "step": 10234
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.937843973931371,
+      "learning_rate": 2.843483771199964e-07,
+      "loss": 0.2367,
+      "step": 10235
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6188196116057207,
+      "learning_rate": 2.842552072647182e-07,
+      "loss": 0.3673,
+      "step": 10236
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.668748716083185,
+      "learning_rate": 2.8416204661354634e-07,
+      "loss": 0.1505,
+      "step": 10237
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6595810651174245,
+      "learning_rate": 2.8406889517045563e-07,
+      "loss": 0.3319,
+      "step": 10238
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.41065654226966897,
+      "learning_rate": 2.839757529394197e-07,
+      "loss": 0.1962,
+      "step": 10239
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3955031550423198,
+      "learning_rate": 2.8388261992441263e-07,
+      "loss": 0.1096,
+      "step": 10240
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6508463178007073,
+      "learning_rate": 2.8378949612940726e-07,
+      "loss": 0.3858,
+      "step": 10241
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.48423807947557324,
+      "learning_rate": 2.836963815583769e-07,
+      "loss": 0.1474,
+      "step": 10242
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8928769138359334,
+      "learning_rate": 2.8360327621529343e-07,
+      "loss": 0.0216,
+      "step": 10243
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.42743112036302267,
+      "learning_rate": 2.835101801041294e-07,
+      "loss": 0.142,
+      "step": 10244
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0621584090643938,
+      "learning_rate": 2.834170932288562e-07,
+      "loss": 0.1203,
+      "step": 10245
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4533056235972029,
+      "learning_rate": 2.8332401559344534e-07,
+      "loss": 0.0911,
+      "step": 10246
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3373391990105874,
+      "learning_rate": 2.8323094720186724e-07,
+      "loss": 0.2234,
+      "step": 10247
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.23324687665964133,
+      "learning_rate": 2.831378880580928e-07,
+      "loss": 0.0763,
+      "step": 10248
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6111975212027193,
+      "learning_rate": 2.8304483816609215e-07,
+      "loss": 0.1486,
+      "step": 10249
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4004848741687492,
+      "learning_rate": 2.8295179752983466e-07,
+      "loss": 0.1131,
+      "step": 10250
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6956776664200824,
+      "learning_rate": 2.828587661532901e-07,
+      "loss": 0.1478,
+      "step": 10251
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8217026929838352,
+      "learning_rate": 2.8276574404042677e-07,
+      "loss": 0.1402,
+      "step": 10252
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6513064563950473,
+      "learning_rate": 2.8267273119521365e-07,
+      "loss": 0.2293,
+      "step": 10253
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6472530386355813,
+      "learning_rate": 2.8257972762161863e-07,
+      "loss": 0.3021,
+      "step": 10254
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7732910536516439,
+      "learning_rate": 2.824867333236096e-07,
+      "loss": 0.1998,
+      "step": 10255
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6853220452332898,
+      "learning_rate": 2.823937483051536e-07,
+      "loss": 0.1586,
+      "step": 10256
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6189092250336454,
+      "learning_rate": 2.82300772570218e-07,
+      "loss": 0.3721,
+      "step": 10257
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3897431660961344,
+      "learning_rate": 2.8220780612276887e-07,
+      "loss": 0.2036,
+      "step": 10258
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.048638907755914,
+      "learning_rate": 2.8211484896677274e-07,
+      "loss": 0.1098,
+      "step": 10259
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.022173529418032,
+      "learning_rate": 2.820219011061949e-07,
+      "loss": 0.2273,
+      "step": 10260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.48019927111320554,
+      "learning_rate": 2.819289625450012e-07,
+      "loss": 0.3013,
+      "step": 10261
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3915675359680851,
+      "learning_rate": 2.8183603328715613e-07,
+      "loss": 0.248,
+      "step": 10262
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3131145352365399,
+      "learning_rate": 2.817431133366246e-07,
+      "loss": 0.1545,
+      "step": 10263
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.41835883686822495,
+      "learning_rate": 2.8165020269737046e-07,
+      "loss": 0.0843,
+      "step": 10264
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.489341998654538,
+      "learning_rate": 2.8155730137335777e-07,
+      "loss": 0.0821,
+      "step": 10265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.31110884249060594,
+      "learning_rate": 2.8146440936854953e-07,
+      "loss": 0.1006,
+      "step": 10266
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2558649006924264,
+      "learning_rate": 2.813715266869089e-07,
+      "loss": 0.0826,
+      "step": 10267
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.138848363680206,
+      "learning_rate": 2.812786533323987e-07,
+      "loss": 0.1607,
+      "step": 10268
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3580398835841955,
+      "learning_rate": 2.811857893089806e-07,
+      "loss": 0.0585,
+      "step": 10269
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5828412543228145,
+      "learning_rate": 2.810929346206168e-07,
+      "loss": 0.1878,
+      "step": 10270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.606115308601487,
+      "learning_rate": 2.810000892712682e-07,
+      "loss": 0.1429,
+      "step": 10271
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.22420079108152827,
+      "learning_rate": 2.809072532648963e-07,
+      "loss": 0.1768,
+      "step": 10272
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.04813094441474049,
+      "learning_rate": 2.808144266054612e-07,
+      "loss": 0.0009,
+      "step": 10273
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.47288197730564774,
+      "learning_rate": 2.8072160929692354e-07,
+      "loss": 0.1842,
+      "step": 10274
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.738235164841927,
+      "learning_rate": 2.806288013432425e-07,
+      "loss": 0.1385,
+      "step": 10275
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7262208566464031,
+      "learning_rate": 2.8053600274837807e-07,
+      "loss": 0.1382,
+      "step": 10276
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 21.905346040021723,
+      "learning_rate": 2.8044321351628897e-07,
+      "loss": 0.2486,
+      "step": 10277
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7732430553059354,
+      "learning_rate": 2.8035043365093347e-07,
+      "loss": 0.2504,
+      "step": 10278
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.705032161380198,
+      "learning_rate": 2.802576631562703e-07,
+      "loss": 0.0047,
+      "step": 10279
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6248934088142746,
+      "learning_rate": 2.801649020362567e-07,
+      "loss": 0.0321,
+      "step": 10280
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5827335660059887,
+      "learning_rate": 2.8007215029485054e-07,
+      "loss": 0.0608,
+      "step": 10281
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8611950012380255,
+      "learning_rate": 2.799794079360083e-07,
+      "loss": 0.18,
+      "step": 10282
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7082351348512694,
+      "learning_rate": 2.798866749636869e-07,
+      "loss": 0.076,
+      "step": 10283
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9008111053865964,
+      "learning_rate": 2.797939513818426e-07,
+      "loss": 0.2403,
+      "step": 10284
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8943813175106697,
+      "learning_rate": 2.7970123719443073e-07,
+      "loss": 0.1797,
+      "step": 10285
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8584149610242897,
+      "learning_rate": 2.79608532405407e-07,
+      "loss": 0.0888,
+      "step": 10286
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 15.404306424246116,
+      "learning_rate": 2.795158370187265e-07,
+      "loss": 0.1976,
+      "step": 10287
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7313523866606231,
+      "learning_rate": 2.794231510383435e-07,
+      "loss": 0.2239,
+      "step": 10288
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 29.744310653057806,
+      "learning_rate": 2.7933047446821236e-07,
+      "loss": 0.1484,
+      "step": 10289
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4091252632914925,
+      "learning_rate": 2.7923780731228664e-07,
+      "loss": 0.3602,
+      "step": 10290
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.17125294876779928,
+      "learning_rate": 2.7914514957452004e-07,
+      "loss": 0.0044,
+      "step": 10291
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8727869656872513,
+      "learning_rate": 2.7905250125886536e-07,
+      "loss": 0.33,
+      "step": 10292
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.40791427587481266,
+      "learning_rate": 2.7895986236927495e-07,
+      "loss": 0.205,
+      "step": 10293
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.808736533405036,
+      "learning_rate": 2.7886723290970134e-07,
+      "loss": 0.3945,
+      "step": 10294
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6901055506409601,
+      "learning_rate": 2.787746128840959e-07,
+      "loss": 0.1194,
+      "step": 10295
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5673610143798768,
+      "learning_rate": 2.7868200229641035e-07,
+      "loss": 0.2125,
+      "step": 10296
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 25.81391849887281,
+      "learning_rate": 2.7858940115059537e-07,
+      "loss": 0.375,
+      "step": 10297
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9861928424107715,
+      "learning_rate": 2.7849680945060176e-07,
+      "loss": 0.0967,
+      "step": 10298
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.45214295832004353,
+      "learning_rate": 2.784042272003794e-07,
+      "loss": 0.1678,
+      "step": 10299
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.084999990659764,
+      "learning_rate": 2.7831165440387825e-07,
+      "loss": 0.3066,
+      "step": 10300
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2193244138626602,
+      "learning_rate": 2.7821909106504747e-07,
+      "loss": 0.3322,
+      "step": 10301
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.810876479458508,
+      "learning_rate": 2.781265371878361e-07,
+      "loss": 0.1077,
+      "step": 10302
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3503260080641253,
+      "learning_rate": 2.7803399277619265e-07,
+      "loss": 0.3605,
+      "step": 10303
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.420953137141911,
+      "learning_rate": 2.779414578340655e-07,
+      "loss": 0.1638,
+      "step": 10304
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8445873763051406,
+      "learning_rate": 2.778489323654022e-07,
+      "loss": 0.2076,
+      "step": 10305
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8129850797536065,
+      "learning_rate": 2.777564163741497e-07,
+      "loss": 0.1513,
+      "step": 10306
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4345052749212925,
+      "learning_rate": 2.7766390986425557e-07,
+      "loss": 0.2743,
+      "step": 10307
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.608516525244101,
+      "learning_rate": 2.775714128396657e-07,
+      "loss": 0.3439,
+      "step": 10308
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3602859943341306,
+      "learning_rate": 2.7747892530432675e-07,
+      "loss": 0.1154,
+      "step": 10309
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8412252413073988,
+      "learning_rate": 2.7738644726218396e-07,
+      "loss": 0.0925,
+      "step": 10310
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4230985042881405,
+      "learning_rate": 2.77293978717183e-07,
+      "loss": 0.0543,
+      "step": 10311
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.730359179745662,
+      "learning_rate": 2.7720151967326845e-07,
+      "loss": 0.083,
+      "step": 10312
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1155228647285207,
+      "learning_rate": 2.771090701343851e-07,
+      "loss": 0.4335,
+      "step": 10313
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6157355401303554,
+      "learning_rate": 2.7701663010447673e-07,
+      "loss": 0.099,
+      "step": 10314
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.607567660844606,
+      "learning_rate": 2.7692419958748734e-07,
+      "loss": 0.0942,
+      "step": 10315
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7522680907313777,
+      "learning_rate": 2.7683177858736e-07,
+      "loss": 0.1124,
+      "step": 10316
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.149502614901415,
+      "learning_rate": 2.767393671080376e-07,
+      "loss": 0.1974,
+      "step": 10317
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7498093110361993,
+      "learning_rate": 2.766469651534624e-07,
+      "loss": 0.2381,
+      "step": 10318
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6138349791888598,
+      "learning_rate": 2.765545727275768e-07,
+      "loss": 0.1838,
+      "step": 10319
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8325056592051004,
+      "learning_rate": 2.7646218983432245e-07,
+      "loss": 0.2273,
+      "step": 10320
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5471731317832644,
+      "learning_rate": 2.7636981647764024e-07,
+      "loss": 0.1571,
+      "step": 10321
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.39690254731037833,
+      "learning_rate": 2.762774526614714e-07,
+      "loss": 0.0376,
+      "step": 10322
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.24004478182410177,
+      "learning_rate": 2.76185098389756e-07,
+      "loss": 0.1198,
+      "step": 10323
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.3566657921106415,
+      "learning_rate": 2.760927536664344e-07,
+      "loss": 0.2131,
+      "step": 10324
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9512974098221918,
+      "learning_rate": 2.7600041849544584e-07,
+      "loss": 0.1373,
+      "step": 10325
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7863187991648417,
+      "learning_rate": 2.7590809288073e-07,
+      "loss": 0.1968,
+      "step": 10326
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.788712920122992,
+      "learning_rate": 2.7581577682622513e-07,
+      "loss": 0.2357,
+      "step": 10327
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.284394747997247,
+      "learning_rate": 2.757234703358701e-07,
+      "loss": 0.1383,
+      "step": 10328
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.363789118705277,
+      "learning_rate": 2.7563117341360244e-07,
+      "loss": 0.0422,
+      "step": 10329
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.31012208836261174,
+      "learning_rate": 2.7553888606336016e-07,
+      "loss": 0.0392,
+      "step": 10330
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7872579617347853,
+      "learning_rate": 2.754466082890801e-07,
+      "loss": 0.0694,
+      "step": 10331
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.14947528566196186,
+      "learning_rate": 2.7535434009469924e-07,
+      "loss": 0.0021,
+      "step": 10332
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.8144802725901945,
+      "learning_rate": 2.7526208148415394e-07,
+      "loss": 0.1461,
+      "step": 10333
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3009842356808723,
+      "learning_rate": 2.7516983246137977e-07,
+      "loss": 0.3036,
+      "step": 10334
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9552782663324134,
+      "learning_rate": 2.750775930303125e-07,
+      "loss": 0.3527,
+      "step": 10335
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1074123808822676,
+      "learning_rate": 2.749853631948875e-07,
+      "loss": 0.1009,
+      "step": 10336
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7717845227744466,
+      "learning_rate": 2.74893142959039e-07,
+      "loss": 0.1866,
+      "step": 10337
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.847477736795236,
+      "learning_rate": 2.7480093232670155e-07,
+      "loss": 0.0834,
+      "step": 10338
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8115874522348948,
+      "learning_rate": 2.747087313018092e-07,
+      "loss": 0.29,
+      "step": 10339
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3672825170076108,
+      "learning_rate": 2.7461653988829515e-07,
+      "loss": 0.1825,
+      "step": 10340
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8866651351700062,
+      "learning_rate": 2.745243580900927e-07,
+      "loss": 0.2698,
+      "step": 10341
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5706974465252797,
+      "learning_rate": 2.7443218591113427e-07,
+      "loss": 0.2585,
+      "step": 10342
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6401080035984237,
+      "learning_rate": 2.7434002335535233e-07,
+      "loss": 0.5122,
+      "step": 10343
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0598099448529625,
+      "learning_rate": 2.742478704266785e-07,
+      "loss": 0.152,
+      "step": 10344
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.573205755417549,
+      "learning_rate": 2.7415572712904454e-07,
+      "loss": 0.0744,
+      "step": 10345
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9795329057276252,
+      "learning_rate": 2.740635934663813e-07,
+      "loss": 0.1945,
+      "step": 10346
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.036694928135541,
+      "learning_rate": 2.739714694426191e-07,
+      "loss": 0.1459,
+      "step": 10347
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1962920417360572,
+      "learning_rate": 2.7387935506168857e-07,
+      "loss": 0.1081,
+      "step": 10348
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.047942062306148,
+      "learning_rate": 2.7378725032751915e-07,
+      "loss": 0.1798,
+      "step": 10349
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8520817851554787,
+      "learning_rate": 2.7369515524404064e-07,
+      "loss": 0.4585,
+      "step": 10350
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.03077299174891,
+      "learning_rate": 2.7360306981518147e-07,
+      "loss": 0.3425,
+      "step": 10351
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4387898763168881,
+      "learning_rate": 2.735109940448706e-07,
+      "loss": 0.0154,
+      "step": 10352
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6529751277320938,
+      "learning_rate": 2.734189279370359e-07,
+      "loss": 0.1431,
+      "step": 10353
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8642363768247914,
+      "learning_rate": 2.733268714956052e-07,
+      "loss": 0.1578,
+      "step": 10354
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5916966740097632,
+      "learning_rate": 2.7323482472450597e-07,
+      "loss": 0.3505,
+      "step": 10355
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9445762314627908,
+      "learning_rate": 2.7314278762766483e-07,
+      "loss": 0.3951,
+      "step": 10356
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7871275307321242,
+      "learning_rate": 2.7305076020900843e-07,
+      "loss": 0.2309,
+      "step": 10357
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1404180942625537,
+      "learning_rate": 2.7295874247246303e-07,
+      "loss": 0.1901,
+      "step": 10358
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4365060851096445,
+      "learning_rate": 2.7286673442195406e-07,
+      "loss": 0.1272,
+      "step": 10359
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6029931644642945,
+      "learning_rate": 2.727747360614066e-07,
+      "loss": 0.1351,
+      "step": 10360
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.47347890544347404,
+      "learning_rate": 2.726827473947458e-07,
+      "loss": 0.3076,
+      "step": 10361
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.47592070419616117,
+      "learning_rate": 2.725907684258959e-07,
+      "loss": 0.0055,
+      "step": 10362
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.371716287788669,
+      "learning_rate": 2.7249879915878117e-07,
+      "loss": 0.0995,
+      "step": 10363
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.834889662434467,
+      "learning_rate": 2.724068395973247e-07,
+      "loss": 0.2137,
+      "step": 10364
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8000913131057464,
+      "learning_rate": 2.7231488974545017e-07,
+      "loss": 0.2315,
+      "step": 10365
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.564255357468439,
+      "learning_rate": 2.7222294960708004e-07,
+      "loss": 0.0632,
+      "step": 10366
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7205279558390081,
+      "learning_rate": 2.721310191861369e-07,
+      "loss": 0.1315,
+      "step": 10367
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7295300427974798,
+      "learning_rate": 2.720390984865424e-07,
+      "loss": 0.2025,
+      "step": 10368
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3406120790995233,
+      "learning_rate": 2.7194718751221836e-07,
+      "loss": 0.3642,
+      "step": 10369
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.258689744960688,
+      "learning_rate": 2.7185528626708553e-07,
+      "loss": 0.1006,
+      "step": 10370
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.43996053820520636,
+      "learning_rate": 2.717633947550651e-07,
+      "loss": 0.1191,
+      "step": 10371
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4180444976354502,
+      "learning_rate": 2.7167151298007686e-07,
+      "loss": 0.0946,
+      "step": 10372
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3772174713897463,
+      "learning_rate": 2.715796409460408e-07,
+      "loss": 0.1083,
+      "step": 10373
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7817672051214644,
+      "learning_rate": 2.714877786568767e-07,
+      "loss": 0.395,
+      "step": 10374
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.325419730471386,
+      "learning_rate": 2.713959261165031e-07,
+      "loss": 0.2452,
+      "step": 10375
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5344302760002922,
+      "learning_rate": 2.7130408332883904e-07,
+      "loss": 0.1106,
+      "step": 10376
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9452148141559257,
+      "learning_rate": 2.7121225029780235e-07,
+      "loss": 0.2956,
+      "step": 10377
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8962247123977751,
+      "learning_rate": 2.7112042702731117e-07,
+      "loss": 0.1983,
+      "step": 10378
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8331461827843374,
+      "learning_rate": 2.7102861352128246e-07,
+      "loss": 0.4281,
+      "step": 10379
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9961725351184338,
+      "learning_rate": 2.709368097836336e-07,
+      "loss": 0.4122,
+      "step": 10380
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1622811223499467,
+      "learning_rate": 2.708450158182807e-07,
+      "loss": 0.0962,
+      "step": 10381
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5044438682947123,
+      "learning_rate": 2.707532316291403e-07,
+      "loss": 0.1427,
+      "step": 10382
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0408060350720334,
+      "learning_rate": 2.7066145722012767e-07,
+      "loss": 0.1227,
+      "step": 10383
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9981343484997779,
+      "learning_rate": 2.7056969259515856e-07,
+      "loss": 0.3258,
+      "step": 10384
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.124303491774027,
+      "learning_rate": 2.704779377581473e-07,
+      "loss": 0.1624,
+      "step": 10385
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8375855156700935,
+      "learning_rate": 2.7038619271300886e-07,
+      "loss": 0.5048,
+      "step": 10386
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3768231972231354,
+      "learning_rate": 2.70294457463657e-07,
+      "loss": 0.1078,
+      "step": 10387
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6236407902616248,
+      "learning_rate": 2.7020273201400514e-07,
+      "loss": 0.282,
+      "step": 10388
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8953272419992317,
+      "learning_rate": 2.7011101636796674e-07,
+      "loss": 0.3133,
+      "step": 10389
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.074712334373866,
+      "learning_rate": 2.700193105294545e-07,
+      "loss": 0.2425,
+      "step": 10390
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7647295096150315,
+      "learning_rate": 2.6992761450238086e-07,
+      "loss": 0.1282,
+      "step": 10391
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6549380344787635,
+      "learning_rate": 2.6983592829065765e-07,
+      "loss": 0.1558,
+      "step": 10392
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.145107928758458,
+      "learning_rate": 2.697442518981966e-07,
+      "loss": 0.2234,
+      "step": 10393
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.47280586014705855,
+      "learning_rate": 2.696525853289084e-07,
+      "loss": 0.057,
+      "step": 10394
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5057971735520573,
+      "learning_rate": 2.6956092858670416e-07,
+      "loss": 0.0897,
+      "step": 10395
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2933726887911243,
+      "learning_rate": 2.6946928167549377e-07,
+      "loss": 0.1144,
+      "step": 10396
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4718624220367011,
+      "learning_rate": 2.693776445991874e-07,
+      "loss": 0.0291,
+      "step": 10397
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0816261079841452,
+      "learning_rate": 2.692860173616942e-07,
+      "loss": 0.1048,
+      "step": 10398
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4821952474066926,
+      "learning_rate": 2.6919439996692343e-07,
+      "loss": 0.1043,
+      "step": 10399
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3739327583129264,
+      "learning_rate": 2.6910279241878337e-07,
+      "loss": 0.2589,
+      "step": 10400
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6883030170882999,
+      "learning_rate": 2.690111947211825e-07,
+      "loss": 0.173,
+      "step": 10401
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5572504594788301,
+      "learning_rate": 2.689196068780285e-07,
+      "loss": 0.2161,
+      "step": 10402
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8375524659754772,
+      "learning_rate": 2.688280288932283e-07,
+      "loss": 0.044,
+      "step": 10403
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1238918985541626,
+      "learning_rate": 2.687364607706893e-07,
+      "loss": 0.2673,
+      "step": 10404
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9614332658496733,
+      "learning_rate": 2.686449025143177e-07,
+      "loss": 0.2368,
+      "step": 10405
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.139127536266855,
+      "learning_rate": 2.685533541280197e-07,
+      "loss": 0.2419,
+      "step": 10406
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9242720014787659,
+      "learning_rate": 2.684618156157008e-07,
+      "loss": 0.3948,
+      "step": 10407
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.721545218688634,
+      "learning_rate": 2.683702869812662e-07,
+      "loss": 0.0945,
+      "step": 10408
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 15.050138743858733,
+      "learning_rate": 2.682787682286209e-07,
+      "loss": 0.1971,
+      "step": 10409
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5187178695507517,
+      "learning_rate": 2.681872593616693e-07,
+      "loss": 0.2237,
+      "step": 10410
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.28910169011590614,
+      "learning_rate": 2.6809576038431504e-07,
+      "loss": 0.0723,
+      "step": 10411
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7739970350627858,
+      "learning_rate": 2.68004271300462e-07,
+      "loss": 0.3179,
+      "step": 10412
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8477935669559543,
+      "learning_rate": 2.6791279211401284e-07,
+      "loss": 0.3444,
+      "step": 10413
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3818642354491543,
+      "learning_rate": 2.678213228288708e-07,
+      "loss": 0.0815,
+      "step": 10414
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.528649656653528,
+      "learning_rate": 2.677298634489379e-07,
+      "loss": 0.3162,
+      "step": 10415
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9286483054980392,
+      "learning_rate": 2.676384139781157e-07,
+      "loss": 0.0572,
+      "step": 10416
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7625097904677913,
+      "learning_rate": 2.675469744203061e-07,
+      "loss": 0.3784,
+      "step": 10417
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2864610662133286,
+      "learning_rate": 2.6745554477940967e-07,
+      "loss": 0.0046,
+      "step": 10418
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6962038626822891,
+      "learning_rate": 2.6736412505932737e-07,
+      "loss": 0.2113,
+      "step": 10419
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0503986618584422,
+      "learning_rate": 2.672727152639589e-07,
+      "loss": 0.1527,
+      "step": 10420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8225215419075138,
+      "learning_rate": 2.6718131539720445e-07,
+      "loss": 0.0783,
+      "step": 10421
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5404830518009177,
+      "learning_rate": 2.6708992546296294e-07,
+      "loss": 0.1638,
+      "step": 10422
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9482240485776333,
+      "learning_rate": 2.6699854546513354e-07,
+      "loss": 0.2334,
+      "step": 10423
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5889709783066235,
+      "learning_rate": 2.6690717540761446e-07,
+      "loss": 0.293,
+      "step": 10424
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.558957254022895,
+      "learning_rate": 2.6681581529430384e-07,
+      "loss": 0.1498,
+      "step": 10425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.316452324124796,
+      "learning_rate": 2.667244651290995e-07,
+      "loss": 0.0172,
+      "step": 10426
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0084936543747833,
+      "learning_rate": 2.666331249158982e-07,
+      "loss": 0.0781,
+      "step": 10427
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1415900491388742,
+      "learning_rate": 2.665417946585972e-07,
+      "loss": 0.3482,
+      "step": 10428
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.230059651396525,
+      "learning_rate": 2.6645047436109224e-07,
+      "loss": 0.1471,
+      "step": 10429
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6010759257443067,
+      "learning_rate": 2.6635916402727975e-07,
+      "loss": 0.1955,
+      "step": 10430
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5991172116467856,
+      "learning_rate": 2.662678636610549e-07,
+      "loss": 0.4037,
+      "step": 10431
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3643282391243785,
+      "learning_rate": 2.6617657326631296e-07,
+      "loss": 0.1668,
+      "step": 10432
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5806538082505955,
+      "learning_rate": 2.6608529284694823e-07,
+      "loss": 0.1467,
+      "step": 10433
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6803427975121148,
+      "learning_rate": 2.6599402240685543e-07,
+      "loss": 0.0592,
+      "step": 10434
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2480218245656194,
+      "learning_rate": 2.659027619499278e-07,
+      "loss": 0.1108,
+      "step": 10435
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7961858815220818,
+      "learning_rate": 2.658115114800591e-07,
+      "loss": 0.2516,
+      "step": 10436
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9066160613403413,
+      "learning_rate": 2.657202710011419e-07,
+      "loss": 0.2253,
+      "step": 10437
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6353668729752076,
+      "learning_rate": 2.6562904051706913e-07,
+      "loss": 0.1116,
+      "step": 10438
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8542947566658707,
+      "learning_rate": 2.655378200317324e-07,
+      "loss": 0.1967,
+      "step": 10439
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6447993647207726,
+      "learning_rate": 2.6544660954902385e-07,
+      "loss": 0.2617,
+      "step": 10440
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8252586569785209,
+      "learning_rate": 2.653554090728342e-07,
+      "loss": 0.2064,
+      "step": 10441
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.551786203138509,
+      "learning_rate": 2.6526421860705473e-07,
+      "loss": 0.1988,
+      "step": 10442
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.888358863306191,
+      "learning_rate": 2.6517303815557533e-07,
+      "loss": 0.168,
+      "step": 10443
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.39348950210940503,
+      "learning_rate": 2.6508186772228626e-07,
+      "loss": 0.2157,
+      "step": 10444
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9599256798437353,
+      "learning_rate": 2.649907073110771e-07,
+      "loss": 0.2491,
+      "step": 10445
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7617142381150503,
+      "learning_rate": 2.648995569258366e-07,
+      "loss": 0.4391,
+      "step": 10446
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.397491569318164,
+      "learning_rate": 2.6480841657045384e-07,
+      "loss": 0.0671,
+      "step": 10447
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3137672202187631,
+      "learning_rate": 2.6471728624881657e-07,
+      "loss": 0.2771,
+      "step": 10448
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9264509751527862,
+      "learning_rate": 2.6462616596481306e-07,
+      "loss": 0.1368,
+      "step": 10449
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1262196550745276,
+      "learning_rate": 2.645350557223303e-07,
+      "loss": 0.0926,
+      "step": 10450
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6099039134737922,
+      "learning_rate": 2.6444395552525556e-07,
+      "loss": 0.1469,
+      "step": 10451
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7466957962216927,
+      "learning_rate": 2.6435286537747507e-07,
+      "loss": 0.1395,
+      "step": 10452
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9883365727329495,
+      "learning_rate": 2.6426178528287526e-07,
+      "loss": 0.1304,
+      "step": 10453
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7112658868253172,
+      "learning_rate": 2.641707152453413e-07,
+      "loss": 0.1583,
+      "step": 10454
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.708473714266294,
+      "learning_rate": 2.64079655268759e-07,
+      "loss": 0.0373,
+      "step": 10455
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.028738953398234,
+      "learning_rate": 2.6398860535701273e-07,
+      "loss": 0.167,
+      "step": 10456
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0053975009335385,
+      "learning_rate": 2.638975655139869e-07,
+      "loss": 0.0174,
+      "step": 10457
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2956561433469352,
+      "learning_rate": 2.6380653574356576e-07,
+      "loss": 0.0927,
+      "step": 10458
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7033163724987198,
+      "learning_rate": 2.637155160496324e-07,
+      "loss": 0.3044,
+      "step": 10459
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7629600639173423,
+      "learning_rate": 2.636245064360701e-07,
+      "loss": 0.3183,
+      "step": 10460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30311969857539905,
+      "learning_rate": 2.635335069067617e-07,
+      "loss": 0.2666,
+      "step": 10461
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8735044339443045,
+      "learning_rate": 2.634425174655891e-07,
+      "loss": 0.1904,
+      "step": 10462
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4713183463796078,
+      "learning_rate": 2.633515381164342e-07,
+      "loss": 0.0081,
+      "step": 10463
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2370586944612736,
+      "learning_rate": 2.632605688631787e-07,
+      "loss": 0.2827,
+      "step": 10464
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6483202060911514,
+      "learning_rate": 2.631696097097029e-07,
+      "loss": 0.1949,
+      "step": 10465
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.606500670229192,
+      "learning_rate": 2.630786606598879e-07,
+      "loss": 0.2757,
+      "step": 10466
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8034149397610436,
+      "learning_rate": 2.629877217176134e-07,
+      "loss": 0.3208,
+      "step": 10467
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0127645399924585,
+      "learning_rate": 2.6289679288675923e-07,
+      "loss": 0.0089,
+      "step": 10468
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1079613960310797,
+      "learning_rate": 2.628058741712046e-07,
+      "loss": 0.0142,
+      "step": 10469
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.124947011569078,
+      "learning_rate": 2.627149655748279e-07,
+      "loss": 0.1319,
+      "step": 10470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8200726489442691,
+      "learning_rate": 2.626240671015081e-07,
+      "loss": 0.0614,
+      "step": 10471
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.110121682498992,
+      "learning_rate": 2.625331787551225e-07,
+      "loss": 0.2593,
+      "step": 10472
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5619631755302068,
+      "learning_rate": 2.62442300539549e-07,
+      "loss": 0.0547,
+      "step": 10473
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5592244554126755,
+      "learning_rate": 2.6235143245866445e-07,
+      "loss": 0.176,
+      "step": 10474
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.31486389004552046,
+      "learning_rate": 2.622605745163456e-07,
+      "loss": 0.2191,
+      "step": 10475
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6316466594089674,
+      "learning_rate": 2.621697267164684e-07,
+      "loss": 0.1195,
+      "step": 10476
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5717075827821232,
+      "learning_rate": 2.62078889062909e-07,
+      "loss": 0.4002,
+      "step": 10477
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8042376967286633,
+      "learning_rate": 2.619880615595422e-07,
+      "loss": 0.2113,
+      "step": 10478
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4112285756893933,
+      "learning_rate": 2.6189724421024315e-07,
+      "loss": 0.158,
+      "step": 10479
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8920847339347455,
+      "learning_rate": 2.618064370188864e-07,
+      "loss": 0.3246,
+      "step": 10480
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4096801217003756,
+      "learning_rate": 2.61715639989346e-07,
+      "loss": 0.0097,
+      "step": 10481
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.45938981760695313,
+      "learning_rate": 2.616248531254953e-07,
+      "loss": 0.2131,
+      "step": 10482
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7588298537674147,
+      "learning_rate": 2.615340764312077e-07,
+      "loss": 0.2523,
+      "step": 10483
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1579948270642288,
+      "learning_rate": 2.614433099103558e-07,
+      "loss": 0.0244,
+      "step": 10484
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4989591486028849,
+      "learning_rate": 2.613525535668116e-07,
+      "loss": 0.0517,
+      "step": 10485
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6484474008402006,
+      "learning_rate": 2.612618074044475e-07,
+      "loss": 0.1227,
+      "step": 10486
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5772858373783285,
+      "learning_rate": 2.611710714271343e-07,
+      "loss": 0.0385,
+      "step": 10487
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9127399629002068,
+      "learning_rate": 2.610803456387436e-07,
+      "loss": 0.2728,
+      "step": 10488
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1753583013101303,
+      "learning_rate": 2.6098963004314536e-07,
+      "loss": 0.0091,
+      "step": 10489
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6615157221653731,
+      "learning_rate": 2.6089892464421025e-07,
+      "loss": 0.1522,
+      "step": 10490
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3786927597294344,
+      "learning_rate": 2.6080822944580736e-07,
+      "loss": 0.0326,
+      "step": 10491
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8774515974536057,
+      "learning_rate": 2.607175444518066e-07,
+      "loss": 0.1332,
+      "step": 10492
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7859571173425213,
+      "learning_rate": 2.606268696660761e-07,
+      "loss": 0.0165,
+      "step": 10493
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4774352893560765,
+      "learning_rate": 2.6053620509248475e-07,
+      "loss": 0.2128,
+      "step": 10494
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8849807851339067,
+      "learning_rate": 2.604455507349001e-07,
+      "loss": 0.2165,
+      "step": 10495
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3053876242631579,
+      "learning_rate": 2.603549065971898e-07,
+      "loss": 0.1507,
+      "step": 10496
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.236602850657951,
+      "learning_rate": 2.6026427268322113e-07,
+      "loss": 0.2823,
+      "step": 10497
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9731953198793252,
+      "learning_rate": 2.6017364899686035e-07,
+      "loss": 0.353,
+      "step": 10498
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.849028799996507,
+      "learning_rate": 2.6008303554197396e-07,
+      "loss": 0.2629,
+      "step": 10499
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.609483280714231,
+      "learning_rate": 2.5999243232242745e-07,
+      "loss": 0.019,
+      "step": 10500
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3888003858011602,
+      "learning_rate": 2.599018393420864e-07,
+      "loss": 0.1603,
+      "step": 10501
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2085305407941447,
+      "learning_rate": 2.5981125660481535e-07,
+      "loss": 0.1441,
+      "step": 10502
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.44957664609367,
+      "learning_rate": 2.5972068411447914e-07,
+      "loss": 0.0748,
+      "step": 10503
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6362945088703087,
+      "learning_rate": 2.596301218749414e-07,
+      "loss": 0.0895,
+      "step": 10504
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7594616526315587,
+      "learning_rate": 2.5953956989006596e-07,
+      "loss": 0.1739,
+      "step": 10505
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0883217998113466,
+      "learning_rate": 2.594490281637157e-07,
+      "loss": 0.0775,
+      "step": 10506
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.490342342348187,
+      "learning_rate": 2.5935849669975373e-07,
+      "loss": 0.2393,
+      "step": 10507
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0084608531775603,
+      "learning_rate": 2.5926797550204175e-07,
+      "loss": 0.2081,
+      "step": 10508
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.577018754148711,
+      "learning_rate": 2.591774645744421e-07,
+      "loss": 0.2774,
+      "step": 10509
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6290297424700538,
+      "learning_rate": 2.5908696392081573e-07,
+      "loss": 0.2795,
+      "step": 10510
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6718054667773171,
+      "learning_rate": 2.5899647354502403e-07,
+      "loss": 0.1014,
+      "step": 10511
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.45768586251432564,
+      "learning_rate": 2.58905993450927e-07,
+      "loss": 0.1617,
+      "step": 10512
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.646220834732189,
+      "learning_rate": 2.588155236423851e-07,
+      "loss": 0.2365,
+      "step": 10513
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7789869069369135,
+      "learning_rate": 2.5872506412325767e-07,
+      "loss": 0.3558,
+      "step": 10514
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.151739340656819,
+      "learning_rate": 2.5863461489740403e-07,
+      "loss": 0.3409,
+      "step": 10515
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5370364545434133,
+      "learning_rate": 2.585441759686831e-07,
+      "loss": 0.2612,
+      "step": 10516
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5427067013972633,
+      "learning_rate": 2.584537473409528e-07,
+      "loss": 0.1106,
+      "step": 10517
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.617124445554166,
+      "learning_rate": 2.5836332901807145e-07,
+      "loss": 0.0732,
+      "step": 10518
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7087900502440856,
+      "learning_rate": 2.58272921003896e-07,
+      "loss": 0.1356,
+      "step": 10519
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.951110245258846,
+      "learning_rate": 2.5818252330228387e-07,
+      "loss": 0.0826,
+      "step": 10520
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1491656278638571,
+      "learning_rate": 2.580921359170912e-07,
+      "loss": 0.076,
+      "step": 10521
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1568027759612993,
+      "learning_rate": 2.580017588521746e-07,
+      "loss": 0.2779,
+      "step": 10522
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8358940530762977,
+      "learning_rate": 2.579113921113891e-07,
+      "loss": 0.5154,
+      "step": 10523
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2170483844360773,
+      "learning_rate": 2.5782103569859054e-07,
+      "loss": 0.0438,
+      "step": 10524
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7479360058026202,
+      "learning_rate": 2.5773068961763343e-07,
+      "loss": 0.2036,
+      "step": 10525
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.617781493911869,
+      "learning_rate": 2.576403538723719e-07,
+      "loss": 0.1514,
+      "step": 10526
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7265854935703562,
+      "learning_rate": 2.5755002846666033e-07,
+      "loss": 0.2367,
+      "step": 10527
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.737124315075192,
+      "learning_rate": 2.5745971340435177e-07,
+      "loss": 0.293,
+      "step": 10528
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6024153526531448,
+      "learning_rate": 2.5736940868929953e-07,
+      "loss": 0.333,
+      "step": 10529
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.599186344179326,
+      "learning_rate": 2.572791143253559e-07,
+      "loss": 0.1295,
+      "step": 10530
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7085186447716115,
+      "learning_rate": 2.5718883031637314e-07,
+      "loss": 0.2426,
+      "step": 10531
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8266295058925617,
+      "learning_rate": 2.5709855666620316e-07,
+      "loss": 0.1897,
+      "step": 10532
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7798735228341888,
+      "learning_rate": 2.5700829337869694e-07,
+      "loss": 0.215,
+      "step": 10533
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.002867064307344,
+      "learning_rate": 2.5691804045770535e-07,
+      "loss": 0.038,
+      "step": 10534
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8655546053823612,
+      "learning_rate": 2.56827797907079e-07,
+      "loss": 0.11,
+      "step": 10535
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.872437944181974,
+      "learning_rate": 2.5673756573066747e-07,
+      "loss": 0.1736,
+      "step": 10536
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3986849148475318,
+      "learning_rate": 2.5664734393232054e-07,
+      "loss": 0.2444,
+      "step": 10537
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5756746969524624,
+      "learning_rate": 2.5655713251588715e-07,
+      "loss": 0.1042,
+      "step": 10538
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2697659380365542,
+      "learning_rate": 2.564669314852157e-07,
+      "loss": 0.0931,
+      "step": 10539
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3407578119448709,
+      "learning_rate": 2.5637674084415473e-07,
+      "loss": 0.2629,
+      "step": 10540
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.39097716840919244,
+      "learning_rate": 2.5628656059655153e-07,
+      "loss": 0.2376,
+      "step": 10541
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6718139343989389,
+      "learning_rate": 2.561963907462537e-07,
+      "loss": 0.2885,
+      "step": 10542
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8800900704244332,
+      "learning_rate": 2.5610623129710784e-07,
+      "loss": 0.1508,
+      "step": 10543
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0095528779448446,
+      "learning_rate": 2.560160822529606e-07,
+      "loss": 0.499,
+      "step": 10544
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7794417974665724,
+      "learning_rate": 2.5592594361765753e-07,
+      "loss": 0.1611,
+      "step": 10545
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18154471465380143,
+      "learning_rate": 2.5583581539504463e-07,
+      "loss": 0.1021,
+      "step": 10546
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7474633808438751,
+      "learning_rate": 2.557456975889664e-07,
+      "loss": 0.0125,
+      "step": 10547
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.031072803768003,
+      "learning_rate": 2.55655590203268e-07,
+      "loss": 0.0846,
+      "step": 10548
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7108431360941682,
+      "learning_rate": 2.55565493241793e-07,
+      "loss": 0.3651,
+      "step": 10549
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6395022418878527,
+      "learning_rate": 2.5547540670838553e-07,
+      "loss": 0.3176,
+      "step": 10550
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2489121349943053,
+      "learning_rate": 2.553853306068888e-07,
+      "loss": 0.3709,
+      "step": 10551
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.126379091094442,
+      "learning_rate": 2.552952649411457e-07,
+      "loss": 0.0413,
+      "step": 10552
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8528101579995798,
+      "learning_rate": 2.5520520971499857e-07,
+      "loss": 0.1108,
+      "step": 10553
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5092444165228933,
+      "learning_rate": 2.551151649322891e-07,
+      "loss": 0.0573,
+      "step": 10554
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.123056991807261,
+      "learning_rate": 2.5502513059685916e-07,
+      "loss": 0.0976,
+      "step": 10555
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.21528667818707273,
+      "learning_rate": 2.549351067125494e-07,
+      "loss": 0.004,
+      "step": 10556
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1022126568579682,
+      "learning_rate": 2.5484509328320083e-07,
+      "loss": 0.2352,
+      "step": 10557
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.963214973942761,
+      "learning_rate": 2.5475509031265316e-07,
+      "loss": 0.1623,
+      "step": 10558
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.33563736633622543,
+      "learning_rate": 2.5466509780474667e-07,
+      "loss": 0.0764,
+      "step": 10559
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9615137551229664,
+      "learning_rate": 2.5457511576332e-07,
+      "loss": 0.2759,
+      "step": 10560
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8480390854122233,
+      "learning_rate": 2.5448514419221243e-07,
+      "loss": 0.2355,
+      "step": 10561
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.800579296234656,
+      "learning_rate": 2.5439518309526203e-07,
+      "loss": 0.1037,
+      "step": 10562
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8229825352930443,
+      "learning_rate": 2.5430523247630703e-07,
+      "loss": 0.1112,
+      "step": 10563
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4302533019832532,
+      "learning_rate": 2.542152923391845e-07,
+      "loss": 0.0144,
+      "step": 10564
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.36507556899829297,
+      "learning_rate": 2.5412536268773186e-07,
+      "loss": 0.1387,
+      "step": 10565
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.479189561026578,
+      "learning_rate": 2.5403544352578544e-07,
+      "loss": 0.3136,
+      "step": 10566
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4284166623846381,
+      "learning_rate": 2.539455348571813e-07,
+      "loss": 0.022,
+      "step": 10567
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.206928190817757,
+      "learning_rate": 2.538556366857556e-07,
+      "loss": 0.2967,
+      "step": 10568
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.25602016189714777,
+      "learning_rate": 2.5376574901534296e-07,
+      "loss": 0.0336,
+      "step": 10569
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.667943144170874,
+      "learning_rate": 2.536758718497787e-07,
+      "loss": 0.0874,
+      "step": 10570
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1676065572146814,
+      "learning_rate": 2.535860051928967e-07,
+      "loss": 0.1725,
+      "step": 10571
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.06383168117739876,
+      "learning_rate": 2.534961490485313e-07,
+      "loss": 0.0017,
+      "step": 10572
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.612009773288431,
+      "learning_rate": 2.534063034205155e-07,
+      "loss": 0.3274,
+      "step": 10573
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5469969393598104,
+      "learning_rate": 2.5331646831268274e-07,
+      "loss": 0.156,
+      "step": 10574
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.353782065378933,
+      "learning_rate": 2.53226643728865e-07,
+      "loss": 0.3062,
+      "step": 10575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6720243725775021,
+      "learning_rate": 2.531368296728951e-07,
+      "loss": 0.1248,
+      "step": 10576
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4131062860017984,
+      "learning_rate": 2.53047026148604e-07,
+      "loss": 0.2135,
+      "step": 10577
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6516688749793922,
+      "learning_rate": 2.5295723315982345e-07,
+      "loss": 0.1384,
+      "step": 10578
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.46519008824416797,
+      "learning_rate": 2.5286745071038396e-07,
+      "loss": 0.1715,
+      "step": 10579
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4264081869372595,
+      "learning_rate": 2.527776788041156e-07,
+      "loss": 0.1294,
+      "step": 10580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1450129377251033,
+      "learning_rate": 2.526879174448486e-07,
+      "loss": 0.4197,
+      "step": 10581
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9483996792354955,
+      "learning_rate": 2.5259816663641205e-07,
+      "loss": 0.1096,
+      "step": 10582
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5704949117557274,
+      "learning_rate": 2.5250842638263526e-07,
+      "loss": 0.0159,
+      "step": 10583
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0299182900440016,
+      "learning_rate": 2.524186966873463e-07,
+      "loss": 0.3862,
+      "step": 10584
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.038388330939207,
+      "learning_rate": 2.5232897755437346e-07,
+      "loss": 0.3125,
+      "step": 10585
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8743035505804668,
+      "learning_rate": 2.522392689875442e-07,
+      "loss": 0.1385,
+      "step": 10586
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1298631708065149,
+      "learning_rate": 2.521495709906861e-07,
+      "loss": 0.0042,
+      "step": 10587
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5824709882596042,
+      "learning_rate": 2.5205988356762536e-07,
+      "loss": 0.1166,
+      "step": 10588
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.152329124092868,
+      "learning_rate": 2.519702067221886e-07,
+      "loss": 0.1667,
+      "step": 10589
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8882923684298127,
+      "learning_rate": 2.5188054045820115e-07,
+      "loss": 0.1068,
+      "step": 10590
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.43831150842493927,
+      "learning_rate": 2.517908847794889e-07,
+      "loss": 0.4529,
+      "step": 10591
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1712969157414352,
+      "learning_rate": 2.517012396898762e-07,
+      "loss": 0.006,
+      "step": 10592
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5727387135701465,
+      "learning_rate": 2.51611605193188e-07,
+      "loss": 0.0746,
+      "step": 10593
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9156049890281388,
+      "learning_rate": 2.515219812932481e-07,
+      "loss": 0.3767,
+      "step": 10594
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.43717320243928,
+      "learning_rate": 2.5143236799387975e-07,
+      "loss": 0.1063,
+      "step": 10595
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3269259054658187,
+      "learning_rate": 2.5134276529890644e-07,
+      "loss": 0.0593,
+      "step": 10596
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3760722979262,
+      "learning_rate": 2.5125317321215046e-07,
+      "loss": 0.2649,
+      "step": 10597
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9689287146562792,
+      "learning_rate": 2.511635917374343e-07,
+      "loss": 0.1306,
+      "step": 10598
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7995139365785888,
+      "learning_rate": 2.510740208785793e-07,
+      "loss": 0.1817,
+      "step": 10599
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1968501978495556,
+      "learning_rate": 2.5098446063940725e-07,
+      "loss": 0.1242,
+      "step": 10600
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.47176432883077934,
+      "learning_rate": 2.508949110237385e-07,
+      "loss": 0.0039,
+      "step": 10601
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8108725101723075,
+      "learning_rate": 2.5080537203539364e-07,
+      "loss": 0.2779,
+      "step": 10602
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8583643549900759,
+      "learning_rate": 2.5071584367819267e-07,
+      "loss": 0.4885,
+      "step": 10603
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26309118018076055,
+      "learning_rate": 2.506263259559548e-07,
+      "loss": 0.0486,
+      "step": 10604
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8450317211287066,
+      "learning_rate": 2.505368188724991e-07,
+      "loss": 0.2276,
+      "step": 10605
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9268324661988067,
+      "learning_rate": 2.5044732243164444e-07,
+      "loss": 0.0802,
+      "step": 10606
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5004466655700627,
+      "learning_rate": 2.5035783663720866e-07,
+      "loss": 0.2071,
+      "step": 10607
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.34058912591187507,
+      "learning_rate": 2.5026836149300924e-07,
+      "loss": 0.198,
+      "step": 10608
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3687588008158,
+      "learning_rate": 2.5017889700286366e-07,
+      "loss": 0.0463,
+      "step": 10609
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3371554072172445,
+      "learning_rate": 2.5008944317058844e-07,
+      "loss": 0.1614,
+      "step": 10610
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7631054936767125,
+      "learning_rate": 2.500000000000001e-07,
+      "loss": 0.2467,
+      "step": 10611
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.570913992693844,
+      "learning_rate": 2.4991056749491414e-07,
+      "loss": 0.0508,
+      "step": 10612
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5580753188516984,
+      "learning_rate": 2.498211456591463e-07,
+      "loss": 0.1841,
+      "step": 10613
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.311823748840983,
+      "learning_rate": 2.4973173449651106e-07,
+      "loss": 0.2012,
+      "step": 10614
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3992919977999924,
+      "learning_rate": 2.496423340108234e-07,
+      "loss": 0.0016,
+      "step": 10615
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7304015725613393,
+      "learning_rate": 2.4955294420589687e-07,
+      "loss": 0.2443,
+      "step": 10616
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7688464051008319,
+      "learning_rate": 2.494635650855453e-07,
+      "loss": 0.1986,
+      "step": 10617
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3838141955276136,
+      "learning_rate": 2.493741966535816e-07,
+      "loss": 0.0887,
+      "step": 10618
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1487147416039272,
+      "learning_rate": 2.4928483891381863e-07,
+      "loss": 0.1145,
+      "step": 10619
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5174528699626204,
+      "learning_rate": 2.4919549187006826e-07,
+      "loss": 0.2569,
+      "step": 10620
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8653498544965659,
+      "learning_rate": 2.491061555261424e-07,
+      "loss": 0.2112,
+      "step": 10621
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8580172433265398,
+      "learning_rate": 2.4901682988585244e-07,
+      "loss": 0.1548,
+      "step": 10622
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7678532818603045,
+      "learning_rate": 2.489275149530089e-07,
+      "loss": 0.2632,
+      "step": 10623
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.852388175325931,
+      "learning_rate": 2.488382107314225e-07,
+      "loss": 0.2685,
+      "step": 10624
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4027321008603786,
+      "learning_rate": 2.487489172249027e-07,
+      "loss": 0.2474,
+      "step": 10625
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.049118659300945,
+      "learning_rate": 2.486596344372594e-07,
+      "loss": 0.1608,
+      "step": 10626
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8708467854010007,
+      "learning_rate": 2.485703623723011e-07,
+      "loss": 0.214,
+      "step": 10627
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4338989155347018,
+      "learning_rate": 2.4848110103383683e-07,
+      "loss": 0.1835,
+      "step": 10628
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0908190163622082,
+      "learning_rate": 2.4839185042567424e-07,
+      "loss": 0.2569,
+      "step": 10629
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6196024690562174,
+      "learning_rate": 2.483026105516212e-07,
+      "loss": 0.111,
+      "step": 10630
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7471693799119881,
+      "learning_rate": 2.4821338141548465e-07,
+      "loss": 0.1557,
+      "step": 10631
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.471578282580022,
+      "learning_rate": 2.481241630210716e-07,
+      "loss": 0.2435,
+      "step": 10632
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.48844098333069136,
+      "learning_rate": 2.4803495537218787e-07,
+      "loss": 0.0122,
+      "step": 10633
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9526328234776237,
+      "learning_rate": 2.4794575847263963e-07,
+      "loss": 0.2759,
+      "step": 10634
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0131510610706103,
+      "learning_rate": 2.478565723262321e-07,
+      "loss": 0.4307,
+      "step": 10635
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9358636516171898,
+      "learning_rate": 2.4776739693676977e-07,
+      "loss": 0.2347,
+      "step": 10636
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.165739279034136,
+      "learning_rate": 2.476782323080574e-07,
+      "loss": 0.2701,
+      "step": 10637
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5396514702615218,
+      "learning_rate": 2.4758907844389905e-07,
+      "loss": 0.1629,
+      "step": 10638
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.320681171798618,
+      "learning_rate": 2.474999353480978e-07,
+      "loss": 0.3476,
+      "step": 10639
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0111865266702131,
+      "learning_rate": 2.47410803024457e-07,
+      "loss": 0.2517,
+      "step": 10640
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0803292059452372,
+      "learning_rate": 2.473216814767792e-07,
+      "loss": 0.1364,
+      "step": 10641
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.501200864153896,
+      "learning_rate": 2.472325707088663e-07,
+      "loss": 0.1743,
+      "step": 10642
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.47594962004949454,
+      "learning_rate": 2.471434707245202e-07,
+      "loss": 0.2192,
+      "step": 10643
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.720951558466603,
+      "learning_rate": 2.4705438152754174e-07,
+      "loss": 0.2886,
+      "step": 10644
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.49906162606421556,
+      "learning_rate": 2.469653031217321e-07,
+      "loss": 0.0609,
+      "step": 10645
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5942776477775826,
+      "learning_rate": 2.4687623551089104e-07,
+      "loss": 0.1346,
+      "step": 10646
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1049866050200776,
+      "learning_rate": 2.4678717869881884e-07,
+      "loss": 0.214,
+      "step": 10647
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0522141300105585,
+      "learning_rate": 2.466981326893146e-07,
+      "loss": 0.0121,
+      "step": 10648
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.595667859330842,
+      "learning_rate": 2.46609097486177e-07,
+      "loss": 0.1439,
+      "step": 10649
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6189656097202366,
+      "learning_rate": 2.465200730932049e-07,
+      "loss": 0.4181,
+      "step": 10650
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4350202001291532,
+      "learning_rate": 2.464310595141959e-07,
+      "loss": 0.3342,
+      "step": 10651
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9072968235831722,
+      "learning_rate": 2.4634205675294777e-07,
+      "loss": 0.5667,
+      "step": 10652
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.433996284579106,
+      "learning_rate": 2.4625306481325727e-07,
+      "loss": 0.0617,
+      "step": 10653
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8957853909505935,
+      "learning_rate": 2.4616408369892126e-07,
+      "loss": 0.2836,
+      "step": 10654
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8956053562597286,
+      "learning_rate": 2.4607511341373557e-07,
+      "loss": 0.1735,
+      "step": 10655
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9634464743769201,
+      "learning_rate": 2.4598615396149597e-07,
+      "loss": 0.4158,
+      "step": 10656
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2599236002697625,
+      "learning_rate": 2.458972053459977e-07,
+      "loss": 0.1456,
+      "step": 10657
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4171246802869317,
+      "learning_rate": 2.4580826757103564e-07,
+      "loss": 0.162,
+      "step": 10658
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.632731797597872,
+      "learning_rate": 2.457193406404036e-07,
+      "loss": 0.1327,
+      "step": 10659
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.463570315065967,
+      "learning_rate": 2.4563042455789593e-07,
+      "loss": 0.2327,
+      "step": 10660
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7475601521375592,
+      "learning_rate": 2.455415193273055e-07,
+      "loss": 0.3889,
+      "step": 10661
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9511196635599971,
+      "learning_rate": 2.4545262495242556e-07,
+      "loss": 0.2351,
+      "step": 10662
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.354419344303781,
+      "learning_rate": 2.4536374143704834e-07,
+      "loss": 0.1062,
+      "step": 10663
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7422148530732846,
+      "learning_rate": 2.452748687849656e-07,
+      "loss": 0.239,
+      "step": 10664
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9694439957071486,
+      "learning_rate": 2.4518600699996925e-07,
+      "loss": 0.2528,
+      "step": 10665
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.38941484134385324,
+      "learning_rate": 2.4509715608584986e-07,
+      "loss": 0.0716,
+      "step": 10666
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53835049220775,
+      "learning_rate": 2.4500831604639843e-07,
+      "loss": 0.2671,
+      "step": 10667
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4995410768084054,
+      "learning_rate": 2.449194868854046e-07,
+      "loss": 0.0492,
+      "step": 10668
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.688961576178236,
+      "learning_rate": 2.448306686066585e-07,
+      "loss": 0.3869,
+      "step": 10669
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7985312278403478,
+      "learning_rate": 2.4474186121394875e-07,
+      "loss": 0.3686,
+      "step": 10670
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.116955759566959,
+      "learning_rate": 2.446530647110646e-07,
+      "loss": 0.3511,
+      "step": 10671
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3933397101762621,
+      "learning_rate": 2.4456427910179374e-07,
+      "loss": 0.1386,
+      "step": 10672
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.24738398884795554,
+      "learning_rate": 2.444755043899243e-07,
+      "loss": 0.1014,
+      "step": 10673
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5006446506991662,
+      "learning_rate": 2.4438674057924365e-07,
+      "loss": 0.2221,
+      "step": 10674
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.689609795630209,
+      "learning_rate": 2.442979876735383e-07,
+      "loss": 0.2426,
+      "step": 10675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3975373833477966,
+      "learning_rate": 2.4420924567659504e-07,
+      "loss": 0.2743,
+      "step": 10676
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.592271887308199,
+      "learning_rate": 2.441205145921994e-07,
+      "loss": 0.1441,
+      "step": 10677
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.36363260404378417,
+      "learning_rate": 2.440317944241372e-07,
+      "loss": 0.0879,
+      "step": 10678
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8731244608591947,
+      "learning_rate": 2.4394308517619293e-07,
+      "loss": 0.2039,
+      "step": 10679
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1850744507161584,
+      "learning_rate": 2.4385438685215165e-07,
+      "loss": 0.0939,
+      "step": 10680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.922497198506668,
+      "learning_rate": 2.437656994557969e-07,
+      "loss": 0.2345,
+      "step": 10681
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9969140916400155,
+      "learning_rate": 2.4367702299091274e-07,
+      "loss": 0.5095,
+      "step": 10682
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6505041781957592,
+      "learning_rate": 2.4358835746128186e-07,
+      "loss": 0.2,
+      "step": 10683
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5692669734993312,
+      "learning_rate": 2.4349970287068725e-07,
+      "loss": 0.1239,
+      "step": 10684
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4186398298078355,
+      "learning_rate": 2.4341105922291076e-07,
+      "loss": 0.1111,
+      "step": 10685
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5287317822196954,
+      "learning_rate": 2.4332242652173455e-07,
+      "loss": 0.1459,
+      "step": 10686
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2826770114363877,
+      "learning_rate": 2.4323380477093934e-07,
+      "loss": 0.5146,
+      "step": 10687
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4946248138867597,
+      "learning_rate": 2.4314519397430646e-07,
+      "loss": 0.0826,
+      "step": 10688
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4667660044199442,
+      "learning_rate": 2.430565941356157e-07,
+      "loss": 0.3147,
+      "step": 10689
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3180618724213369,
+      "learning_rate": 2.4296800525864736e-07,
+      "loss": 0.2997,
+      "step": 10690
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.671803698899248,
+      "learning_rate": 2.4287942734718044e-07,
+      "loss": 0.2124,
+      "step": 10691
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.36901619813841613,
+      "learning_rate": 2.4279086040499394e-07,
+      "loss": 0.1636,
+      "step": 10692
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.125111527344528,
+      "learning_rate": 2.427023044358667e-07,
+      "loss": 0.195,
+      "step": 10693
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5461668614907347,
+      "learning_rate": 2.4261375944357607e-07,
+      "loss": 0.3006,
+      "step": 10694
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.36798634203906,
+      "learning_rate": 2.4252522543190017e-07,
+      "loss": 0.1184,
+      "step": 10695
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9025822251560608,
+      "learning_rate": 2.4243670240461556e-07,
+      "loss": 0.2122,
+      "step": 10696
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2385314046509113,
+      "learning_rate": 2.4234819036549916e-07,
+      "loss": 0.3901,
+      "step": 10697
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.509657875785536,
+      "learning_rate": 2.422596893183266e-07,
+      "loss": 0.2317,
+      "step": 10698
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2800454280133113,
+      "learning_rate": 2.421711992668741e-07,
+      "loss": 0.3699,
+      "step": 10699
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9437888866957841,
+      "learning_rate": 2.420827202149164e-07,
+      "loss": 0.1995,
+      "step": 10700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.7303262059632845,
+      "learning_rate": 2.419942521662285e-07,
+      "loss": 0.1273,
+      "step": 10701
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.096351887508121,
+      "learning_rate": 2.419057951245842e-07,
+      "loss": 0.1125,
+      "step": 10702
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.0985929843028215,
+      "learning_rate": 2.418173490937578e-07,
+      "loss": 0.0228,
+      "step": 10703
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3791057710799882,
+      "learning_rate": 2.417289140775222e-07,
+      "loss": 0.13,
+      "step": 10704
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.587168419506907,
+      "learning_rate": 2.4164049007965023e-07,
+      "loss": 0.2207,
+      "step": 10705
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6911265857887287,
+      "learning_rate": 2.415520771039144e-07,
+      "loss": 0.4728,
+      "step": 10706
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4532718760205119,
+      "learning_rate": 2.414636751540864e-07,
+      "loss": 0.1215,
+      "step": 10707
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0060944484966456,
+      "learning_rate": 2.413752842339377e-07,
+      "loss": 0.2105,
+      "step": 10708
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9268099357017542,
+      "learning_rate": 2.4128690434723947e-07,
+      "loss": 0.2565,
+      "step": 10709
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2520195283701715,
+      "learning_rate": 2.411985354977618e-07,
+      "loss": 0.2335,
+      "step": 10710
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0977907578748356,
+      "learning_rate": 2.411101776892749e-07,
+      "loss": 0.0649,
+      "step": 10711
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.7218064755685125,
+      "learning_rate": 2.410218309255484e-07,
+      "loss": 0.0321,
+      "step": 10712
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7625233089280851,
+      "learning_rate": 2.40933495210351e-07,
+      "loss": 0.1131,
+      "step": 10713
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0855095888509445,
+      "learning_rate": 2.408451705474517e-07,
+      "loss": 0.114,
+      "step": 10714
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1842261955563496,
+      "learning_rate": 2.407568569406182e-07,
+      "loss": 0.0506,
+      "step": 10715
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.38620217678899577,
+      "learning_rate": 2.406685543936185e-07,
+      "loss": 0.0084,
+      "step": 10716
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5924790759334992,
+      "learning_rate": 2.405802629102196e-07,
+      "loss": 0.0855,
+      "step": 10717
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5337281680297228,
+      "learning_rate": 2.4049198249418803e-07,
+      "loss": 0.2453,
+      "step": 10718
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5770910264205996,
+      "learning_rate": 2.404037131492903e-07,
+      "loss": 0.1786,
+      "step": 10719
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8558103821620981,
+      "learning_rate": 2.403154548792918e-07,
+      "loss": 0.172,
+      "step": 10720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8557960799844515,
+      "learning_rate": 2.402272076879583e-07,
+      "loss": 0.1622,
+      "step": 10721
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.80220731283959,
+      "learning_rate": 2.401389715790541e-07,
+      "loss": 0.2381,
+      "step": 10722
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.076657403162898,
+      "learning_rate": 2.4005074655634393e-07,
+      "loss": 0.3262,
+      "step": 10723
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3533423049794706,
+      "learning_rate": 2.3996253262359133e-07,
+      "loss": 0.2099,
+      "step": 10724
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5668099182172478,
+      "learning_rate": 2.3987432978455995e-07,
+      "loss": 0.1811,
+      "step": 10725
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5371098300354171,
+      "learning_rate": 2.3978613804301246e-07,
+      "loss": 0.4094,
+      "step": 10726
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.838505960343988,
+      "learning_rate": 2.3969795740271145e-07,
+      "loss": 0.2311,
+      "step": 10727
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.667570296835776,
+      "learning_rate": 2.3960978786741874e-07,
+      "loss": 0.1847,
+      "step": 10728
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5828788587723586,
+      "learning_rate": 2.3952162944089616e-07,
+      "loss": 0.1691,
+      "step": 10729
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9476949999625318,
+      "learning_rate": 2.3943348212690433e-07,
+      "loss": 0.3167,
+      "step": 10730
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8783290069773235,
+      "learning_rate": 2.3934534592920413e-07,
+      "loss": 0.1312,
+      "step": 10731
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.081775542810553,
+      "learning_rate": 2.392572208515555e-07,
+      "loss": 0.397,
+      "step": 10732
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5762764388574344,
+      "learning_rate": 2.3916910689771773e-07,
+      "loss": 0.0017,
+      "step": 10733
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7355044397297507,
+      "learning_rate": 2.390810040714504e-07,
+      "loss": 0.2677,
+      "step": 10734
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5964938536065715,
+      "learning_rate": 2.389929123765118e-07,
+      "loss": 0.0749,
+      "step": 10735
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7905799500317098,
+      "learning_rate": 2.3890483181666036e-07,
+      "loss": 0.1636,
+      "step": 10736
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.36713854956127034,
+      "learning_rate": 2.3881676239565356e-07,
+      "loss": 0.0206,
+      "step": 10737
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.347921022010893,
+      "learning_rate": 2.3872870411724887e-07,
+      "loss": 0.1419,
+      "step": 10738
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6525773147906483,
+      "learning_rate": 2.386406569852027e-07,
+      "loss": 0.4326,
+      "step": 10739
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5300134958343342,
+      "learning_rate": 2.3855262100327167e-07,
+      "loss": 0.1283,
+      "step": 10740
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8747068130142694,
+      "learning_rate": 2.3846459617521123e-07,
+      "loss": 0.0996,
+      "step": 10741
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8111566195875461,
+      "learning_rate": 2.3837658250477704e-07,
+      "loss": 0.2887,
+      "step": 10742
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6136206640377274,
+      "learning_rate": 2.382885799957236e-07,
+      "loss": 0.2516,
+      "step": 10743
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.35564433647555616,
+      "learning_rate": 2.3820058865180548e-07,
+      "loss": 0.1259,
+      "step": 10744
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.30969938791369,
+      "learning_rate": 2.3811260847677668e-07,
+      "loss": 0.0474,
+      "step": 10745
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.982035371250566,
+      "learning_rate": 2.3802463947439028e-07,
+      "loss": 0.2785,
+      "step": 10746
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5894925748322754,
+      "learning_rate": 2.3793668164839954e-07,
+      "loss": 0.3994,
+      "step": 10747
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8366667867151247,
+      "learning_rate": 2.3784873500255666e-07,
+      "loss": 0.0728,
+      "step": 10748
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4545383376816283,
+      "learning_rate": 2.3776079954061385e-07,
+      "loss": 0.2984,
+      "step": 10749
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0425947107482705,
+      "learning_rate": 2.3767287526632234e-07,
+      "loss": 0.4034,
+      "step": 10750
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0457992927079485,
+      "learning_rate": 2.3758496218343355e-07,
+      "loss": 0.0778,
+      "step": 10751
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5266467923509143,
+      "learning_rate": 2.3749706029569754e-07,
+      "loss": 0.2213,
+      "step": 10752
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.39514589901137304,
+      "learning_rate": 2.3740916960686486e-07,
+      "loss": 0.0478,
+      "step": 10753
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7346820868868575,
+      "learning_rate": 2.373212901206847e-07,
+      "loss": 0.1854,
+      "step": 10754
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.701917774586411,
+      "learning_rate": 2.372334218409065e-07,
+      "loss": 0.0562,
+      "step": 10755
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.841902311073271,
+      "learning_rate": 2.3714556477127861e-07,
+      "loss": 0.3454,
+      "step": 10756
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9812225346183727,
+      "learning_rate": 2.3705771891554948e-07,
+      "loss": 0.2398,
+      "step": 10757
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5400856289738735,
+      "learning_rate": 2.3696988427746667e-07,
+      "loss": 0.0455,
+      "step": 10758
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0962111199743132,
+      "learning_rate": 2.3688206086077716e-07,
+      "loss": 0.2525,
+      "step": 10759
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5658892060409247,
+      "learning_rate": 2.3679424866922805e-07,
+      "loss": 0.2609,
+      "step": 10760
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8507019580896611,
+      "learning_rate": 2.3670644770656517e-07,
+      "loss": 0.417,
+      "step": 10761
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.051728849599386,
+      "learning_rate": 2.3661865797653458e-07,
+      "loss": 0.2295,
+      "step": 10762
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3231275283644623,
+      "learning_rate": 2.3653087948288147e-07,
+      "loss": 0.1155,
+      "step": 10763
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.17446752350544095,
+      "learning_rate": 2.3644311222935088e-07,
+      "loss": 0.0846,
+      "step": 10764
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2839051739428037,
+      "learning_rate": 2.3635535621968678e-07,
+      "loss": 0.1953,
+      "step": 10765
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.34255789800130965,
+      "learning_rate": 2.362676114576333e-07,
+      "loss": 0.1236,
+      "step": 10766
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9428923488254283,
+      "learning_rate": 2.3617987794693357e-07,
+      "loss": 0.0381,
+      "step": 10767
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0815575565411712,
+      "learning_rate": 2.3609215569133074e-07,
+      "loss": 0.2537,
+      "step": 10768
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2725050593944605,
+      "learning_rate": 2.3600444469456688e-07,
+      "loss": 0.0969,
+      "step": 10769
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.320496287192785,
+      "learning_rate": 2.3591674496038428e-07,
+      "loss": 0.0124,
+      "step": 10770
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3960283998846838,
+      "learning_rate": 2.3582905649252404e-07,
+      "loss": 0.3139,
+      "step": 10771
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9080930515887877,
+      "learning_rate": 2.357413792947275e-07,
+      "loss": 0.3573,
+      "step": 10772
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.022924753631725,
+      "learning_rate": 2.3565371337073493e-07,
+      "loss": 0.2909,
+      "step": 10773
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.192332589148641,
+      "learning_rate": 2.355660587242862e-07,
+      "loss": 0.3369,
+      "step": 10774
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5991252494796782,
+      "learning_rate": 2.3547841535912112e-07,
+      "loss": 0.3739,
+      "step": 10775
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3910590544628323,
+      "learning_rate": 2.3539078327897843e-07,
+      "loss": 0.1886,
+      "step": 10776
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5203742713285756,
+      "learning_rate": 2.35303162487597e-07,
+      "loss": 0.1422,
+      "step": 10777
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.009354655764911,
+      "learning_rate": 2.3521555298871466e-07,
+      "loss": 0.0497,
+      "step": 10778
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4855409972449557,
+      "learning_rate": 2.3512795478606905e-07,
+      "loss": 0.3071,
+      "step": 10779
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6035946861246638,
+      "learning_rate": 2.350403678833976e-07,
+      "loss": 0.2212,
+      "step": 10780
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.43697590959260857,
+      "learning_rate": 2.3495279228443643e-07,
+      "loss": 0.0041,
+      "step": 10781
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.761721033718258,
+      "learning_rate": 2.3486522799292202e-07,
+      "loss": 0.3627,
+      "step": 10782
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8386367542748692,
+      "learning_rate": 2.347776750125901e-07,
+      "loss": 0.2445,
+      "step": 10783
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.670499795764103,
+      "learning_rate": 2.346901333471756e-07,
+      "loss": 0.187,
+      "step": 10784
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6148808205684445,
+      "learning_rate": 2.346026030004135e-07,
+      "loss": 0.0973,
+      "step": 10785
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4434147233251777,
+      "learning_rate": 2.3451508397603785e-07,
+      "loss": 0.1887,
+      "step": 10786
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.134159405329597,
+      "learning_rate": 2.3442757627778227e-07,
+      "loss": 0.2923,
+      "step": 10787
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9278861527308889,
+      "learning_rate": 2.343400799093803e-07,
+      "loss": 0.1494,
+      "step": 10788
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6902228714397468,
+      "learning_rate": 2.3425259487456439e-07,
+      "loss": 0.1855,
+      "step": 10789
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7393512624553055,
+      "learning_rate": 2.341651211770672e-07,
+      "loss": 0.0571,
+      "step": 10790
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2855576620004634,
+      "learning_rate": 2.340776588206202e-07,
+      "loss": 0.3084,
+      "step": 10791
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.208128575753836,
+      "learning_rate": 2.3399020780895495e-07,
+      "loss": 0.2589,
+      "step": 10792
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.068699523543247,
+      "learning_rate": 2.33902768145802e-07,
+      "loss": 0.1964,
+      "step": 10793
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.622485600677907,
+      "learning_rate": 2.338153398348921e-07,
+      "loss": 0.1287,
+      "step": 10794
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8692939431777988,
+      "learning_rate": 2.3372792287995473e-07,
+      "loss": 0.2133,
+      "step": 10795
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.51608069027204,
+      "learning_rate": 2.336405172847196e-07,
+      "loss": 0.4074,
+      "step": 10796
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.036266952579445,
+      "learning_rate": 2.3355312305291526e-07,
+      "loss": 0.2128,
+      "step": 10797
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.727967646911707,
+      "learning_rate": 2.3346574018827037e-07,
+      "loss": 0.3428,
+      "step": 10798
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3616547796227862,
+      "learning_rate": 2.3337836869451294e-07,
+      "loss": 0.2615,
+      "step": 10799
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.447780603213033,
+      "learning_rate": 2.332910085753701e-07,
+      "loss": 0.1512,
+      "step": 10800
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9427087413710671,
+      "learning_rate": 2.3320365983456918e-07,
+      "loss": 0.267,
+      "step": 10801
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.508688698731216,
+      "learning_rate": 2.3311632247583623e-07,
+      "loss": 0.3934,
+      "step": 10802
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9400835084229531,
+      "learning_rate": 2.330289965028977e-07,
+      "loss": 0.3936,
+      "step": 10803
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.14746324843519315,
+      "learning_rate": 2.3294168191947867e-07,
+      "loss": 0.0015,
+      "step": 10804
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1419398616777385,
+      "learning_rate": 2.3285437872930452e-07,
+      "loss": 0.1238,
+      "step": 10805
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7147358235020845,
+      "learning_rate": 2.3276708693609942e-07,
+      "loss": 0.3919,
+      "step": 10806
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5466781663745297,
+      "learning_rate": 2.3267980654358782e-07,
+      "loss": 0.1859,
+      "step": 10807
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.135015677342683,
+      "learning_rate": 2.325925375554928e-07,
+      "loss": 0.2184,
+      "step": 10808
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.826877852968951,
+      "learning_rate": 2.3250527997553794e-07,
+      "loss": 0.2951,
+      "step": 10809
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5517360802098473,
+      "learning_rate": 2.324180338074453e-07,
+      "loss": 0.0864,
+      "step": 10810
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1360274930485523,
+      "learning_rate": 2.3233079905493747e-07,
+      "loss": 0.2908,
+      "step": 10811
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7841358997135572,
+      "learning_rate": 2.322435757217357e-07,
+      "loss": 0.0843,
+      "step": 10812
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.419906399602515,
+      "learning_rate": 2.3215636381156135e-07,
+      "loss": 0.0172,
+      "step": 10813
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0568777606934483,
+      "learning_rate": 2.3206916332813481e-07,
+      "loss": 0.2334,
+      "step": 10814
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2047534685914261,
+      "learning_rate": 2.3198197427517658e-07,
+      "loss": 0.0967,
+      "step": 10815
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7356720773304193,
+      "learning_rate": 2.3189479665640588e-07,
+      "loss": 0.1231,
+      "step": 10816
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9034519010183022,
+      "learning_rate": 2.3180763047554213e-07,
+      "loss": 0.1016,
+      "step": 10817
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.186849603714974,
+      "learning_rate": 2.3172047573630416e-07,
+      "loss": 0.1949,
+      "step": 10818
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9214053839934992,
+      "learning_rate": 2.3163333244240984e-07,
+      "loss": 0.1373,
+      "step": 10819
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.678931933551839,
+      "learning_rate": 2.3154620059757723e-07,
+      "loss": 0.0885,
+      "step": 10820
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9878662707130919,
+      "learning_rate": 2.3145908020552318e-07,
+      "loss": 0.3005,
+      "step": 10821
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.47761683792383697,
+      "learning_rate": 2.3137197126996482e-07,
+      "loss": 0.2039,
+      "step": 10822
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3679451424814524,
+      "learning_rate": 2.3128487379461798e-07,
+      "loss": 0.1906,
+      "step": 10823
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0166061374711701,
+      "learning_rate": 2.3119778778319877e-07,
+      "loss": 0.2979,
+      "step": 10824
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.216325736730003,
+      "learning_rate": 2.311107132394221e-07,
+      "loss": 0.29,
+      "step": 10825
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.482344311760232,
+      "learning_rate": 2.3102365016700315e-07,
+      "loss": 0.1998,
+      "step": 10826
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8380530062664374,
+      "learning_rate": 2.3093659856965596e-07,
+      "loss": 0.0783,
+      "step": 10827
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6588364842038741,
+      "learning_rate": 2.3084955845109416e-07,
+      "loss": 0.1647,
+      "step": 10828
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2935703272833226,
+      "learning_rate": 2.3076252981503154e-07,
+      "loss": 0.0233,
+      "step": 10829
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.47247374926229946,
+      "learning_rate": 2.3067551266518037e-07,
+      "loss": 0.1596,
+      "step": 10830
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9566299353721107,
+      "learning_rate": 2.305885070052534e-07,
+      "loss": 0.532,
+      "step": 10831
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8751567444118277,
+      "learning_rate": 2.305015128389622e-07,
+      "loss": 0.1801,
+      "step": 10832
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5742905626408723,
+      "learning_rate": 2.3041453017001815e-07,
+      "loss": 0.0756,
+      "step": 10833
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.38439957698341465,
+      "learning_rate": 2.3032755900213223e-07,
+      "loss": 0.2279,
+      "step": 10834
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18531328291539825,
+      "learning_rate": 2.302405993390149e-07,
+      "loss": 0.1327,
+      "step": 10835
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.60385236376719,
+      "learning_rate": 2.3015365118437574e-07,
+      "loss": 0.2272,
+      "step": 10836
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6486143980048088,
+      "learning_rate": 2.3006671454192444e-07,
+      "loss": 0.2544,
+      "step": 10837
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6628286454833892,
+      "learning_rate": 2.2997978941536955e-07,
+      "loss": 0.4387,
+      "step": 10838
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.450387559009505,
+      "learning_rate": 2.2989287580841981e-07,
+      "loss": 0.1608,
+      "step": 10839
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.560444908888854,
+      "learning_rate": 2.2980597372478282e-07,
+      "loss": 0.1453,
+      "step": 10840
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7652646336107973,
+      "learning_rate": 2.2971908316816635e-07,
+      "loss": 0.4461,
+      "step": 10841
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.641306707553857,
+      "learning_rate": 2.2963220414227708e-07,
+      "loss": 0.2179,
+      "step": 10842
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1646865114753244,
+      "learning_rate": 2.2954533665082132e-07,
+      "loss": 0.1115,
+      "step": 10843
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5558736048088738,
+      "learning_rate": 2.2945848069750538e-07,
+      "loss": 0.0706,
+      "step": 10844
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0829731712122816,
+      "learning_rate": 2.2937163628603434e-07,
+      "loss": 0.1866,
+      "step": 10845
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7158053352819196,
+      "learning_rate": 2.2928480342011347e-07,
+      "loss": 0.4227,
+      "step": 10846
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5870431288921848,
+      "learning_rate": 2.2919798210344694e-07,
+      "loss": 0.1103,
+      "step": 10847
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.518641319599222,
+      "learning_rate": 2.2911117233973905e-07,
+      "loss": 0.1914,
+      "step": 10848
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 23.7094736476411,
+      "learning_rate": 2.2902437413269293e-07,
+      "loss": 0.0784,
+      "step": 10849
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.037859043760364,
+      "learning_rate": 2.2893758748601172e-07,
+      "loss": 0.2207,
+      "step": 10850
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5423624933920133,
+      "learning_rate": 2.2885081240339809e-07,
+      "loss": 0.3081,
+      "step": 10851
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5829197775705188,
+      "learning_rate": 2.2876404888855372e-07,
+      "loss": 0.1074,
+      "step": 10852
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.54594707465134,
+      "learning_rate": 2.286772969451803e-07,
+      "loss": 0.1657,
+      "step": 10853
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7077938283153192,
+      "learning_rate": 2.28590556576979e-07,
+      "loss": 0.2317,
+      "step": 10854
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8975389192393,
+      "learning_rate": 2.2850382778765014e-07,
+      "loss": 0.227,
+      "step": 10855
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.35601280740426633,
+      "learning_rate": 2.2841711058089358e-07,
+      "loss": 0.2147,
+      "step": 10856
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9712915419698354,
+      "learning_rate": 2.2833040496040922e-07,
+      "loss": 0.1006,
+      "step": 10857
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9216084415325392,
+      "learning_rate": 2.2824371092989574e-07,
+      "loss": 0.1229,
+      "step": 10858
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6577588361517682,
+      "learning_rate": 2.2815702849305202e-07,
+      "loss": 0.2169,
+      "step": 10859
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28307925170456116,
+      "learning_rate": 2.2807035765357573e-07,
+      "loss": 0.105,
+      "step": 10860
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.071701063090803,
+      "learning_rate": 2.2798369841516484e-07,
+      "loss": 0.0923,
+      "step": 10861
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.607538185220634,
+      "learning_rate": 2.27897050781516e-07,
+      "loss": 0.4002,
+      "step": 10862
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5820780420207257,
+      "learning_rate": 2.2781041475632617e-07,
+      "loss": 0.1472,
+      "step": 10863
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6536469069476663,
+      "learning_rate": 2.2772379034329103e-07,
+      "loss": 0.1771,
+      "step": 10864
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8923911434698747,
+      "learning_rate": 2.2763717754610652e-07,
+      "loss": 0.2508,
+      "step": 10865
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.736036544110356,
+      "learning_rate": 2.2755057636846737e-07,
+      "loss": 0.0673,
+      "step": 10866
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.13211128099521,
+      "learning_rate": 2.2746398681406847e-07,
+      "loss": 0.2356,
+      "step": 10867
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.47241730232045226,
+      "learning_rate": 2.2737740888660356e-07,
+      "loss": 0.1102,
+      "step": 10868
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4254458946121775,
+      "learning_rate": 2.272908425897665e-07,
+      "loss": 0.2835,
+      "step": 10869
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7895794471402303,
+      "learning_rate": 2.2720428792725044e-07,
+      "loss": 0.2342,
+      "step": 10870
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6911187701869503,
+      "learning_rate": 2.2711774490274766e-07,
+      "loss": 0.0379,
+      "step": 10871
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.949568196928449,
+      "learning_rate": 2.2703121351995064e-07,
+      "loss": 0.3727,
+      "step": 10872
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6162927476970472,
+      "learning_rate": 2.269446937825506e-07,
+      "loss": 0.218,
+      "step": 10873
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7294745669729295,
+      "learning_rate": 2.2685818569423904e-07,
+      "loss": 0.1375,
+      "step": 10874
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7484061186990685,
+      "learning_rate": 2.2677168925870615e-07,
+      "loss": 0.5,
+      "step": 10875
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.86320809631737,
+      "learning_rate": 2.2668520447964246e-07,
+      "loss": 0.1564,
+      "step": 10876
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3315311656569516,
+      "learning_rate": 2.2659873136073715e-07,
+      "loss": 0.0126,
+      "step": 10877
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.46869695401621636,
+      "learning_rate": 2.265122699056798e-07,
+      "loss": 0.1943,
+      "step": 10878
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8318506761180977,
+      "learning_rate": 2.2642582011815858e-07,
+      "loss": 0.312,
+      "step": 10879
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7972055452495757,
+      "learning_rate": 2.2633938200186203e-07,
+      "loss": 0.1897,
+      "step": 10880
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8801000207391041,
+      "learning_rate": 2.2625295556047736e-07,
+      "loss": 0.2444,
+      "step": 10881
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4887721157384821,
+      "learning_rate": 2.261665407976921e-07,
+      "loss": 0.1556,
+      "step": 10882
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0530104310733335,
+      "learning_rate": 2.260801377171927e-07,
+      "loss": 0.1596,
+      "step": 10883
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3159872398057218,
+      "learning_rate": 2.2599374632266511e-07,
+      "loss": 0.1374,
+      "step": 10884
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.611180244577775,
+      "learning_rate": 2.259073666177951e-07,
+      "loss": 0.2115,
+      "step": 10885
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0627240304031664,
+      "learning_rate": 2.2582099860626797e-07,
+      "loss": 0.0233,
+      "step": 10886
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6653248825304307,
+      "learning_rate": 2.257346422917681e-07,
+      "loss": 0.1105,
+      "step": 10887
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9950886138225368,
+      "learning_rate": 2.2564829767797965e-07,
+      "loss": 0.1554,
+      "step": 10888
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.46491460431634946,
+      "learning_rate": 2.2556196476858657e-07,
+      "loss": 0.0984,
+      "step": 10889
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.293669330584775,
+      "learning_rate": 2.2547564356727155e-07,
+      "loss": 0.2248,
+      "step": 10890
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.088819784265053,
+      "learning_rate": 2.2538933407771766e-07,
+      "loss": 0.0916,
+      "step": 10891
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.110023347891709,
+      "learning_rate": 2.2530303630360664e-07,
+      "loss": 0.0023,
+      "step": 10892
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9093858880895302,
+      "learning_rate": 2.2521675024862047e-07,
+      "loss": 0.3635,
+      "step": 10893
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7695666115831168,
+      "learning_rate": 2.2513047591643992e-07,
+      "loss": 0.13,
+      "step": 10894
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.32660121975726,
+      "learning_rate": 2.2504421331074602e-07,
+      "loss": 0.1307,
+      "step": 10895
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5935487063161089,
+      "learning_rate": 2.249579624352187e-07,
+      "loss": 0.181,
+      "step": 10896
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7901801674782892,
+      "learning_rate": 2.2487172329353742e-07,
+      "loss": 0.239,
+      "step": 10897
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7021555975048261,
+      "learning_rate": 2.2478549588938166e-07,
+      "loss": 0.233,
+      "step": 10898
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5346683612322068,
+      "learning_rate": 2.2469928022642975e-07,
+      "loss": 0.1757,
+      "step": 10899
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5942130167894475,
+      "learning_rate": 2.2461307630836019e-07,
+      "loss": 0.4709,
+      "step": 10900
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7249753989402374,
+      "learning_rate": 2.2452688413885013e-07,
+      "loss": 0.0288,
+      "step": 10901
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.289366193772501,
+      "learning_rate": 2.244407037215772e-07,
+      "loss": 0.1315,
+      "step": 10902
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.954304440433352,
+      "learning_rate": 2.2435453506021756e-07,
+      "loss": 0.1926,
+      "step": 10903
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2178358002772534,
+      "learning_rate": 2.2426837815844763e-07,
+      "loss": 0.1126,
+      "step": 10904
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.44983159921230526,
+      "learning_rate": 2.2418223301994293e-07,
+      "loss": 0.1848,
+      "step": 10905
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1048548497985862,
+      "learning_rate": 2.2409609964837883e-07,
+      "loss": 0.1964,
+      "step": 10906
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.967727743154967,
+      "learning_rate": 2.2400997804742961e-07,
+      "loss": 0.1552,
+      "step": 10907
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.13480417299554695,
+      "learning_rate": 2.239238682207697e-07,
+      "loss": 0.0921,
+      "step": 10908
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4985602003788864,
+      "learning_rate": 2.238377701720725e-07,
+      "loss": 0.1493,
+      "step": 10909
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.815357305600492,
+      "learning_rate": 2.237516839050111e-07,
+      "loss": 0.3343,
+      "step": 10910
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9960945528187201,
+      "learning_rate": 2.2366560942325828e-07,
+      "loss": 0.2111,
+      "step": 10911
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4115713506311849,
+      "learning_rate": 2.2357954673048591e-07,
+      "loss": 0.0647,
+      "step": 10912
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0007216294650185,
+      "learning_rate": 2.2349349583036598e-07,
+      "loss": 0.1253,
+      "step": 10913
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.821385428051082,
+      "learning_rate": 2.2340745672656914e-07,
+      "loss": 0.0354,
+      "step": 10914
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.586873288271951,
+      "learning_rate": 2.2332142942276639e-07,
+      "loss": 0.0837,
+      "step": 10915
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7014841527889852,
+      "learning_rate": 2.2323541392262745e-07,
+      "loss": 0.196,
+      "step": 10916
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.967562145076491,
+      "learning_rate": 2.2314941022982237e-07,
+      "loss": 0.3864,
+      "step": 10917
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9130360703446458,
+      "learning_rate": 2.230634183480198e-07,
+      "loss": 0.2439,
+      "step": 10918
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5969503230851588,
+      "learning_rate": 2.2297743828088867e-07,
+      "loss": 0.0972,
+      "step": 10919
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.496305039123906,
+      "learning_rate": 2.228914700320967e-07,
+      "loss": 0.1178,
+      "step": 10920
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2448276710422292,
+      "learning_rate": 2.2280551360531168e-07,
+      "loss": 0.0335,
+      "step": 10921
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0203334600594705,
+      "learning_rate": 2.227195690042009e-07,
+      "loss": 0.195,
+      "step": 10922
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7703721701369108,
+      "learning_rate": 2.2263363623243054e-07,
+      "loss": 0.2266,
+      "step": 10923
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.049210719399762,
+      "learning_rate": 2.22547715293667e-07,
+      "loss": 0.0976,
+      "step": 10924
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.135960356496401,
+      "learning_rate": 2.2246180619157546e-07,
+      "loss": 0.0766,
+      "step": 10925
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.31557956273237464,
+      "learning_rate": 2.2237590892982138e-07,
+      "loss": 0.2627,
+      "step": 10926
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.608551494496266,
+      "learning_rate": 2.22290023512069e-07,
+      "loss": 0.1915,
+      "step": 10927
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.35726329297766696,
+      "learning_rate": 2.2220414994198262e-07,
+      "loss": 0.1034,
+      "step": 10928
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.886101922517762,
+      "learning_rate": 2.2211828822322544e-07,
+      "loss": 0.0925,
+      "step": 10929
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1361921865049236,
+      "learning_rate": 2.220324383594609e-07,
+      "loss": 0.104,
+      "step": 10930
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.17799373833339124,
+      "learning_rate": 2.2194660035435115e-07,
+      "loss": 0.021,
+      "step": 10931
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.42703875159493,
+      "learning_rate": 2.2186077421155853e-07,
+      "loss": 0.0421,
+      "step": 10932
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6544839328708031,
+      "learning_rate": 2.2177495993474426e-07,
+      "loss": 0.0781,
+      "step": 10933
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5053969140401248,
+      "learning_rate": 2.2168915752756966e-07,
+      "loss": 0.1136,
+      "step": 10934
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4204335796883389,
+      "learning_rate": 2.2160336699369496e-07,
+      "loss": 0.2632,
+      "step": 10935
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1065380945059267,
+      "learning_rate": 2.215175883367804e-07,
+      "loss": 0.2669,
+      "step": 10936
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.061545908147558,
+      "learning_rate": 2.2143182156048533e-07,
+      "loss": 0.0885,
+      "step": 10937
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7378857275320583,
+      "learning_rate": 2.213460666684686e-07,
+      "loss": 0.2206,
+      "step": 10938
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.050980108589889,
+      "learning_rate": 2.2126032366438884e-07,
+      "loss": 0.18,
+      "step": 10939
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3873979101593528,
+      "learning_rate": 2.2117459255190395e-07,
+      "loss": 0.2091,
+      "step": 10940
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.81192411700409,
+      "learning_rate": 2.2108887333467168e-07,
+      "loss": 0.2052,
+      "step": 10941
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1025429172170629,
+      "learning_rate": 2.2100316601634856e-07,
+      "loss": 0.2563,
+      "step": 10942
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6293557518228154,
+      "learning_rate": 2.2091747060059141e-07,
+      "loss": 0.3961,
+      "step": 10943
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5601243607070677,
+      "learning_rate": 2.2083178709105583e-07,
+      "loss": 0.1867,
+      "step": 10944
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.018311267316071,
+      "learning_rate": 2.2074611549139754e-07,
+      "loss": 0.2061,
+      "step": 10945
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7922089652905071,
+      "learning_rate": 2.206604558052712e-07,
+      "loss": 0.1672,
+      "step": 10946
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7070052739808517,
+      "learning_rate": 2.2057480803633154e-07,
+      "loss": 0.2281,
+      "step": 10947
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7016793010266547,
+      "learning_rate": 2.204891721882321e-07,
+      "loss": 0.2014,
+      "step": 10948
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8402525364923409,
+      "learning_rate": 2.2040354826462664e-07,
+      "loss": 0.2437,
+      "step": 10949
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5643340185649347,
+      "learning_rate": 2.2031793626916768e-07,
+      "loss": 0.1824,
+      "step": 10950
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8265373224511069,
+      "learning_rate": 2.2023233620550797e-07,
+      "loss": 0.1035,
+      "step": 10951
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.348838993143465,
+      "learning_rate": 2.2014674807729923e-07,
+      "loss": 0.0987,
+      "step": 10952
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7134912311271417,
+      "learning_rate": 2.2006117188819257e-07,
+      "loss": 0.2111,
+      "step": 10953
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.039798570227404,
+      "learning_rate": 2.1997560764183926e-07,
+      "loss": 0.1505,
+      "step": 10954
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2631960550321324,
+      "learning_rate": 2.1989005534188927e-07,
+      "loss": 0.1331,
+      "step": 10955
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.41869817689472827,
+      "learning_rate": 2.198045149919926e-07,
+      "loss": 0.1506,
+      "step": 10956
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6820577255079567,
+      "learning_rate": 2.197189865957988e-07,
+      "loss": 0.0211,
+      "step": 10957
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8595514898710638,
+      "learning_rate": 2.1963347015695627e-07,
+      "loss": 0.1329,
+      "step": 10958
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4475700125823077,
+      "learning_rate": 2.195479656791135e-07,
+      "loss": 0.2221,
+      "step": 10959
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7665736691231024,
+      "learning_rate": 2.1946247316591843e-07,
+      "loss": 0.0957,
+      "step": 10960
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5043782022654189,
+      "learning_rate": 2.193769926210181e-07,
+      "loss": 0.0801,
+      "step": 10961
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1340649885550248,
+      "learning_rate": 2.1929152404805956e-07,
+      "loss": 0.0045,
+      "step": 10962
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9196294150621533,
+      "learning_rate": 2.1920606745068864e-07,
+      "loss": 0.2139,
+      "step": 10963
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5396771173298256,
+      "learning_rate": 2.1912062283255162e-07,
+      "loss": 0.0046,
+      "step": 10964
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6335613203423425,
+      "learning_rate": 2.1903519019729343e-07,
+      "loss": 0.1939,
+      "step": 10965
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7146708155614051,
+      "learning_rate": 2.1894976954855865e-07,
+      "loss": 0.3615,
+      "step": 10966
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.39300506891577236,
+      "learning_rate": 2.188643608899919e-07,
+      "loss": 0.2438,
+      "step": 10967
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.23469147985959943,
+      "learning_rate": 2.1877896422523644e-07,
+      "loss": 0.0026,
+      "step": 10968
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.659546256345295,
+      "learning_rate": 2.1869357955793594e-07,
+      "loss": 0.1957,
+      "step": 10969
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1810606116773537,
+      "learning_rate": 2.186082068917326e-07,
+      "loss": 0.3302,
+      "step": 10970
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2789898034190668,
+      "learning_rate": 2.18522846230269e-07,
+      "loss": 0.1007,
+      "step": 10971
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8152696515338478,
+      "learning_rate": 2.1843749757718642e-07,
+      "loss": 0.2352,
+      "step": 10972
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.506016568426306,
+      "learning_rate": 2.1835216093612646e-07,
+      "loss": 0.098,
+      "step": 10973
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3521813685806305,
+      "learning_rate": 2.182668363107293e-07,
+      "loss": 0.0706,
+      "step": 10974
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.784555173968003,
+      "learning_rate": 2.1818152370463523e-07,
+      "loss": 0.0722,
+      "step": 10975
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.952886471051704,
+      "learning_rate": 2.1809622312148402e-07,
+      "loss": 0.2057,
+      "step": 10976
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.772372475430498,
+      "learning_rate": 2.180109345649145e-07,
+      "loss": 0.1928,
+      "step": 10977
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3889697536058163,
+      "learning_rate": 2.1792565803856556e-07,
+      "loss": 0.0522,
+      "step": 10978
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8443163771593503,
+      "learning_rate": 2.178403935460749e-07,
+      "loss": 0.1573,
+      "step": 10979
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.549392684872409,
+      "learning_rate": 2.1775514109108046e-07,
+      "loss": 0.1038,
+      "step": 10980
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2069482018807078,
+      "learning_rate": 2.176699006772189e-07,
+      "loss": 0.2588,
+      "step": 10981
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.342735477229369,
+      "learning_rate": 2.175846723081271e-07,
+      "loss": 0.3343,
+      "step": 10982
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6949236444913175,
+      "learning_rate": 2.1749945598744075e-07,
+      "loss": 0.0976,
+      "step": 10983
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4281699533945873,
+      "learning_rate": 2.1741425171879563e-07,
+      "loss": 0.2621,
+      "step": 10984
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0020370195970825,
+      "learning_rate": 2.1732905950582648e-07,
+      "loss": 0.1802,
+      "step": 10985
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0932002911407386,
+      "learning_rate": 2.1724387935216802e-07,
+      "loss": 0.1798,
+      "step": 10986
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2957926083056028,
+      "learning_rate": 2.171587112614539e-07,
+      "loss": 0.2829,
+      "step": 10987
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.253034759683275,
+      "learning_rate": 2.1707355523731796e-07,
+      "loss": 0.1463,
+      "step": 10988
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.34662369243637214,
+      "learning_rate": 2.1698841128339273e-07,
+      "loss": 0.0779,
+      "step": 10989
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8748200788607298,
+      "learning_rate": 2.1690327940331093e-07,
+      "loss": 0.163,
+      "step": 10990
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9269464993348918,
+      "learning_rate": 2.168181596007042e-07,
+      "loss": 0.118,
+      "step": 10991
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5689704725961378,
+      "learning_rate": 2.167330518792042e-07,
+      "loss": 0.1489,
+      "step": 10992
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1384217527504339,
+      "learning_rate": 2.1664795624244147e-07,
+      "loss": 0.0393,
+      "step": 10993
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7500700286501576,
+      "learning_rate": 2.1656287269404656e-07,
+      "loss": 0.1911,
+      "step": 10994
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1563941203100896,
+      "learning_rate": 2.1647780123764942e-07,
+      "loss": 0.1127,
+      "step": 10995
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7440228102081966,
+      "learning_rate": 2.1639274187687906e-07,
+      "loss": 0.0976,
+      "step": 10996
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3423135838816568,
+      "learning_rate": 2.1630769461536463e-07,
+      "loss": 0.0711,
+      "step": 10997
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6920210219335216,
+      "learning_rate": 2.1622265945673406e-07,
+      "loss": 0.2363,
+      "step": 10998
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.477111108642288,
+      "learning_rate": 2.1613763640461552e-07,
+      "loss": 0.1506,
+      "step": 10999
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.42656076104498575,
+      "learning_rate": 2.1605262546263587e-07,
+      "loss": 0.047,
+      "step": 11000
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8438581654792345,
+      "learning_rate": 2.1596762663442213e-07,
+      "loss": 0.0957,
+      "step": 11001
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.809254864804685,
+      "learning_rate": 2.158826399236003e-07,
+      "loss": 0.1341,
+      "step": 11002
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8724682548909874,
+      "learning_rate": 2.1579766533379635e-07,
+      "loss": 0.3811,
+      "step": 11003
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2784828289476305,
+      "learning_rate": 2.1571270286863514e-07,
+      "loss": 0.2726,
+      "step": 11004
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.325186343899838,
+      "learning_rate": 2.156277525317417e-07,
+      "loss": 0.2104,
+      "step": 11005
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7809667381097108,
+      "learning_rate": 2.1554281432674e-07,
+      "loss": 0.203,
+      "step": 11006
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5387177855910273,
+      "learning_rate": 2.1545788825725348e-07,
+      "loss": 0.2158,
+      "step": 11007
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7752514065628473,
+      "learning_rate": 2.1537297432690566e-07,
+      "loss": 0.0828,
+      "step": 11008
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6494091713904622,
+      "learning_rate": 2.152880725393187e-07,
+      "loss": 0.1067,
+      "step": 11009
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8775734501206596,
+      "learning_rate": 2.152031828981149e-07,
+      "loss": 0.4599,
+      "step": 11010
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4376476542465838,
+      "learning_rate": 2.151183054069159e-07,
+      "loss": 0.1312,
+      "step": 11011
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.1787140873996134,
+      "learning_rate": 2.1503344006934283e-07,
+      "loss": 0.1617,
+      "step": 11012
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9153597476126069,
+      "learning_rate": 2.1494858688901585e-07,
+      "loss": 0.2475,
+      "step": 11013
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7634750254238873,
+      "learning_rate": 2.1486374586955535e-07,
+      "loss": 0.1348,
+      "step": 11014
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2705824767832749,
+      "learning_rate": 2.1477891701458052e-07,
+      "loss": 0.098,
+      "step": 11015
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5972037289624058,
+      "learning_rate": 2.1469410032771056e-07,
+      "loss": 0.0561,
+      "step": 11016
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.543764948171805,
+      "learning_rate": 2.146092958125637e-07,
+      "loss": 0.1716,
+      "step": 11017
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.20522270832625228,
+      "learning_rate": 2.145245034727582e-07,
+      "loss": 0.0167,
+      "step": 11018
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2037803633222162,
+      "learning_rate": 2.1443972331191118e-07,
+      "loss": 0.0692,
+      "step": 11019
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4281920119549205,
+      "learning_rate": 2.1435495533363946e-07,
+      "loss": 0.1621,
+      "step": 11020
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1511074131109007,
+      "learning_rate": 2.1427019954155979e-07,
+      "loss": 0.3641,
+      "step": 11021
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4625029465831232,
+      "learning_rate": 2.1418545593928756e-07,
+      "loss": 0.3489,
+      "step": 11022
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3047263123238109,
+      "learning_rate": 2.1410072453043853e-07,
+      "loss": 0.1995,
+      "step": 11023
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.369664270063948,
+      "learning_rate": 2.1401600531862713e-07,
+      "loss": 0.1859,
+      "step": 11024
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.689149430334241,
+      "learning_rate": 2.1393129830746804e-07,
+      "loss": 0.1718,
+      "step": 11025
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6200892886489244,
+      "learning_rate": 2.1384660350057465e-07,
+      "loss": 0.1687,
+      "step": 11026
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.706330514924541,
+      "learning_rate": 2.1376192090156043e-07,
+      "loss": 0.1297,
+      "step": 11027
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0425024830100322,
+      "learning_rate": 2.1367725051403817e-07,
+      "loss": 0.3272,
+      "step": 11028
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.946566279496054,
+      "learning_rate": 2.1359259234161985e-07,
+      "loss": 0.1678,
+      "step": 11029
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.48638936657029846,
+      "learning_rate": 2.135079463879173e-07,
+      "loss": 0.0274,
+      "step": 11030
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5038371680377733,
+      "learning_rate": 2.134233126565419e-07,
+      "loss": 0.1905,
+      "step": 11031
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9346990238157602,
+      "learning_rate": 2.1333869115110382e-07,
+      "loss": 0.5061,
+      "step": 11032
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0362277480262794,
+      "learning_rate": 2.1325408187521364e-07,
+      "loss": 0.167,
+      "step": 11033
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.861370803338594,
+      "learning_rate": 2.1316948483248082e-07,
+      "loss": 0.2061,
+      "step": 11034
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7567563176852425,
+      "learning_rate": 2.1308490002651413e-07,
+      "loss": 0.2294,
+      "step": 11035
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.30553706264457836,
+      "learning_rate": 2.1300032746092257e-07,
+      "loss": 0.0625,
+      "step": 11036
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7933743165339,
+      "learning_rate": 2.129157671393138e-07,
+      "loss": 0.1537,
+      "step": 11037
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.335773703985514,
+      "learning_rate": 2.1283121906529571e-07,
+      "loss": 0.3533,
+      "step": 11038
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4964445480855572,
+      "learning_rate": 2.1274668324247491e-07,
+      "loss": 0.2921,
+      "step": 11039
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.930152862201858,
+      "learning_rate": 2.126621596744582e-07,
+      "loss": 0.1646,
+      "step": 11040
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.248724251053854,
+      "learning_rate": 2.1257764836485127e-07,
+      "loss": 0.2651,
+      "step": 11041
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3113870893272026,
+      "learning_rate": 2.1249314931725974e-07,
+      "loss": 0.2208,
+      "step": 11042
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4539614691084134,
+      "learning_rate": 2.1240866253528832e-07,
+      "loss": 0.2331,
+      "step": 11043
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.506758943529494,
+      "learning_rate": 2.1232418802254165e-07,
+      "loss": 0.3359,
+      "step": 11044
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.48030673453778105,
+      "learning_rate": 2.1223972578262322e-07,
+      "loss": 0.0153,
+      "step": 11045
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8853162385891539,
+      "learning_rate": 2.1215527581913655e-07,
+      "loss": 0.1202,
+      "step": 11046
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7863355873264025,
+      "learning_rate": 2.1207083813568466e-07,
+      "loss": 0.2481,
+      "step": 11047
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0899537164999757,
+      "learning_rate": 2.1198641273586947e-07,
+      "loss": 0.2548,
+      "step": 11048
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.11510099776902,
+      "learning_rate": 2.1190199962329302e-07,
+      "loss": 0.026,
+      "step": 11049
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6448255195451303,
+      "learning_rate": 2.1181759880155625e-07,
+      "loss": 0.2012,
+      "step": 11050
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3651832298687614,
+      "learning_rate": 2.1173321027426022e-07,
+      "loss": 0.0762,
+      "step": 11051
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7536512939188647,
+      "learning_rate": 2.1164883404500476e-07,
+      "loss": 0.1388,
+      "step": 11052
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9228609114338816,
+      "learning_rate": 2.1156447011738986e-07,
+      "loss": 0.064,
+      "step": 11053
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2032538379935707,
+      "learning_rate": 2.1148011849501436e-07,
+      "loss": 0.1504,
+      "step": 11054
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7894577062338723,
+      "learning_rate": 2.1139577918147711e-07,
+      "loss": 0.3238,
+      "step": 11055
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2614407387535882,
+      "learning_rate": 2.1131145218037593e-07,
+      "loss": 0.0887,
+      "step": 11056
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.045079877596098,
+      "learning_rate": 2.1122713749530875e-07,
+      "loss": 0.2197,
+      "step": 11057
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5737296747467094,
+      "learning_rate": 2.1114283512987218e-07,
+      "loss": 0.223,
+      "step": 11058
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.231003810412058,
+      "learning_rate": 2.1105854508766312e-07,
+      "loss": 0.2532,
+      "step": 11059
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 16.87974904872534,
+      "learning_rate": 2.1097426737227724e-07,
+      "loss": 0.1487,
+      "step": 11060
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6033239733857554,
+      "learning_rate": 2.1089000198731027e-07,
+      "loss": 0.0204,
+      "step": 11061
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7984236334952097,
+      "learning_rate": 2.1080574893635683e-07,
+      "loss": 0.1942,
+      "step": 11062
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9847381829793511,
+      "learning_rate": 2.1072150822301167e-07,
+      "loss": 0.0893,
+      "step": 11063
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7396772250488312,
+      "learning_rate": 2.1063727985086827e-07,
+      "loss": 0.1299,
+      "step": 11064
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5982581324806512,
+      "learning_rate": 2.1055306382352022e-07,
+      "loss": 0.2079,
+      "step": 11065
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.247004817230564,
+      "learning_rate": 2.1046886014456054e-07,
+      "loss": 0.2978,
+      "step": 11066
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2188244200151201,
+      "learning_rate": 2.1038466881758116e-07,
+      "loss": 0.2978,
+      "step": 11067
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0301709760929476,
+      "learning_rate": 2.1030048984617416e-07,
+      "loss": 0.2754,
+      "step": 11068
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.343686261228263,
+      "learning_rate": 2.102163232339304e-07,
+      "loss": 0.2055,
+      "step": 11069
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0485427604293511,
+      "learning_rate": 2.1013216898444109e-07,
+      "loss": 0.3121,
+      "step": 11070
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1570134171494262,
+      "learning_rate": 2.1004802710129592e-07,
+      "loss": 0.1822,
+      "step": 11071
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5084985592317486,
+      "learning_rate": 2.0996389758808498e-07,
+      "loss": 0.0521,
+      "step": 11072
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4328263965227327,
+      "learning_rate": 2.0987978044839705e-07,
+      "loss": 0.0071,
+      "step": 11073
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.42636946733427256,
+      "learning_rate": 2.0979567568582108e-07,
+      "loss": 0.1459,
+      "step": 11074
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8818847458483785,
+      "learning_rate": 2.0971158330394495e-07,
+      "loss": 0.4688,
+      "step": 11075
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6775631918795026,
+      "learning_rate": 2.096275033063561e-07,
+      "loss": 0.019,
+      "step": 11076
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3100042645637124,
+      "learning_rate": 2.0954343569664185e-07,
+      "loss": 0.1081,
+      "step": 11077
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3202207840361624,
+      "learning_rate": 2.094593804783883e-07,
+      "loss": 0.2358,
+      "step": 11078
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1424883672675796,
+      "learning_rate": 2.0937533765518185e-07,
+      "loss": 0.0199,
+      "step": 11079
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.759146111867244,
+      "learning_rate": 2.0929130723060752e-07,
+      "loss": 0.257,
+      "step": 11080
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8304879441262054,
+      "learning_rate": 2.0920728920825043e-07,
+      "loss": 0.1505,
+      "step": 11081
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.29801680293990124,
+      "learning_rate": 2.0912328359169495e-07,
+      "loss": 0.0428,
+      "step": 11082
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3461982760630799,
+      "learning_rate": 2.090392903845251e-07,
+      "loss": 0.0044,
+      "step": 11083
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0081329296310395,
+      "learning_rate": 2.0895530959032388e-07,
+      "loss": 0.3769,
+      "step": 11084
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8026771160791514,
+      "learning_rate": 2.0887134121267437e-07,
+      "loss": 0.2868,
+      "step": 11085
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.45385461846890257,
+      "learning_rate": 2.0878738525515853e-07,
+      "loss": 0.2828,
+      "step": 11086
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.41979306423280627,
+      "learning_rate": 2.087034417213584e-07,
+      "loss": 0.2649,
+      "step": 11087
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.705398689810544,
+      "learning_rate": 2.0861951061485504e-07,
+      "loss": 0.1281,
+      "step": 11088
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0492582653181994,
+      "learning_rate": 2.0853559193922892e-07,
+      "loss": 0.2019,
+      "step": 11089
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.050281563306293,
+      "learning_rate": 2.084516856980606e-07,
+      "loss": 0.2285,
+      "step": 11090
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6808190367480089,
+      "learning_rate": 2.083677918949292e-07,
+      "loss": 0.011,
+      "step": 11091
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.999061353887269,
+      "learning_rate": 2.0828391053341427e-07,
+      "loss": 0.1825,
+      "step": 11092
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2763628861309195,
+      "learning_rate": 2.0820004161709393e-07,
+      "loss": 0.2497,
+      "step": 11093
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.360521841048718,
+      "learning_rate": 2.0811618514954665e-07,
+      "loss": 0.158,
+      "step": 11094
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9218195814754128,
+      "learning_rate": 2.0803234113434942e-07,
+      "loss": 0.1443,
+      "step": 11095
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6991316452549218,
+      "learning_rate": 2.0794850957507965e-07,
+      "loss": 0.1132,
+      "step": 11096
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.744254439839101,
+      "learning_rate": 2.0786469047531341e-07,
+      "loss": 0.2021,
+      "step": 11097
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8041557797444049,
+      "learning_rate": 2.0778088383862686e-07,
+      "loss": 0.3431,
+      "step": 11098
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6138279733706382,
+      "learning_rate": 2.0769708966859512e-07,
+      "loss": 0.1224,
+      "step": 11099
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6659031859543398,
+      "learning_rate": 2.0761330796879306e-07,
+      "loss": 0.1869,
+      "step": 11100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.576919203865755,
+      "learning_rate": 2.075295387427951e-07,
+      "loss": 0.3169,
+      "step": 11101
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0270864257226362,
+      "learning_rate": 2.0744578199417519e-07,
+      "loss": 0.1338,
+      "step": 11102
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9645019146817351,
+      "learning_rate": 2.0736203772650622e-07,
+      "loss": 0.2736,
+      "step": 11103
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1034516439933453,
+      "learning_rate": 2.0727830594336088e-07,
+      "loss": 0.3325,
+      "step": 11104
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6279409222422017,
+      "learning_rate": 2.0719458664831164e-07,
+      "loss": 0.1725,
+      "step": 11105
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0000270743273245,
+      "learning_rate": 2.0711087984492976e-07,
+      "loss": 0.3225,
+      "step": 11106
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.489929596811043,
+      "learning_rate": 2.0702718553678672e-07,
+      "loss": 0.1607,
+      "step": 11107
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5464027832140982,
+      "learning_rate": 2.0694350372745267e-07,
+      "loss": 0.0064,
+      "step": 11108
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4753609948604213,
+      "learning_rate": 2.068598344204981e-07,
+      "loss": 0.1583,
+      "step": 11109
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0716566064726825,
+      "learning_rate": 2.0677617761949206e-07,
+      "loss": 0.2261,
+      "step": 11110
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.31560982199835835,
+      "learning_rate": 2.066925333280039e-07,
+      "loss": 0.0317,
+      "step": 11111
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8557832334185258,
+      "learning_rate": 2.0660890154960175e-07,
+      "loss": 0.2705,
+      "step": 11112
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.16970641697493014,
+      "learning_rate": 2.0652528228785382e-07,
+      "loss": 0.1145,
+      "step": 11113
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7323255733169622,
+      "learning_rate": 2.0644167554632712e-07,
+      "loss": 0.1083,
+      "step": 11114
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.423845452669845,
+      "learning_rate": 2.0635808132858883e-07,
+      "loss": 0.2142,
+      "step": 11115
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7703557994381832,
+      "learning_rate": 2.0627449963820494e-07,
+      "loss": 0.0975,
+      "step": 11116
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.397097604885472,
+      "learning_rate": 2.0619093047874136e-07,
+      "loss": 0.3368,
+      "step": 11117
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.648458511106753,
+      "learning_rate": 2.0610737385376348e-07,
+      "loss": 0.0169,
+      "step": 11118
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4159630617314296,
+      "learning_rate": 2.0602382976683564e-07,
+      "loss": 0.3397,
+      "step": 11119
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.40372083127383873,
+      "learning_rate": 2.0594029822152238e-07,
+      "loss": 0.1969,
+      "step": 11120
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3412550776805925,
+      "learning_rate": 2.0585677922138694e-07,
+      "loss": 0.3395,
+      "step": 11121
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8002680589122609,
+      "learning_rate": 2.057732727699928e-07,
+      "loss": 0.3023,
+      "step": 11122
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.79165682852856,
+      "learning_rate": 2.056897788709021e-07,
+      "loss": 0.2777,
+      "step": 11123
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6588782726645641,
+      "learning_rate": 2.0560629752767727e-07,
+      "loss": 0.2617,
+      "step": 11124
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9772970872969657,
+      "learning_rate": 2.0552282874387944e-07,
+      "loss": 0.1517,
+      "step": 11125
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.680089317186349,
+      "learning_rate": 2.0543937252306986e-07,
+      "loss": 0.1008,
+      "step": 11126
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.16421178474496412,
+      "learning_rate": 2.0535592886880858e-07,
+      "loss": 0.0016,
+      "step": 11127
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0169762939722602,
+      "learning_rate": 2.0527249778465595e-07,
+      "loss": 0.3602,
+      "step": 11128
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4693658530747357,
+      "learning_rate": 2.051890792741708e-07,
+      "loss": 0.1672,
+      "step": 11129
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.377894875031313,
+      "learning_rate": 2.0510567334091233e-07,
+      "loss": 0.3077,
+      "step": 11130
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9291163569811182,
+      "learning_rate": 2.050222799884387e-07,
+      "loss": 0.1542,
+      "step": 11131
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8849080450106809,
+      "learning_rate": 2.0493889922030738e-07,
+      "loss": 0.2231,
+      "step": 11132
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9583189714214655,
+      "learning_rate": 2.048555310400758e-07,
+      "loss": 0.3332,
+      "step": 11133
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.33436524652268007,
+      "learning_rate": 2.0477217545130072e-07,
+      "loss": 0.0046,
+      "step": 11134
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3954086713344425,
+      "learning_rate": 2.0468883245753794e-07,
+      "loss": 0.0819,
+      "step": 11135
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.39441446712178774,
+      "learning_rate": 2.0460550206234323e-07,
+      "loss": 0.2405,
+      "step": 11136
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7869768509866532,
+      "learning_rate": 2.0452218426927182e-07,
+      "loss": 0.1657,
+      "step": 11137
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.31599890181855755,
+      "learning_rate": 2.0443887908187778e-07,
+      "loss": 0.1003,
+      "step": 11138
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9200400092405193,
+      "learning_rate": 2.0435558650371553e-07,
+      "loss": 0.2184,
+      "step": 11139
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5725922337780283,
+      "learning_rate": 2.042723065383381e-07,
+      "loss": 0.2143,
+      "step": 11140
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8036918149895718,
+      "learning_rate": 2.0418903918929875e-07,
+      "loss": 0.1671,
+      "step": 11141
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0754672093221846,
+      "learning_rate": 2.0410578446014943e-07,
+      "loss": 0.1978,
+      "step": 11142
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.009860347668434,
+      "learning_rate": 2.0402254235444237e-07,
+      "loss": 0.263,
+      "step": 11143
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0077324420036544,
+      "learning_rate": 2.0393931287572863e-07,
+      "loss": 0.4108,
+      "step": 11144
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7975073861023216,
+      "learning_rate": 2.0385609602755877e-07,
+      "loss": 0.2923,
+      "step": 11145
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7617043569621263,
+      "learning_rate": 2.0377289181348338e-07,
+      "loss": 0.2321,
+      "step": 11146
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8149967886397724,
+      "learning_rate": 2.0368970023705174e-07,
+      "loss": 0.1248,
+      "step": 11147
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3126689947528291,
+      "learning_rate": 2.0360652130181329e-07,
+      "loss": 0.0073,
+      "step": 11148
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5347447803184486,
+      "learning_rate": 2.0352335501131634e-07,
+      "loss": 0.0089,
+      "step": 11149
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1174960807730312,
+      "learning_rate": 2.0344020136910915e-07,
+      "loss": 0.1651,
+      "step": 11150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3891793644539086,
+      "learning_rate": 2.0335706037873907e-07,
+      "loss": 0.1932,
+      "step": 11151
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4109825230115582,
+      "learning_rate": 2.0327393204375303e-07,
+      "loss": 0.0103,
+      "step": 11152
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.44008456427840037,
+      "learning_rate": 2.0319081636769775e-07,
+      "loss": 0.0624,
+      "step": 11153
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6804454130050315,
+      "learning_rate": 2.0310771335411876e-07,
+      "loss": 0.1748,
+      "step": 11154
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6616182441697905,
+      "learning_rate": 2.0302462300656148e-07,
+      "loss": 0.1779,
+      "step": 11155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.194886232200801,
+      "learning_rate": 2.02941545328571e-07,
+      "loss": 0.0157,
+      "step": 11156
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0210193060768187,
+      "learning_rate": 2.0285848032369136e-07,
+      "loss": 0.1991,
+      "step": 11157
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.42220279880617184,
+      "learning_rate": 2.027754279954661e-07,
+      "loss": 0.0531,
+      "step": 11158
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3413695279247972,
+      "learning_rate": 2.0269238834743873e-07,
+      "loss": 0.0783,
+      "step": 11159
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1783040252237527,
+      "learning_rate": 2.026093613831516e-07,
+      "loss": 0.0881,
+      "step": 11160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.352888585774589,
+      "learning_rate": 2.0252634710614708e-07,
+      "loss": 0.1885,
+      "step": 11161
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7125407256154703,
+      "learning_rate": 2.0244334551996644e-07,
+      "loss": 0.207,
+      "step": 11162
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0810572743366624,
+      "learning_rate": 2.02360356628151e-07,
+      "loss": 0.1821,
+      "step": 11163
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9800343033106649,
+      "learning_rate": 2.0227738043424092e-07,
+      "loss": 0.0958,
+      "step": 11164
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1925215461899459,
+      "learning_rate": 2.0219441694177646e-07,
+      "loss": 0.13,
+      "step": 11165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.488269609454193,
+      "learning_rate": 2.0211146615429663e-07,
+      "loss": 0.0184,
+      "step": 11166
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5923949634576661,
+      "learning_rate": 2.0202852807534072e-07,
+      "loss": 0.1575,
+      "step": 11167
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0325760284176058,
+      "learning_rate": 2.0194560270844656e-07,
+      "loss": 0.2335,
+      "step": 11168
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.616832799167069,
+      "learning_rate": 2.0186269005715238e-07,
+      "loss": 0.0999,
+      "step": 11169
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.24812581154032837,
+      "learning_rate": 2.0177979012499496e-07,
+      "loss": 0.0109,
+      "step": 11170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7395334250891737,
+      "learning_rate": 2.0169690291551122e-07,
+      "loss": 0.1651,
+      "step": 11171
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.678953458118523,
+      "learning_rate": 2.0161402843223746e-07,
+      "loss": 0.2679,
+      "step": 11172
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2235245198901974,
+      "learning_rate": 2.0153116667870884e-07,
+      "loss": 0.1278,
+      "step": 11173
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.413046949698867,
+      "learning_rate": 2.0144831765846086e-07,
+      "loss": 0.0327,
+      "step": 11174
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.374152723337225,
+      "learning_rate": 2.0136548137502762e-07,
+      "loss": 0.0914,
+      "step": 11175
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5635293048822592,
+      "learning_rate": 2.0128265783194347e-07,
+      "loss": 0.2037,
+      "step": 11176
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.696824857039066,
+      "learning_rate": 2.0119984703274145e-07,
+      "loss": 0.2893,
+      "step": 11177
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4448810036308517,
+      "learning_rate": 2.0111704898095484e-07,
+      "loss": 0.299,
+      "step": 11178
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4842321762944772,
+      "learning_rate": 2.010342636801155e-07,
+      "loss": 0.2445,
+      "step": 11179
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2315378854292929,
+      "learning_rate": 2.009514911337557e-07,
+      "loss": 0.1501,
+      "step": 11180
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9680214184034847,
+      "learning_rate": 2.0086873134540622e-07,
+      "loss": 0.1119,
+      "step": 11181
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6481634616597933,
+      "learning_rate": 2.007859843185982e-07,
+      "loss": 0.3295,
+      "step": 11182
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8337402788030434,
+      "learning_rate": 2.0070325005686146e-07,
+      "loss": 0.4265,
+      "step": 11183
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1277671168274772,
+      "learning_rate": 2.006205285637258e-07,
+      "loss": 0.1004,
+      "step": 11184
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1033904171117666,
+      "learning_rate": 2.0053781984272028e-07,
+      "loss": 0.2123,
+      "step": 11185
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.831684019145874,
+      "learning_rate": 2.0045512389737317e-07,
+      "loss": 0.1505,
+      "step": 11186
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9398543804586794,
+      "learning_rate": 2.0037244073121268e-07,
+      "loss": 0.0956,
+      "step": 11187
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.33626345490035,
+      "learning_rate": 2.0028977034776617e-07,
+      "loss": 0.1765,
+      "step": 11188
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7330600531312977,
+      "learning_rate": 2.0020711275056068e-07,
+      "loss": 0.2654,
+      "step": 11189
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.071430163760048,
+      "learning_rate": 2.0012446794312232e-07,
+      "loss": 0.1128,
+      "step": 11190
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4393179279597202,
+      "learning_rate": 2.000418359289771e-07,
+      "loss": 0.0701,
+      "step": 11191
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7261389888422688,
+      "learning_rate": 1.9995921671165e-07,
+      "loss": 0.2169,
+      "step": 11192
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6970623023922435,
+      "learning_rate": 1.9987661029466606e-07,
+      "loss": 0.4113,
+      "step": 11193
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2783452929849227,
+      "learning_rate": 1.9979401668154905e-07,
+      "loss": 0.1111,
+      "step": 11194
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7138841297634648,
+      "learning_rate": 1.9971143587582296e-07,
+      "loss": 0.1301,
+      "step": 11195
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7634042944827867,
+      "learning_rate": 1.9962886788101047e-07,
+      "loss": 0.3775,
+      "step": 11196
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8514063734546307,
+      "learning_rate": 1.9954631270063455e-07,
+      "loss": 0.3355,
+      "step": 11197
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.36701638430003297,
+      "learning_rate": 1.9946377033821682e-07,
+      "loss": 0.1232,
+      "step": 11198
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8138968763301555,
+      "learning_rate": 1.9938124079727868e-07,
+      "loss": 0.2306,
+      "step": 11199
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.957162666514625,
+      "learning_rate": 1.9929872408134128e-07,
+      "loss": 0.01,
+      "step": 11200
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.663272276955558,
+      "learning_rate": 1.9921622019392465e-07,
+      "loss": 0.3103,
+      "step": 11201
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6185255139594323,
+      "learning_rate": 1.9913372913854887e-07,
+      "loss": 0.1556,
+      "step": 11202
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.517211480643752,
+      "learning_rate": 1.9905125091873286e-07,
+      "loss": 0.1336,
+      "step": 11203
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.136301256768579,
+      "learning_rate": 1.9896878553799552e-07,
+      "loss": 0.0989,
+      "step": 11204
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5318233109332234,
+      "learning_rate": 1.9888633299985501e-07,
+      "loss": 0.0444,
+      "step": 11205
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6815359606972862,
+      "learning_rate": 1.988038933078287e-07,
+      "loss": 0.1588,
+      "step": 11206
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.064440810281863,
+      "learning_rate": 1.9872146646543385e-07,
+      "loss": 0.1855,
+      "step": 11207
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9657875998630698,
+      "learning_rate": 1.98639052476187e-07,
+      "loss": 0.1648,
+      "step": 11208
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6574026522270735,
+      "learning_rate": 1.9855665134360384e-07,
+      "loss": 0.2409,
+      "step": 11209
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4096503991301412,
+      "learning_rate": 1.984742630712001e-07,
+      "loss": 0.1128,
+      "step": 11210
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5949327327874128,
+      "learning_rate": 1.983918876624902e-07,
+      "loss": 0.1528,
+      "step": 11211
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2938873804657227,
+      "learning_rate": 1.9830952512098887e-07,
+      "loss": 0.3501,
+      "step": 11212
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7971925638047858,
+      "learning_rate": 1.9822717545020968e-07,
+      "loss": 0.1957,
+      "step": 11213
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7505609000325931,
+      "learning_rate": 1.9814483865366565e-07,
+      "loss": 0.452,
+      "step": 11214
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1554099347542885,
+      "learning_rate": 1.9806251473486985e-07,
+      "loss": 0.4074,
+      "step": 11215
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.612928200475958,
+      "learning_rate": 1.9798020369733387e-07,
+      "loss": 0.2485,
+      "step": 11216
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1516989032552714,
+      "learning_rate": 1.9789790554456975e-07,
+      "loss": 0.2485,
+      "step": 11217
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7193849460650937,
+      "learning_rate": 1.9781562028008815e-07,
+      "loss": 0.2538,
+      "step": 11218
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1335841146218795,
+      "learning_rate": 1.9773334790739977e-07,
+      "loss": 0.3156,
+      "step": 11219
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9071096834049142,
+      "learning_rate": 1.9765108843001422e-07,
+      "loss": 0.2181,
+      "step": 11220
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4201123633870905,
+      "learning_rate": 1.9756884185144124e-07,
+      "loss": 0.126,
+      "step": 11221
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.74766202741582,
+      "learning_rate": 1.9748660817518924e-07,
+      "loss": 0.1011,
+      "step": 11222
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5162123783421584,
+      "learning_rate": 1.9740438740476667e-07,
+      "loss": 0.0893,
+      "step": 11223
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8705645587691554,
+      "learning_rate": 1.9732217954368142e-07,
+      "loss": 0.2566,
+      "step": 11224
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7949028927690419,
+      "learning_rate": 1.9723998459544027e-07,
+      "loss": 0.2387,
+      "step": 11225
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.079413507175972,
+      "learning_rate": 1.9715780256355013e-07,
+      "loss": 0.1022,
+      "step": 11226
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1397131273361476,
+      "learning_rate": 1.970756334515168e-07,
+      "loss": 0.0882,
+      "step": 11227
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6633143255294004,
+      "learning_rate": 1.969934772628461e-07,
+      "loss": 0.0782,
+      "step": 11228
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7519215211373192,
+      "learning_rate": 1.9691133400104254e-07,
+      "loss": 0.145,
+      "step": 11229
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8070537655189025,
+      "learning_rate": 1.9682920366961098e-07,
+      "loss": 0.2521,
+      "step": 11230
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1612635234765778,
+      "learning_rate": 1.9674708627205484e-07,
+      "loss": 0.4225,
+      "step": 11231
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6298576287659362,
+      "learning_rate": 1.9666498181187775e-07,
+      "loss": 0.4885,
+      "step": 11232
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6332193205907097,
+      "learning_rate": 1.9658289029258218e-07,
+      "loss": 0.1921,
+      "step": 11233
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9558343194950308,
+      "learning_rate": 1.9650081171767063e-07,
+      "loss": 0.2455,
+      "step": 11234
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.19114937827609,
+      "learning_rate": 1.964187460906444e-07,
+      "loss": 0.1037,
+      "step": 11235
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9883718376886304,
+      "learning_rate": 1.9633669341500492e-07,
+      "loss": 0.2626,
+      "step": 11236
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5499604013139279,
+      "learning_rate": 1.962546536942523e-07,
+      "loss": 0.3044,
+      "step": 11237
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9565536379392627,
+      "learning_rate": 1.9617262693188703e-07,
+      "loss": 0.0567,
+      "step": 11238
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.45903994597422537,
+      "learning_rate": 1.96090613131408e-07,
+      "loss": 0.0997,
+      "step": 11239
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9446472463424355,
+      "learning_rate": 1.9600861229631456e-07,
+      "loss": 0.2746,
+      "step": 11240
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0899076842399467,
+      "learning_rate": 1.9592662443010466e-07,
+      "loss": 0.1573,
+      "step": 11241
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.718774540486392,
+      "learning_rate": 1.9584464953627621e-07,
+      "loss": 0.135,
+      "step": 11242
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6652962312181652,
+      "learning_rate": 1.957626876183266e-07,
+      "loss": 0.1435,
+      "step": 11243
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.600922276086795,
+      "learning_rate": 1.9568073867975217e-07,
+      "loss": 0.2011,
+      "step": 11244
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7627624830365518,
+      "learning_rate": 1.9559880272404937e-07,
+      "loss": 0.1666,
+      "step": 11245
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34654085612832575,
+      "learning_rate": 1.9551687975471337e-07,
+      "loss": 0.0986,
+      "step": 11246
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.44899840641322625,
+      "learning_rate": 1.9543496977523954e-07,
+      "loss": 0.2565,
+      "step": 11247
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.913897145425443,
+      "learning_rate": 1.95353072789122e-07,
+      "loss": 0.0238,
+      "step": 11248
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5162736989873675,
+      "learning_rate": 1.9527118879985498e-07,
+      "loss": 0.0036,
+      "step": 11249
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4226692172058273,
+      "learning_rate": 1.9518931781093145e-07,
+      "loss": 0.2843,
+      "step": 11250
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7849040542404795,
+      "learning_rate": 1.9510745982584452e-07,
+      "loss": 0.1482,
+      "step": 11251
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9902365652357928,
+      "learning_rate": 1.9502561484808612e-07,
+      "loss": 0.0985,
+      "step": 11252
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4831877618226033,
+      "learning_rate": 1.9494378288114816e-07,
+      "loss": 0.2056,
+      "step": 11253
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2272332823805097,
+      "learning_rate": 1.948619639285217e-07,
+      "loss": 0.257,
+      "step": 11254
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3251243194056916,
+      "learning_rate": 1.9478015799369711e-07,
+      "loss": 0.2012,
+      "step": 11255
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0066410634636844,
+      "learning_rate": 1.9469836508016475e-07,
+      "loss": 0.242,
+      "step": 11256
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.44470279731483114,
+      "learning_rate": 1.9461658519141368e-07,
+      "loss": 0.1918,
+      "step": 11257
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9107649630220954,
+      "learning_rate": 1.9453481833093298e-07,
+      "loss": 0.2812,
+      "step": 11258
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8946927979470785,
+      "learning_rate": 1.94453064502211e-07,
+      "loss": 0.1912,
+      "step": 11259
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.46426163624716443,
+      "learning_rate": 1.9437132370873565e-07,
+      "loss": 0.0628,
+      "step": 11260
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6507929639380663,
+      "learning_rate": 1.9428959595399385e-07,
+      "loss": 0.2525,
+      "step": 11261
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8699678919806615,
+      "learning_rate": 1.9420788124147264e-07,
+      "loss": 0.1519,
+      "step": 11262
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8073902475761016,
+      "learning_rate": 1.9412617957465777e-07,
+      "loss": 0.0252,
+      "step": 11263
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34797149864818405,
+      "learning_rate": 1.9404449095703512e-07,
+      "loss": 0.1325,
+      "step": 11264
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8124671326892062,
+      "learning_rate": 1.9396281539208937e-07,
+      "loss": 0.2661,
+      "step": 11265
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.250268017497989,
+      "learning_rate": 1.9388115288330526e-07,
+      "loss": 0.0244,
+      "step": 11266
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6363677745593981,
+      "learning_rate": 1.9379950343416656e-07,
+      "loss": 0.1499,
+      "step": 11267
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.158412443101855,
+      "learning_rate": 1.9371786704815645e-07,
+      "loss": 0.3666,
+      "step": 11268
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3719723092275262,
+      "learning_rate": 1.93636243728758e-07,
+      "loss": 0.2845,
+      "step": 11269
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8101719245667807,
+      "learning_rate": 1.9355463347945305e-07,
+      "loss": 0.2269,
+      "step": 11270
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3636657007926116,
+      "learning_rate": 1.934730363037237e-07,
+      "loss": 0.2541,
+      "step": 11271
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5334193361930671,
+      "learning_rate": 1.933914522050506e-07,
+      "loss": 0.1261,
+      "step": 11272
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9624972864915775,
+      "learning_rate": 1.933098811869147e-07,
+      "loss": 0.2005,
+      "step": 11273
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4614518387255773,
+      "learning_rate": 1.9322832325279558e-07,
+      "loss": 0.1322,
+      "step": 11274
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.35987495361670174,
+      "learning_rate": 1.931467784061731e-07,
+      "loss": 0.0917,
+      "step": 11275
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3751024634717288,
+      "learning_rate": 1.930652466505257e-07,
+      "loss": 0.2347,
+      "step": 11276
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.791205175829184,
+      "learning_rate": 1.9298372798933193e-07,
+      "loss": 0.1651,
+      "step": 11277
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5227542350670976,
+      "learning_rate": 1.9290222242606946e-07,
+      "loss": 0.0922,
+      "step": 11278
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.39238810146444875,
+      "learning_rate": 1.9282072996421577e-07,
+      "loss": 0.1329,
+      "step": 11279
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8366118709173087,
+      "learning_rate": 1.92739250607247e-07,
+      "loss": 0.1964,
+      "step": 11280
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8545381031191396,
+      "learning_rate": 1.9265778435863967e-07,
+      "loss": 0.2031,
+      "step": 11281
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5449701366996338,
+      "learning_rate": 1.925763312218691e-07,
+      "loss": 0.1661,
+      "step": 11282
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3373603343141682,
+      "learning_rate": 1.9249489120041007e-07,
+      "loss": 0.4064,
+      "step": 11283
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7810633429509313,
+      "learning_rate": 1.924134642977373e-07,
+      "loss": 0.2826,
+      "step": 11284
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5539469380350881,
+      "learning_rate": 1.9233205051732431e-07,
+      "loss": 0.3932,
+      "step": 11285
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0544716698876138,
+      "learning_rate": 1.9225064986264473e-07,
+      "loss": 0.2043,
+      "step": 11286
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4614833600306727,
+      "learning_rate": 1.9216926233717084e-07,
+      "loss": 0.0585,
+      "step": 11287
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2431582573504985,
+      "learning_rate": 1.920878879443753e-07,
+      "loss": 0.3204,
+      "step": 11288
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7717165785479521,
+      "learning_rate": 1.9200652668772922e-07,
+      "loss": 0.4098,
+      "step": 11289
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.27975978791258,
+      "learning_rate": 1.9192517857070402e-07,
+      "loss": 0.3203,
+      "step": 11290
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5398482857016399,
+      "learning_rate": 1.9184384359676986e-07,
+      "loss": 0.1996,
+      "step": 11291
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8720094451731067,
+      "learning_rate": 1.9176252176939696e-07,
+      "loss": 0.3041,
+      "step": 11292
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.40526003036035,
+      "learning_rate": 1.916812130920543e-07,
+      "loss": 0.2855,
+      "step": 11293
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7492185619361499,
+      "learning_rate": 1.9159991756821097e-07,
+      "loss": 0.1638,
+      "step": 11294
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.8863006868943675,
+      "learning_rate": 1.9151863520133527e-07,
+      "loss": 0.0419,
+      "step": 11295
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2509224150024811,
+      "learning_rate": 1.914373659948945e-07,
+      "loss": 0.0916,
+      "step": 11296
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.734187662025367,
+      "learning_rate": 1.9135610995235618e-07,
+      "loss": 0.2635,
+      "step": 11297
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18010752757270784,
+      "learning_rate": 1.9127486707718648e-07,
+      "loss": 0.0547,
+      "step": 11298
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3652400883110703,
+      "learning_rate": 1.9119363737285177e-07,
+      "loss": 0.0909,
+      "step": 11299
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8036609510321918,
+      "learning_rate": 1.911124208428171e-07,
+      "loss": 0.1154,
+      "step": 11300
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9157834260756635,
+      "learning_rate": 1.9103121749054767e-07,
+      "loss": 0.1498,
+      "step": 11301
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.44321398656095967,
+      "learning_rate": 1.9095002731950738e-07,
+      "loss": 0.1158,
+      "step": 11302
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6170540323583107,
+      "learning_rate": 1.9086885033316042e-07,
+      "loss": 0.2886,
+      "step": 11303
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.156481405504817,
+      "learning_rate": 1.9078768653496957e-07,
+      "loss": 0.3127,
+      "step": 11304
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.371487411194727,
+      "learning_rate": 1.9070653592839774e-07,
+      "loss": 0.0578,
+      "step": 11305
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4820827101920844,
+      "learning_rate": 1.906253985169067e-07,
+      "loss": 0.2871,
+      "step": 11306
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.743525703499758,
+      "learning_rate": 1.9054427430395825e-07,
+      "loss": 0.2672,
+      "step": 11307
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3482562172310517,
+      "learning_rate": 1.904631632930131e-07,
+      "loss": 0.0837,
+      "step": 11308
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7090816574899531,
+      "learning_rate": 1.9038206548753156e-07,
+      "loss": 0.3691,
+      "step": 11309
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6656375800217357,
+      "learning_rate": 1.9030098089097345e-07,
+      "loss": 0.1905,
+      "step": 11310
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9876900600288355,
+      "learning_rate": 1.9021990950679821e-07,
+      "loss": 0.0927,
+      "step": 11311
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6965926457653281,
+      "learning_rate": 1.901388513384643e-07,
+      "loss": 0.1765,
+      "step": 11312
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.423903362526783,
+      "learning_rate": 1.900578063894298e-07,
+      "loss": 0.0063,
+      "step": 11313
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.227600379231426,
+      "learning_rate": 1.8997677466315253e-07,
+      "loss": 0.2589,
+      "step": 11314
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3112065002380018,
+      "learning_rate": 1.8989575616308916e-07,
+      "loss": 0.1935,
+      "step": 11315
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5071865985285988,
+      "learning_rate": 1.898147508926964e-07,
+      "loss": 0.5045,
+      "step": 11316
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6361604339442415,
+      "learning_rate": 1.8973375885542963e-07,
+      "loss": 0.2456,
+      "step": 11317
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.49806507043030224,
+      "learning_rate": 1.8965278005474473e-07,
+      "loss": 0.2286,
+      "step": 11318
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5761747681338725,
+      "learning_rate": 1.8957181449409582e-07,
+      "loss": 0.0065,
+      "step": 11319
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7775147011624235,
+      "learning_rate": 1.894908621769376e-07,
+      "loss": 0.0907,
+      "step": 11320
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3591939137242455,
+      "learning_rate": 1.8940992310672315e-07,
+      "loss": 0.2522,
+      "step": 11321
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.975254351357844,
+      "learning_rate": 1.89328997286906e-07,
+      "loss": 0.2037,
+      "step": 11322
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0850047961635327,
+      "learning_rate": 1.892480847209383e-07,
+      "loss": 0.183,
+      "step": 11323
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9839293972390819,
+      "learning_rate": 1.8916718541227185e-07,
+      "loss": 0.3225,
+      "step": 11324
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7497703514236261,
+      "learning_rate": 1.8908629936435827e-07,
+      "loss": 0.1687,
+      "step": 11325
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2926129797374673,
+      "learning_rate": 1.8900542658064805e-07,
+      "loss": 0.1619,
+      "step": 11326
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8220806442208703,
+      "learning_rate": 1.8892456706459163e-07,
+      "loss": 0.133,
+      "step": 11327
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0574883016902499,
+      "learning_rate": 1.8884372081963835e-07,
+      "loss": 0.0633,
+      "step": 11328
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.775531667402893,
+      "learning_rate": 1.8876288784923745e-07,
+      "loss": 0.0403,
+      "step": 11329
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.709690766797677,
+      "learning_rate": 1.8868206815683763e-07,
+      "loss": 0.0914,
+      "step": 11330
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9142820871214081,
+      "learning_rate": 1.8860126174588636e-07,
+      "loss": 0.2878,
+      "step": 11331
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.953645651570006,
+      "learning_rate": 1.8852046861983134e-07,
+      "loss": 0.0113,
+      "step": 11332
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5171814683896292,
+      "learning_rate": 1.884396887821194e-07,
+      "loss": 0.1535,
+      "step": 11333
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9780494805901289,
+      "learning_rate": 1.883589222361965e-07,
+      "loss": 0.3118,
+      "step": 11334
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4719540877982585,
+      "learning_rate": 1.8827816898550863e-07,
+      "loss": 0.1827,
+      "step": 11335
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5489758011674478,
+      "learning_rate": 1.8819742903350068e-07,
+      "loss": 0.1459,
+      "step": 11336
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3944660254622898,
+      "learning_rate": 1.8811670238361703e-07,
+      "loss": 0.1734,
+      "step": 11337
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.577398856012256,
+      "learning_rate": 1.8803598903930205e-07,
+      "loss": 0.0038,
+      "step": 11338
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.821320998314224,
+      "learning_rate": 1.8795528900399872e-07,
+      "loss": 0.2236,
+      "step": 11339
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7722991064448721,
+      "learning_rate": 1.878746022811502e-07,
+      "loss": 0.1255,
+      "step": 11340
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.36684766591212525,
+      "learning_rate": 1.8779392887419843e-07,
+      "loss": 0.1759,
+      "step": 11341
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3443206899330606,
+      "learning_rate": 1.8771326878658545e-07,
+      "loss": 0.374,
+      "step": 11342
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8821809580335769,
+      "learning_rate": 1.8763262202175202e-07,
+      "loss": 0.2174,
+      "step": 11343
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 18.54331626413916,
+      "learning_rate": 1.8755198858313903e-07,
+      "loss": 0.1073,
+      "step": 11344
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3406909347627889,
+      "learning_rate": 1.874713684741861e-07,
+      "loss": 0.1718,
+      "step": 11345
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7917407665217508,
+      "learning_rate": 1.8739076169833308e-07,
+      "loss": 0.1452,
+      "step": 11346
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7331813446289264,
+      "learning_rate": 1.8731016825901842e-07,
+      "loss": 0.2365,
+      "step": 11347
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.423763078498533,
+      "learning_rate": 1.8722958815968054e-07,
+      "loss": 0.1211,
+      "step": 11348
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.238856469415873,
+      "learning_rate": 1.871490214037572e-07,
+      "loss": 0.0339,
+      "step": 11349
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.38855925532029184,
+      "learning_rate": 1.8706846799468568e-07,
+      "loss": 0.0896,
+      "step": 11350
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.112503105090476,
+      "learning_rate": 1.8698792793590235e-07,
+      "loss": 0.1634,
+      "step": 11351
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6381406907226377,
+      "learning_rate": 1.8690740123084315e-07,
+      "loss": 0.0893,
+      "step": 11352
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9092041439344964,
+      "learning_rate": 1.868268878829437e-07,
+      "loss": 0.2248,
+      "step": 11353
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.6039574825094425,
+      "learning_rate": 1.8674638789563869e-07,
+      "loss": 0.1151,
+      "step": 11354
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7360579684385566,
+      "learning_rate": 1.866659012723626e-07,
+      "loss": 0.2566,
+      "step": 11355
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8850195778522192,
+      "learning_rate": 1.8658542801654887e-07,
+      "loss": 0.1602,
+      "step": 11356
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7711100564924428,
+      "learning_rate": 1.8650496813163096e-07,
+      "loss": 0.0459,
+      "step": 11357
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.383151805890166,
+      "learning_rate": 1.864245216210412e-07,
+      "loss": 0.0569,
+      "step": 11358
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6917029957197063,
+      "learning_rate": 1.8634408848821186e-07,
+      "loss": 0.0785,
+      "step": 11359
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8159387198461046,
+      "learning_rate": 1.8626366873657413e-07,
+      "loss": 0.1036,
+      "step": 11360
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.240459356947686,
+      "learning_rate": 1.8618326236955906e-07,
+      "loss": 0.1267,
+      "step": 11361
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.495084438111064,
+      "learning_rate": 1.8610286939059676e-07,
+      "loss": 0.1815,
+      "step": 11362
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.692074811260936,
+      "learning_rate": 1.860224898031172e-07,
+      "loss": 0.1091,
+      "step": 11363
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4801473968529875,
+      "learning_rate": 1.8594212361054922e-07,
+      "loss": 0.0339,
+      "step": 11364
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0618883813365898,
+      "learning_rate": 1.8586177081632158e-07,
+      "loss": 0.3165,
+      "step": 11365
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 21.182772692848125,
+      "learning_rate": 1.8578143142386248e-07,
+      "loss": 0.2846,
+      "step": 11366
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5496385039487668,
+      "learning_rate": 1.8570110543659907e-07,
+      "loss": 0.4394,
+      "step": 11367
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8117317447610425,
+      "learning_rate": 1.856207928579584e-07,
+      "loss": 0.1669,
+      "step": 11368
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6162402209721237,
+      "learning_rate": 1.8554049369136655e-07,
+      "loss": 0.0026,
+      "step": 11369
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7646794922934054,
+      "learning_rate": 1.8546020794024954e-07,
+      "loss": 0.1577,
+      "step": 11370
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.79595816006261,
+      "learning_rate": 1.853799356080322e-07,
+      "loss": 0.3337,
+      "step": 11371
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1560967344600477,
+      "learning_rate": 1.8529967669813945e-07,
+      "loss": 0.2162,
+      "step": 11372
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9609528498824436,
+      "learning_rate": 1.8521943121399497e-07,
+      "loss": 0.3063,
+      "step": 11373
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7273985933698603,
+      "learning_rate": 1.8513919915902248e-07,
+      "loss": 0.2677,
+      "step": 11374
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8282398453406957,
+      "learning_rate": 1.8505898053664455e-07,
+      "loss": 0.28,
+      "step": 11375
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.672463712150755,
+      "learning_rate": 1.849787753502838e-07,
+      "loss": 0.1784,
+      "step": 11376
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5570384879018646,
+      "learning_rate": 1.848985836033617e-07,
+      "loss": 0.1562,
+      "step": 11377
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.666323930178118,
+      "learning_rate": 1.8481840529929938e-07,
+      "loss": 0.223,
+      "step": 11378
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0651202670771267,
+      "learning_rate": 1.8473824044151758e-07,
+      "loss": 0.0999,
+      "step": 11379
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.857961344166354,
+      "learning_rate": 1.8465808903343606e-07,
+      "loss": 0.1848,
+      "step": 11380
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3676116604325516,
+      "learning_rate": 1.8457795107847435e-07,
+      "loss": 0.0898,
+      "step": 11381
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.171026555137782,
+      "learning_rate": 1.8449782658005152e-07,
+      "loss": 0.3671,
+      "step": 11382
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5106940084247752,
+      "learning_rate": 1.8441771554158554e-07,
+      "loss": 0.1456,
+      "step": 11383
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7994769509285218,
+      "learning_rate": 1.843376179664941e-07,
+      "loss": 0.2311,
+      "step": 11384
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5945871463968089,
+      "learning_rate": 1.842575338581947e-07,
+      "loss": 0.2475,
+      "step": 11385
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3367533168219845,
+      "learning_rate": 1.8417746322010342e-07,
+      "loss": 0.2572,
+      "step": 11386
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.212180939838843,
+      "learning_rate": 1.8409740605563662e-07,
+      "loss": 0.0048,
+      "step": 11387
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.529879221363101,
+      "learning_rate": 1.8401736236820931e-07,
+      "loss": 0.0993,
+      "step": 11388
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5938335393745138,
+      "learning_rate": 1.8393733216123675e-07,
+      "loss": 0.2476,
+      "step": 11389
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2951656965333097,
+      "learning_rate": 1.8385731543813278e-07,
+      "loss": 0.0676,
+      "step": 11390
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1143514561504615,
+      "learning_rate": 1.837773122023114e-07,
+      "loss": 0.1534,
+      "step": 11391
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3308697602123853,
+      "learning_rate": 1.8369732245718562e-07,
+      "loss": 0.1973,
+      "step": 11392
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.354457923485755,
+      "learning_rate": 1.836173462061677e-07,
+      "loss": 0.0941,
+      "step": 11393
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6274520460016499,
+      "learning_rate": 1.8353738345267e-07,
+      "loss": 0.1106,
+      "step": 11394
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5221636998211159,
+      "learning_rate": 1.834574342001035e-07,
+      "loss": 0.3306,
+      "step": 11395
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.324097496094751,
+      "learning_rate": 1.8337749845187934e-07,
+      "loss": 0.0962,
+      "step": 11396
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.675354180486095,
+      "learning_rate": 1.8329757621140746e-07,
+      "loss": 0.0131,
+      "step": 11397
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5524860551788853,
+      "learning_rate": 1.8321766748209778e-07,
+      "loss": 0.1393,
+      "step": 11398
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5590398221356457,
+      "learning_rate": 1.8313777226735904e-07,
+      "loss": 0.1419,
+      "step": 11399
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5652153307301052,
+      "learning_rate": 1.8305789057059995e-07,
+      "loss": 0.1761,
+      "step": 11400
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1758393193850527,
+      "learning_rate": 1.8297802239522847e-07,
+      "loss": 0.133,
+      "step": 11401
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7437972372207307,
+      "learning_rate": 1.8289816774465178e-07,
+      "loss": 0.1527,
+      "step": 11402
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3463829145471895,
+      "learning_rate": 1.8281832662227665e-07,
+      "loss": 0.6341,
+      "step": 11403
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2889171271947935,
+      "learning_rate": 1.8273849903150946e-07,
+      "loss": 0.0184,
+      "step": 11404
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3106307738070322,
+      "learning_rate": 1.8265868497575576e-07,
+      "loss": 0.0877,
+      "step": 11405
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5275145363797861,
+      "learning_rate": 1.8257888445842023e-07,
+      "loss": 0.2441,
+      "step": 11406
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.11567962623004249,
+      "learning_rate": 1.824990974829078e-07,
+      "loss": 0.0019,
+      "step": 11407
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3797639917060166,
+      "learning_rate": 1.82419324052622e-07,
+      "loss": 0.1215,
+      "step": 11408
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.989969077827051,
+      "learning_rate": 1.823395641709664e-07,
+      "loss": 0.3756,
+      "step": 11409
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8942893852458406,
+      "learning_rate": 1.8225981784134336e-07,
+      "loss": 0.0271,
+      "step": 11410
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9581676941039796,
+      "learning_rate": 1.8218008506715544e-07,
+      "loss": 0.0194,
+      "step": 11411
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.600710384600168,
+      "learning_rate": 1.8210036585180383e-07,
+      "loss": 0.0841,
+      "step": 11412
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.112113519964818,
+      "learning_rate": 1.8202066019868979e-07,
+      "loss": 0.4596,
+      "step": 11413
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0098642959598396,
+      "learning_rate": 1.8194096811121346e-07,
+      "loss": 0.1137,
+      "step": 11414
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.250568542399658,
+      "learning_rate": 1.8186128959277497e-07,
+      "loss": 0.0919,
+      "step": 11415
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9247788958931612,
+      "learning_rate": 1.8178162464677328e-07,
+      "loss": 0.2528,
+      "step": 11416
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4988442802089735,
+      "learning_rate": 1.817019732766073e-07,
+      "loss": 0.0702,
+      "step": 11417
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0537970175279456,
+      "learning_rate": 1.8162233548567486e-07,
+      "loss": 0.3904,
+      "step": 11418
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4204321740646035,
+      "learning_rate": 1.8154271127737357e-07,
+      "loss": 0.0892,
+      "step": 11419
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7301438334478592,
+      "learning_rate": 1.814631006551006e-07,
+      "loss": 0.1586,
+      "step": 11420
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.607254441125836,
+      "learning_rate": 1.813835036222519e-07,
+      "loss": 0.2037,
+      "step": 11421
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4555645094974445,
+      "learning_rate": 1.8130392018222362e-07,
+      "loss": 0.1012,
+      "step": 11422
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5832270929414122,
+      "learning_rate": 1.812243503384106e-07,
+      "loss": 0.2479,
+      "step": 11423
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9202720334256785,
+      "learning_rate": 1.811447940942078e-07,
+      "loss": 0.2745,
+      "step": 11424
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8422475057212134,
+      "learning_rate": 1.810652514530089e-07,
+      "loss": 0.1688,
+      "step": 11425
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.267528395624066,
+      "learning_rate": 1.8098572241820764e-07,
+      "loss": 0.1757,
+      "step": 11426
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.42829894387744827,
+      "learning_rate": 1.8090620699319658e-07,
+      "loss": 0.1465,
+      "step": 11427
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3229782021811989,
+      "learning_rate": 1.8082670518136839e-07,
+      "loss": 0.1382,
+      "step": 11428
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.010708927575749,
+      "learning_rate": 1.807472169861144e-07,
+      "loss": 0.4057,
+      "step": 11429
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.46407104850424413,
+      "learning_rate": 1.806677424108261e-07,
+      "loss": 0.1411,
+      "step": 11430
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3753019193240372,
+      "learning_rate": 1.8058828145889367e-07,
+      "loss": 0.2436,
+      "step": 11431
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.601650084403245,
+      "learning_rate": 1.8050883413370738e-07,
+      "loss": 0.0615,
+      "step": 11432
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.10839587137874689,
+      "learning_rate": 1.8042940043865655e-07,
+      "loss": 0.0044,
+      "step": 11433
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.94888379640534,
+      "learning_rate": 1.8034998037712967e-07,
+      "loss": 0.2202,
+      "step": 11434
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0849138426443623,
+      "learning_rate": 1.8027057395251528e-07,
+      "loss": 0.4132,
+      "step": 11435
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4985725447357996,
+      "learning_rate": 1.8019118116820091e-07,
+      "loss": 0.1153,
+      "step": 11436
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0127380697798904,
+      "learning_rate": 1.8011180202757382e-07,
+      "loss": 0.0446,
+      "step": 11437
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.21527906885469447,
+      "learning_rate": 1.8003243653402013e-07,
+      "loss": 0.1046,
+      "step": 11438
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5323627967733878,
+      "learning_rate": 1.7995308469092608e-07,
+      "loss": 0.0247,
+      "step": 11439
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1324239374290164,
+      "learning_rate": 1.7987374650167664e-07,
+      "loss": 0.1062,
+      "step": 11440
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2843530420456393,
+      "learning_rate": 1.797944219696569e-07,
+      "loss": 0.0342,
+      "step": 11441
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1114536875301562,
+      "learning_rate": 1.797151110982506e-07,
+      "loss": 0.2956,
+      "step": 11442
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5086169328961192,
+      "learning_rate": 1.7963581389084175e-07,
+      "loss": 0.1714,
+      "step": 11443
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.919644543589692,
+      "learning_rate": 1.7955653035081287e-07,
+      "loss": 0.249,
+      "step": 11444
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.42747007523985503,
+      "learning_rate": 1.7947726048154676e-07,
+      "loss": 0.0351,
+      "step": 11445
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7128928853881998,
+      "learning_rate": 1.793980042864251e-07,
+      "loss": 0.2259,
+      "step": 11446
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.726557080274698,
+      "learning_rate": 1.7931876176882883e-07,
+      "loss": 0.1929,
+      "step": 11447
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1988796467009433,
+      "learning_rate": 1.79239532932139e-07,
+      "loss": 0.255,
+      "step": 11448
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0306298807776058,
+      "learning_rate": 1.791603177797354e-07,
+      "loss": 0.3409,
+      "step": 11449
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8206713277340335,
+      "learning_rate": 1.7908111631499772e-07,
+      "loss": 0.2763,
+      "step": 11450
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.906822347935492,
+      "learning_rate": 1.7900192854130464e-07,
+      "loss": 0.1326,
+      "step": 11451
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7884778066634541,
+      "learning_rate": 1.789227544620347e-07,
+      "loss": 0.1866,
+      "step": 11452
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4145810436910572,
+      "learning_rate": 1.7884359408056532e-07,
+      "loss": 0.031,
+      "step": 11453
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6210063643046614,
+      "learning_rate": 1.7876444740027384e-07,
+      "loss": 0.1273,
+      "step": 11454
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9570963601069166,
+      "learning_rate": 1.7868531442453677e-07,
+      "loss": 0.23,
+      "step": 11455
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9909922353503083,
+      "learning_rate": 1.7860619515673032e-07,
+      "loss": 0.3218,
+      "step": 11456
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.436526150111342,
+      "learning_rate": 1.785270896002295e-07,
+      "loss": 0.0065,
+      "step": 11457
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4145354786040616,
+      "learning_rate": 1.784479977584094e-07,
+      "loss": 0.1989,
+      "step": 11458
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0745897571282355,
+      "learning_rate": 1.78368919634644e-07,
+      "loss": 0.1913,
+      "step": 11459
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6166655301826944,
+      "learning_rate": 1.7828985523230722e-07,
+      "loss": 0.3184,
+      "step": 11460
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5085237659998391,
+      "learning_rate": 1.782108045547719e-07,
+      "loss": 0.0833,
+      "step": 11461
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3943343845860209,
+      "learning_rate": 1.7813176760541037e-07,
+      "loss": 0.2125,
+      "step": 11462
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.625016401705777,
+      "learning_rate": 1.7805274438759482e-07,
+      "loss": 0.0823,
+      "step": 11463
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.275557332623727,
+      "learning_rate": 1.7797373490469624e-07,
+      "loss": 0.2726,
+      "step": 11464
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0708090834780841,
+      "learning_rate": 1.7789473916008568e-07,
+      "loss": 0.3473,
+      "step": 11465
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.224580636042428,
+      "learning_rate": 1.7781575715713287e-07,
+      "loss": 0.2007,
+      "step": 11466
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8447087835760301,
+      "learning_rate": 1.7773678889920774e-07,
+      "loss": 0.0485,
+      "step": 11467
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.765010029835095,
+      "learning_rate": 1.776578343896788e-07,
+      "loss": 0.0837,
+      "step": 11468
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4329833036547865,
+      "learning_rate": 1.775788936319148e-07,
+      "loss": 0.1843,
+      "step": 11469
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5821034965963179,
+      "learning_rate": 1.7749996662928317e-07,
+      "loss": 0.1486,
+      "step": 11470
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.621258290149381,
+      "learning_rate": 1.7742105338515135e-07,
+      "loss": 0.3242,
+      "step": 11471
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.129873094277355,
+      "learning_rate": 1.7734215390288598e-07,
+      "loss": 0.2612,
+      "step": 11472
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.49494560339519383,
+      "learning_rate": 1.7726326818585275e-07,
+      "loss": 0.0985,
+      "step": 11473
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.08218488388321017,
+      "learning_rate": 1.771843962374175e-07,
+      "loss": 0.0018,
+      "step": 11474
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6028904083027404,
+      "learning_rate": 1.7710553806094465e-07,
+      "loss": 0.2327,
+      "step": 11475
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.263165722367507,
+      "learning_rate": 1.770266936597988e-07,
+      "loss": 0.2769,
+      "step": 11476
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.504074391140485,
+      "learning_rate": 1.7694786303734326e-07,
+      "loss": 0.2694,
+      "step": 11477
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7620545644939543,
+      "learning_rate": 1.7686904619694154e-07,
+      "loss": 0.0712,
+      "step": 11478
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7622744474898832,
+      "learning_rate": 1.7679024314195567e-07,
+      "loss": 0.2432,
+      "step": 11479
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9690519051497881,
+      "learning_rate": 1.7671145387574798e-07,
+      "loss": 0.4384,
+      "step": 11480
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6862667021500114,
+      "learning_rate": 1.7663267840167934e-07,
+      "loss": 0.3638,
+      "step": 11481
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.46122680833365964,
+      "learning_rate": 1.7655391672311088e-07,
+      "loss": 0.1538,
+      "step": 11482
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.401978850357725,
+      "learning_rate": 1.7647516884340235e-07,
+      "loss": 0.2038,
+      "step": 11483
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1790502517707164,
+      "learning_rate": 1.763964347659137e-07,
+      "loss": 0.267,
+      "step": 11484
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9168656808320996,
+      "learning_rate": 1.7631771449400346e-07,
+      "loss": 0.2097,
+      "step": 11485
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5945793788160334,
+      "learning_rate": 1.7623900803103036e-07,
+      "loss": 0.2263,
+      "step": 11486
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7329442615887327,
+      "learning_rate": 1.7616031538035187e-07,
+      "loss": 0.1219,
+      "step": 11487
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7730545468778828,
+      "learning_rate": 1.7608163654532548e-07,
+      "loss": 0.3914,
+      "step": 11488
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5839277464470503,
+      "learning_rate": 1.760029715293075e-07,
+      "loss": 0.0079,
+      "step": 11489
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5553975994262143,
+      "learning_rate": 1.7592432033565407e-07,
+      "loss": 0.2914,
+      "step": 11490
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.576759306034706,
+      "learning_rate": 1.7584568296772073e-07,
+      "loss": 0.0139,
+      "step": 11491
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7765750848745309,
+      "learning_rate": 1.7576705942886206e-07,
+      "loss": 0.1938,
+      "step": 11492
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7595096649079054,
+      "learning_rate": 1.7568844972243257e-07,
+      "loss": 0.0817,
+      "step": 11493
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6963008825521375,
+      "learning_rate": 1.7560985385178561e-07,
+      "loss": 0.1989,
+      "step": 11494
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.610266027254652,
+      "learning_rate": 1.7553127182027456e-07,
+      "loss": 0.2918,
+      "step": 11495
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4262986009962106,
+      "learning_rate": 1.7545270363125153e-07,
+      "loss": 0.1154,
+      "step": 11496
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1687380590174075,
+      "learning_rate": 1.7537414928806876e-07,
+      "loss": 0.299,
+      "step": 11497
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3245236831780917,
+      "learning_rate": 1.7529560879407718e-07,
+      "loss": 0.1032,
+      "step": 11498
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3085804959006284,
+      "learning_rate": 1.7521708215262788e-07,
+      "loss": 0.0966,
+      "step": 11499
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5337805967239032,
+      "learning_rate": 1.751385693670705e-07,
+      "loss": 0.0953,
+      "step": 11500
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 15681,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "total_flos": 1464249587613696.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}