diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,18963 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9987519068090416,
+  "eval_steps": 500,
+  "global_step": 2703,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001109416169740674,
+      "grad_norm": 0.06963343173265457,
+      "learning_rate": 1.107011070110701e-06,
+      "loss": 0.5928,
+      "step": 1
+    },
+    {
+      "epoch": 0.002218832339481348,
+      "grad_norm": 0.08431016653776169,
+      "learning_rate": 2.214022140221402e-06,
+      "loss": 0.7915,
+      "step": 2
+    },
+    {
+      "epoch": 0.003328248509222022,
+      "grad_norm": 0.06818119436502457,
+      "learning_rate": 3.321033210332103e-06,
+      "loss": 0.6594,
+      "step": 3
+    },
+    {
+      "epoch": 0.004437664678962696,
+      "grad_norm": 0.08563707023859024,
+      "learning_rate": 4.428044280442804e-06,
+      "loss": 0.5879,
+      "step": 4
+    },
+    {
+      "epoch": 0.00554708084870337,
+      "grad_norm": 0.053242169320583344,
+      "learning_rate": 5.535055350553505e-06,
+      "loss": 0.6156,
+      "step": 5
+    },
+    {
+      "epoch": 0.006656497018444044,
+      "grad_norm": 0.06964253634214401,
+      "learning_rate": 6.642066420664206e-06,
+      "loss": 0.7194,
+      "step": 6
+    },
+    {
+      "epoch": 0.007765913188184718,
+      "grad_norm": 0.07358719408512115,
+      "learning_rate": 7.749077490774907e-06,
+      "loss": 0.6253,
+      "step": 7
+    },
+    {
+      "epoch": 0.008875329357925392,
+      "grad_norm": 0.04413611814379692,
+      "learning_rate": 8.856088560885607e-06,
+      "loss": 0.5087,
+      "step": 8
+    },
+    {
+      "epoch": 0.009984745527666065,
+      "grad_norm": 0.06562013179063797,
+      "learning_rate": 9.96309963099631e-06,
+      "loss": 0.7106,
+      "step": 9
+    },
+    {
+      "epoch": 0.01109416169740674,
+      "grad_norm": 0.06969469785690308,
+      "learning_rate": 1.107011070110701e-05,
+      "loss": 0.5231,
+      "step": 10
+    },
+    {
+      "epoch": 0.012203577867147413,
+      "grad_norm": 0.06136368215084076,
+      "learning_rate": 1.2177121771217711e-05,
+      "loss": 0.5043,
+      "step": 11
+    },
+    {
+      "epoch": 0.013312994036888088,
+      "grad_norm": 0.22236867249011993,
+      "learning_rate": 1.3284132841328412e-05,
+      "loss": 1.1332,
+      "step": 12
+    },
+    {
+      "epoch": 0.014422410206628761,
+      "grad_norm": 0.05163656920194626,
+      "learning_rate": 1.4391143911439114e-05,
+      "loss": 0.6196,
+      "step": 13
+    },
+    {
+      "epoch": 0.015531826376369436,
+      "grad_norm": 0.07425156235694885,
+      "learning_rate": 1.5498154981549814e-05,
+      "loss": 0.6824,
+      "step": 14
+    },
+    {
+      "epoch": 0.01664124254611011,
+      "grad_norm": 0.10796231776475906,
+      "learning_rate": 1.6605166051660514e-05,
+      "loss": 0.6668,
+      "step": 15
+    },
+    {
+      "epoch": 0.017750658715850784,
+      "grad_norm": 0.08407709002494812,
+      "learning_rate": 1.7712177121771215e-05,
+      "loss": 0.6919,
+      "step": 16
+    },
+    {
+      "epoch": 0.01886007488559146,
+      "grad_norm": 0.09712005406618118,
+      "learning_rate": 1.8819188191881916e-05,
+      "loss": 0.6337,
+      "step": 17
+    },
+    {
+      "epoch": 0.01996949105533213,
+      "grad_norm": 0.09893519431352615,
+      "learning_rate": 1.992619926199262e-05,
+      "loss": 0.5787,
+      "step": 18
+    },
+    {
+      "epoch": 0.021078907225072805,
+      "grad_norm": 0.06742904335260391,
+      "learning_rate": 2.1033210332103317e-05,
+      "loss": 0.5236,
+      "step": 19
+    },
+    {
+      "epoch": 0.02218832339481348,
+      "grad_norm": 0.09684479981660843,
+      "learning_rate": 2.214022140221402e-05,
+      "loss": 0.8853,
+      "step": 20
+    },
+    {
+      "epoch": 0.023297739564554155,
+      "grad_norm": 0.07542835175991058,
+      "learning_rate": 2.3247232472324722e-05,
+      "loss": 0.5122,
+      "step": 21
+    },
+    {
+      "epoch": 0.024407155734294826,
+      "grad_norm": 0.10576613247394562,
+      "learning_rate": 2.4354243542435423e-05,
+      "loss": 0.5764,
+      "step": 22
+    },
+    {
+      "epoch": 0.0255165719040355,
+      "grad_norm": 0.09776762127876282,
+      "learning_rate": 2.5461254612546123e-05,
+      "loss": 0.5908,
+      "step": 23
+    },
+    {
+      "epoch": 0.026625988073776176,
+      "grad_norm": 0.11171472072601318,
+      "learning_rate": 2.6568265682656824e-05,
+      "loss": 0.6293,
+      "step": 24
+    },
+    {
+      "epoch": 0.02773540424351685,
+      "grad_norm": 0.07923085242509842,
+      "learning_rate": 2.7675276752767525e-05,
+      "loss": 0.5343,
+      "step": 25
+    },
+    {
+      "epoch": 0.028844820413257522,
+      "grad_norm": 0.13260161876678467,
+      "learning_rate": 2.878228782287823e-05,
+      "loss": 0.6485,
+      "step": 26
+    },
+    {
+      "epoch": 0.029954236582998197,
+      "grad_norm": 0.08974039554595947,
+      "learning_rate": 2.9889298892988926e-05,
+      "loss": 0.5814,
+      "step": 27
+    },
+    {
+      "epoch": 0.031063652752738872,
+      "grad_norm": 0.08319084346294403,
+      "learning_rate": 3.099630996309963e-05,
+      "loss": 0.5042,
+      "step": 28
+    },
+    {
+      "epoch": 0.03217306892247954,
+      "grad_norm": 0.21256183087825775,
+      "learning_rate": 3.2103321033210324e-05,
+      "loss": 0.7798,
+      "step": 29
+    },
+    {
+      "epoch": 0.03328248509222022,
+      "grad_norm": 0.15112732350826263,
+      "learning_rate": 3.321033210332103e-05,
+      "loss": 0.6845,
+      "step": 30
+    },
+    {
+      "epoch": 0.03439190126196089,
+      "grad_norm": 0.09808047860860825,
+      "learning_rate": 3.431734317343173e-05,
+      "loss": 0.631,
+      "step": 31
+    },
+    {
+      "epoch": 0.03550131743170157,
+      "grad_norm": 0.12207363545894623,
+      "learning_rate": 3.542435424354243e-05,
+      "loss": 0.6708,
+      "step": 32
+    },
+    {
+      "epoch": 0.03661073360144224,
+      "grad_norm": 0.11131646484136581,
+      "learning_rate": 3.6531365313653134e-05,
+      "loss": 0.4812,
+      "step": 33
+    },
+    {
+      "epoch": 0.03772014977118292,
+      "grad_norm": 0.15473362803459167,
+      "learning_rate": 3.763837638376383e-05,
+      "loss": 0.5202,
+      "step": 34
+    },
+    {
+      "epoch": 0.03882956594092359,
+      "grad_norm": 0.1214248314499855,
+      "learning_rate": 3.8745387453874535e-05,
+      "loss": 0.5629,
+      "step": 35
+    },
+    {
+      "epoch": 0.03993898211066426,
+      "grad_norm": 0.12858295440673828,
+      "learning_rate": 3.985239852398524e-05,
+      "loss": 0.6535,
+      "step": 36
+    },
+    {
+      "epoch": 0.041048398280404935,
+      "grad_norm": 0.23708263039588928,
+      "learning_rate": 4.0959409594095944e-05,
+      "loss": 0.7139,
+      "step": 37
+    },
+    {
+      "epoch": 0.04215781445014561,
+      "grad_norm": 0.11044642329216003,
+      "learning_rate": 4.2066420664206634e-05,
+      "loss": 0.5751,
+      "step": 38
+    },
+    {
+      "epoch": 0.043267230619886285,
+      "grad_norm": 0.23254919052124023,
+      "learning_rate": 4.317343173431734e-05,
+      "loss": 0.4629,
+      "step": 39
+    },
+    {
+      "epoch": 0.04437664678962696,
+      "grad_norm": 0.16974490880966187,
+      "learning_rate": 4.428044280442804e-05,
+      "loss": 0.7469,
+      "step": 40
+    },
+    {
+      "epoch": 0.045486062959367635,
+      "grad_norm": 0.1873021423816681,
+      "learning_rate": 4.538745387453874e-05,
+      "loss": 0.6589,
+      "step": 41
+    },
+    {
+      "epoch": 0.04659547912910831,
+      "grad_norm": 0.17700991034507751,
+      "learning_rate": 4.6494464944649444e-05,
+      "loss": 0.7062,
+      "step": 42
+    },
+    {
+      "epoch": 0.04770489529884898,
+      "grad_norm": 0.20159931480884552,
+      "learning_rate": 4.760147601476014e-05,
+      "loss": 0.6955,
+      "step": 43
+    },
+    {
+      "epoch": 0.04881431146858965,
+      "grad_norm": 0.18304705619812012,
+      "learning_rate": 4.8708487084870845e-05,
+      "loss": 0.6543,
+      "step": 44
+    },
+    {
+      "epoch": 0.04992372763833033,
+      "grad_norm": 0.1452740579843521,
+      "learning_rate": 4.981549815498154e-05,
+      "loss": 0.5875,
+      "step": 45
+    },
+    {
+      "epoch": 0.051033143808071,
+      "grad_norm": 0.16509367525577545,
+      "learning_rate": 5.092250922509225e-05,
+      "loss": 0.4751,
+      "step": 46
+    },
+    {
+      "epoch": 0.05214255997781168,
+      "grad_norm": 0.19247905910015106,
+      "learning_rate": 5.202952029520295e-05,
+      "loss": 0.4989,
+      "step": 47
+    },
+    {
+      "epoch": 0.05325197614755235,
+      "grad_norm": 0.20170418918132782,
+      "learning_rate": 5.313653136531365e-05,
+      "loss": 0.6385,
+      "step": 48
+    },
+    {
+      "epoch": 0.05436139231729303,
+      "grad_norm": 0.21076518297195435,
+      "learning_rate": 5.4243542435424346e-05,
+      "loss": 0.4535,
+      "step": 49
+    },
+    {
+      "epoch": 0.0554708084870337,
+      "grad_norm": 0.16010227799415588,
+      "learning_rate": 5.535055350553505e-05,
+      "loss": 0.5345,
+      "step": 50
+    },
+    {
+      "epoch": 0.05658022465677437,
+      "grad_norm": 0.21507202088832855,
+      "learning_rate": 5.6457564575645754e-05,
+      "loss": 0.5626,
+      "step": 51
+    },
+    {
+      "epoch": 0.057689640826515044,
+      "grad_norm": 0.2764558494091034,
+      "learning_rate": 5.756457564575646e-05,
+      "loss": 0.7726,
+      "step": 52
+    },
+    {
+      "epoch": 0.05879905699625572,
+      "grad_norm": 0.19346186518669128,
+      "learning_rate": 5.867158671586715e-05,
+      "loss": 0.6975,
+      "step": 53
+    },
+    {
+      "epoch": 0.059908473165996394,
+      "grad_norm": 0.15796959400177002,
+      "learning_rate": 5.977859778597785e-05,
+      "loss": 0.5838,
+      "step": 54
+    },
+    {
+      "epoch": 0.06101788933573707,
+      "grad_norm": 0.26421332359313965,
+      "learning_rate": 6.088560885608856e-05,
+      "loss": 0.4547,
+      "step": 55
+    },
+    {
+      "epoch": 0.062127305505477744,
+      "grad_norm": 0.15716983377933502,
+      "learning_rate": 6.199261992619925e-05,
+      "loss": 0.5811,
+      "step": 56
+    },
+    {
+      "epoch": 0.06323672167521842,
+      "grad_norm": 0.31450599431991577,
+      "learning_rate": 6.309963099630996e-05,
+      "loss": 0.6177,
+      "step": 57
+    },
+    {
+      "epoch": 0.06434613784495909,
+      "grad_norm": 0.1722506731748581,
+      "learning_rate": 6.420664206642065e-05,
+      "loss": 0.4451,
+      "step": 58
+    },
+    {
+      "epoch": 0.06545555401469977,
+      "grad_norm": 0.19947190582752228,
+      "learning_rate": 6.531365313653135e-05,
+      "loss": 0.6172,
+      "step": 59
+    },
+    {
+      "epoch": 0.06656497018444044,
+      "grad_norm": 0.24372801184654236,
+      "learning_rate": 6.642066420664206e-05,
+      "loss": 0.7055,
+      "step": 60
+    },
+    {
+      "epoch": 0.06767438635418112,
+      "grad_norm": 0.2935488224029541,
+      "learning_rate": 6.752767527675276e-05,
+      "loss": 0.6359,
+      "step": 61
+    },
+    {
+      "epoch": 0.06878380252392179,
+      "grad_norm": 0.30399176478385925,
+      "learning_rate": 6.863468634686347e-05,
+      "loss": 0.6141,
+      "step": 62
+    },
+    {
+      "epoch": 0.06989321869366245,
+      "grad_norm": 0.2780139446258545,
+      "learning_rate": 6.974169741697416e-05,
+      "loss": 0.6173,
+      "step": 63
+    },
+    {
+      "epoch": 0.07100263486340314,
+      "grad_norm": 0.22119928896427155,
+      "learning_rate": 7.084870848708486e-05,
+      "loss": 0.6311,
+      "step": 64
+    },
+    {
+      "epoch": 0.0721120510331438,
+      "grad_norm": 0.17710696160793304,
+      "learning_rate": 7.195571955719556e-05,
+      "loss": 0.5011,
+      "step": 65
+    },
+    {
+      "epoch": 0.07322146720288449,
+      "grad_norm": 0.24316059052944183,
+      "learning_rate": 7.306273062730627e-05,
+      "loss": 0.5671,
+      "step": 66
+    },
+    {
+      "epoch": 0.07433088337262515,
+      "grad_norm": 0.2396726906299591,
+      "learning_rate": 7.416974169741697e-05,
+      "loss": 0.5996,
+      "step": 67
+    },
+    {
+      "epoch": 0.07544029954236584,
+      "grad_norm": 0.25187021493911743,
+      "learning_rate": 7.527675276752766e-05,
+      "loss": 0.4867,
+      "step": 68
+    },
+    {
+      "epoch": 0.0765497157121065,
+      "grad_norm": 0.18911105394363403,
+      "learning_rate": 7.638376383763837e-05,
+      "loss": 0.5302,
+      "step": 69
+    },
+    {
+      "epoch": 0.07765913188184718,
+      "grad_norm": 0.2970142364501953,
+      "learning_rate": 7.749077490774907e-05,
+      "loss": 0.6817,
+      "step": 70
+    },
+    {
+      "epoch": 0.07876854805158785,
+      "grad_norm": 0.255367249250412,
+      "learning_rate": 7.859778597785978e-05,
+      "loss": 0.53,
+      "step": 71
+    },
+    {
+      "epoch": 0.07987796422132852,
+      "grad_norm": 0.2670186758041382,
+      "learning_rate": 7.970479704797048e-05,
+      "loss": 0.6765,
+      "step": 72
+    },
+    {
+      "epoch": 0.0809873803910692,
+      "grad_norm": 0.29176634550094604,
+      "learning_rate": 8.081180811808118e-05,
+      "loss": 0.6788,
+      "step": 73
+    },
+    {
+      "epoch": 0.08209679656080987,
+      "grad_norm": 0.1941952258348465,
+      "learning_rate": 8.191881918819189e-05,
+      "loss": 0.5154,
+      "step": 74
+    },
+    {
+      "epoch": 0.08320621273055055,
+      "grad_norm": 0.2700209617614746,
+      "learning_rate": 8.302583025830258e-05,
+      "loss": 0.6123,
+      "step": 75
+    },
+    {
+      "epoch": 0.08431562890029122,
+      "grad_norm": 0.2537286579608917,
+      "learning_rate": 8.413284132841327e-05,
+      "loss": 0.6377,
+      "step": 76
+    },
+    {
+      "epoch": 0.0854250450700319,
+      "grad_norm": 0.3083483874797821,
+      "learning_rate": 8.523985239852397e-05,
+      "loss": 0.6263,
+      "step": 77
+    },
+    {
+      "epoch": 0.08653446123977257,
+      "grad_norm": 0.1838807612657547,
+      "learning_rate": 8.634686346863468e-05,
+      "loss": 0.4246,
+      "step": 78
+    },
+    {
+      "epoch": 0.08764387740951324,
+      "grad_norm": 0.19544388353824615,
+      "learning_rate": 8.745387453874538e-05,
+      "loss": 0.6333,
+      "step": 79
+    },
+    {
+      "epoch": 0.08875329357925392,
+      "grad_norm": 0.11637648195028305,
+      "learning_rate": 8.856088560885608e-05,
+      "loss": 0.3302,
+      "step": 80
+    },
+    {
+      "epoch": 0.08986270974899459,
+      "grad_norm": 0.22084759175777435,
+      "learning_rate": 8.966789667896679e-05,
+      "loss": 0.5961,
+      "step": 81
+    },
+    {
+      "epoch": 0.09097212591873527,
+      "grad_norm": 0.20843982696533203,
+      "learning_rate": 9.077490774907748e-05,
+      "loss": 0.4923,
+      "step": 82
+    },
+    {
+      "epoch": 0.09208154208847594,
+      "grad_norm": 0.18717604875564575,
+      "learning_rate": 9.188191881918818e-05,
+      "loss": 0.7945,
+      "step": 83
+    },
+    {
+      "epoch": 0.09319095825821662,
+      "grad_norm": 0.5555791854858398,
+      "learning_rate": 9.298892988929889e-05,
+      "loss": 0.7418,
+      "step": 84
+    },
+    {
+      "epoch": 0.09430037442795729,
+      "grad_norm": 0.35459521412849426,
+      "learning_rate": 9.409594095940959e-05,
+      "loss": 0.4774,
+      "step": 85
+    },
+    {
+      "epoch": 0.09540979059769795,
+      "grad_norm": 0.33607375621795654,
+      "learning_rate": 9.520295202952028e-05,
+      "loss": 0.6068,
+      "step": 86
+    },
+    {
+      "epoch": 0.09651920676743864,
+      "grad_norm": 0.2963770031929016,
+      "learning_rate": 9.630996309963099e-05,
+      "loss": 0.6738,
+      "step": 87
+    },
+    {
+      "epoch": 0.0976286229371793,
+      "grad_norm": 0.1896924376487732,
+      "learning_rate": 9.741697416974169e-05,
+      "loss": 0.4507,
+      "step": 88
+    },
+    {
+      "epoch": 0.09873803910691999,
+      "grad_norm": 0.17777574062347412,
+      "learning_rate": 9.852398523985238e-05,
+      "loss": 0.7229,
+      "step": 89
+    },
+    {
+      "epoch": 0.09984745527666065,
+      "grad_norm": 0.2619733512401581,
+      "learning_rate": 9.963099630996309e-05,
+      "loss": 0.6666,
+      "step": 90
+    },
+    {
+      "epoch": 0.10095687144640134,
+      "grad_norm": 0.24837014079093933,
+      "learning_rate": 0.00010073800738007379,
+      "loss": 0.6532,
+      "step": 91
+    },
+    {
+      "epoch": 0.102066287616142,
+      "grad_norm": 0.16982443630695343,
+      "learning_rate": 0.0001018450184501845,
+      "loss": 0.6299,
+      "step": 92
+    },
+    {
+      "epoch": 0.10317570378588269,
+      "grad_norm": 0.25121009349823,
+      "learning_rate": 0.0001029520295202952,
+      "loss": 0.5286,
+      "step": 93
+    },
+    {
+      "epoch": 0.10428511995562335,
+      "grad_norm": 0.17865754663944244,
+      "learning_rate": 0.0001040590405904059,
+      "loss": 0.6088,
+      "step": 94
+    },
+    {
+      "epoch": 0.10539453612536402,
+      "grad_norm": 0.21300899982452393,
+      "learning_rate": 0.0001051660516605166,
+      "loss": 0.5066,
+      "step": 95
+    },
+    {
+      "epoch": 0.1065039522951047,
+      "grad_norm": 0.18651628494262695,
+      "learning_rate": 0.0001062730627306273,
+      "loss": 0.5439,
+      "step": 96
+    },
+    {
+      "epoch": 0.10761336846484537,
+      "grad_norm": 0.18930092453956604,
+      "learning_rate": 0.00010738007380073799,
+      "loss": 0.4745,
+      "step": 97
+    },
+    {
+      "epoch": 0.10872278463458605,
+      "grad_norm": 0.1967380791902542,
+      "learning_rate": 0.00010848708487084869,
+      "loss": 0.5471,
+      "step": 98
+    },
+    {
+      "epoch": 0.10983220080432672,
+      "grad_norm": 0.16241402924060822,
+      "learning_rate": 0.0001095940959409594,
+      "loss": 0.6002,
+      "step": 99
+    },
+    {
+      "epoch": 0.1109416169740674,
+      "grad_norm": 0.22282204031944275,
+      "learning_rate": 0.0001107011070110701,
+      "loss": 0.4687,
+      "step": 100
+    },
+    {
+      "epoch": 0.11205103314380807,
+      "grad_norm": 0.1546042561531067,
+      "learning_rate": 0.0001118081180811808,
+      "loss": 0.599,
+      "step": 101
+    },
+    {
+      "epoch": 0.11316044931354874,
+      "grad_norm": 0.19606682658195496,
+      "learning_rate": 0.00011291512915129151,
+      "loss": 0.5543,
+      "step": 102
+    },
+    {
+      "epoch": 0.11426986548328942,
+      "grad_norm": 0.16278928518295288,
+      "learning_rate": 0.00011402214022140221,
+      "loss": 0.4827,
+      "step": 103
+    },
+    {
+      "epoch": 0.11537928165303009,
+      "grad_norm": 0.18518386781215668,
+      "learning_rate": 0.00011512915129151292,
+      "loss": 0.6719,
+      "step": 104
+    },
+    {
+      "epoch": 0.11648869782277077,
+      "grad_norm": 0.3365796208381653,
+      "learning_rate": 0.00011623616236162362,
+      "loss": 0.3853,
+      "step": 105
+    },
+    {
+      "epoch": 0.11759811399251144,
+      "grad_norm": 0.15232332050800323,
+      "learning_rate": 0.0001173431734317343,
+      "loss": 0.6667,
+      "step": 106
+    },
+    {
+      "epoch": 0.11870753016225212,
+      "grad_norm": 0.15742583572864532,
+      "learning_rate": 0.000118450184501845,
+      "loss": 0.6997,
+      "step": 107
+    },
+    {
+      "epoch": 0.11981694633199279,
+      "grad_norm": 0.18328164517879486,
+      "learning_rate": 0.0001195571955719557,
+      "loss": 0.4386,
+      "step": 108
+    },
+    {
+      "epoch": 0.12092636250173346,
+      "grad_norm": 0.19277305901050568,
+      "learning_rate": 0.00012066420664206641,
+      "loss": 0.4237,
+      "step": 109
+    },
+    {
+      "epoch": 0.12203577867147414,
+      "grad_norm": 0.20428280532360077,
+      "learning_rate": 0.00012177121771217711,
+      "loss": 0.8255,
+      "step": 110
+    },
+    {
+      "epoch": 0.1231451948412148,
+      "grad_norm": 0.21956562995910645,
+      "learning_rate": 0.0001228782287822878,
+      "loss": 0.5919,
+      "step": 111
+    },
+    {
+      "epoch": 0.12425461101095549,
+      "grad_norm": 0.16980376839637756,
+      "learning_rate": 0.0001239852398523985,
+      "loss": 0.3784,
+      "step": 112
+    },
+    {
+      "epoch": 0.12536402718069617,
+      "grad_norm": 0.14789903163909912,
+      "learning_rate": 0.0001250922509225092,
+      "loss": 0.4501,
+      "step": 113
+    },
+    {
+      "epoch": 0.12647344335043684,
+      "grad_norm": 0.22972984611988068,
+      "learning_rate": 0.00012619926199261992,
+      "loss": 0.5988,
+      "step": 114
+    },
+    {
+      "epoch": 0.1275828595201775,
+      "grad_norm": 0.345342218875885,
+      "learning_rate": 0.00012730627306273062,
+      "loss": 0.6277,
+      "step": 115
+    },
+    {
+      "epoch": 0.12869227568991817,
+      "grad_norm": 0.22726920247077942,
+      "learning_rate": 0.0001284132841328413,
+      "loss": 0.7774,
+      "step": 116
+    },
+    {
+      "epoch": 0.12980169185965884,
+      "grad_norm": 0.18863454461097717,
+      "learning_rate": 0.000129520295202952,
+      "loss": 0.8846,
+      "step": 117
+    },
+    {
+      "epoch": 0.13091110802939954,
+      "grad_norm": 0.14772231876850128,
+      "learning_rate": 0.0001306273062730627,
+      "loss": 0.546,
+      "step": 118
+    },
+    {
+      "epoch": 0.1320205241991402,
+      "grad_norm": 0.16707932949066162,
+      "learning_rate": 0.0001317343173431734,
+      "loss": 0.4935,
+      "step": 119
+    },
+    {
+      "epoch": 0.13312994036888087,
+      "grad_norm": 0.17367342114448547,
+      "learning_rate": 0.00013284132841328411,
+      "loss": 0.5853,
+      "step": 120
+    },
+    {
+      "epoch": 0.13423935653862154,
+      "grad_norm": 0.20932604372501373,
+      "learning_rate": 0.00013394833948339482,
+      "loss": 0.5861,
+      "step": 121
+    },
+    {
+      "epoch": 0.13534877270836224,
+      "grad_norm": 0.19261795282363892,
+      "learning_rate": 0.00013505535055350552,
+      "loss": 0.5912,
+      "step": 122
+    },
+    {
+      "epoch": 0.1364581888781029,
+      "grad_norm": 0.16762207448482513,
+      "learning_rate": 0.00013616236162361623,
+      "loss": 0.6856,
+      "step": 123
+    },
+    {
+      "epoch": 0.13756760504784357,
+      "grad_norm": 0.22033065557479858,
+      "learning_rate": 0.00013726937269372693,
+      "loss": 0.6217,
+      "step": 124
+    },
+    {
+      "epoch": 0.13867702121758424,
+      "grad_norm": 0.1888565719127655,
+      "learning_rate": 0.00013837638376383763,
+      "loss": 0.4837,
+      "step": 125
+    },
+    {
+      "epoch": 0.1397864373873249,
+      "grad_norm": 0.13172639906406403,
+      "learning_rate": 0.0001394833948339483,
+      "loss": 0.4298,
+      "step": 126
+    },
+    {
+      "epoch": 0.1408958535570656,
+      "grad_norm": 0.20497459173202515,
+      "learning_rate": 0.00014059040590405902,
+      "loss": 0.7032,
+      "step": 127
+    },
+    {
+      "epoch": 0.14200526972680627,
+      "grad_norm": 0.6339800357818604,
+      "learning_rate": 0.00014169741697416972,
+      "loss": 0.5669,
+      "step": 128
+    },
+    {
+      "epoch": 0.14311468589654694,
+      "grad_norm": 0.25784072279930115,
+      "learning_rate": 0.00014280442804428042,
+      "loss": 0.6115,
+      "step": 129
+    },
+    {
+      "epoch": 0.1442241020662876,
+      "grad_norm": 0.1341128647327423,
+      "learning_rate": 0.00014391143911439113,
+      "loss": 0.5487,
+      "step": 130
+    },
+    {
+      "epoch": 0.1453335182360283,
+      "grad_norm": 0.15084423124790192,
+      "learning_rate": 0.00014501845018450183,
+      "loss": 0.5858,
+      "step": 131
+    },
+    {
+      "epoch": 0.14644293440576897,
+      "grad_norm": 0.3109219968318939,
+      "learning_rate": 0.00014612546125461254,
+      "loss": 0.5663,
+      "step": 132
+    },
+    {
+      "epoch": 0.14755235057550964,
+      "grad_norm": 0.20024478435516357,
+      "learning_rate": 0.00014723247232472324,
+      "loss": 0.49,
+      "step": 133
+    },
+    {
+      "epoch": 0.1486617667452503,
+      "grad_norm": 0.2169143110513687,
+      "learning_rate": 0.00014833948339483394,
+      "loss": 0.5955,
+      "step": 134
+    },
+    {
+      "epoch": 0.14977118291499097,
+      "grad_norm": 0.12287949025630951,
+      "learning_rate": 0.00014944649446494465,
+      "loss": 0.5406,
+      "step": 135
+    },
+    {
+      "epoch": 0.15088059908473167,
+      "grad_norm": 0.1676974594593048,
+      "learning_rate": 0.00015055350553505533,
+      "loss": 0.5796,
+      "step": 136
+    },
+    {
+      "epoch": 0.15199001525447234,
+      "grad_norm": 0.18620753288269043,
+      "learning_rate": 0.00015166051660516606,
+      "loss": 0.4466,
+      "step": 137
+    },
+    {
+      "epoch": 0.153099431424213,
+      "grad_norm": 0.2430751472711563,
+      "learning_rate": 0.00015276752767527673,
+      "loss": 0.5272,
+      "step": 138
+    },
+    {
+      "epoch": 0.15420884759395367,
+      "grad_norm": 0.21565105020999908,
+      "learning_rate": 0.00015387453874538746,
+      "loss": 0.7445,
+      "step": 139
+    },
+    {
+      "epoch": 0.15531826376369437,
+      "grad_norm": 0.14998437464237213,
+      "learning_rate": 0.00015498154981549814,
+      "loss": 0.7069,
+      "step": 140
+    },
+    {
+      "epoch": 0.15642767993343504,
+      "grad_norm": 0.22991682589054108,
+      "learning_rate": 0.00015608856088560882,
+      "loss": 0.5109,
+      "step": 141
+    },
+    {
+      "epoch": 0.1575370961031757,
+      "grad_norm": 0.21704424917697906,
+      "learning_rate": 0.00015719557195571955,
+      "loss": 0.5226,
+      "step": 142
+    },
+    {
+      "epoch": 0.15864651227291637,
+      "grad_norm": 0.141441211104393,
+      "learning_rate": 0.00015830258302583023,
+      "loss": 0.6123,
+      "step": 143
+    },
+    {
+      "epoch": 0.15975592844265704,
+      "grad_norm": 0.14867329597473145,
+      "learning_rate": 0.00015940959409594096,
+      "loss": 0.606,
+      "step": 144
+    },
+    {
+      "epoch": 0.16086534461239774,
+      "grad_norm": 0.37929290533065796,
+      "learning_rate": 0.00016051660516605164,
+      "loss": 0.5971,
+      "step": 145
+    },
+    {
+      "epoch": 0.1619747607821384,
+      "grad_norm": 0.19486796855926514,
+      "learning_rate": 0.00016162361623616237,
+      "loss": 0.554,
+      "step": 146
+    },
+    {
+      "epoch": 0.16308417695187907,
+      "grad_norm": 0.16117985546588898,
+      "learning_rate": 0.00016273062730627304,
+      "loss": 0.5379,
+      "step": 147
+    },
+    {
+      "epoch": 0.16419359312161974,
+      "grad_norm": 0.21857589483261108,
+      "learning_rate": 0.00016383763837638377,
+      "loss": 0.6984,
+      "step": 148
+    },
+    {
+      "epoch": 0.1653030092913604,
+      "grad_norm": 0.18841134011745453,
+      "learning_rate": 0.00016494464944649445,
+      "loss": 0.7714,
+      "step": 149
+    },
+    {
+      "epoch": 0.1664124254611011,
+      "grad_norm": 0.16624371707439423,
+      "learning_rate": 0.00016605166051660516,
+      "loss": 0.5173,
+      "step": 150
+    },
+    {
+      "epoch": 0.16752184163084177,
+      "grad_norm": 0.17951254546642303,
+      "learning_rate": 0.00016715867158671586,
+      "loss": 0.8636,
+      "step": 151
+    },
+    {
+      "epoch": 0.16863125780058244,
+      "grad_norm": 0.1839190572500229,
+      "learning_rate": 0.00016826568265682654,
+      "loss": 0.4827,
+      "step": 152
+    },
+    {
+      "epoch": 0.1697406739703231,
+      "grad_norm": 0.12314256280660629,
+      "learning_rate": 0.00016937269372693727,
+      "loss": 0.512,
+      "step": 153
+    },
+    {
+      "epoch": 0.1708500901400638,
+      "grad_norm": 0.12132935225963593,
+      "learning_rate": 0.00017047970479704795,
+      "loss": 0.5284,
+      "step": 154
+    },
+    {
+      "epoch": 0.17195950630980447,
+      "grad_norm": 0.29458367824554443,
+      "learning_rate": 0.00017158671586715868,
+      "loss": 0.703,
+      "step": 155
+    },
+    {
+      "epoch": 0.17306892247954514,
+      "grad_norm": 0.15419234335422516,
+      "learning_rate": 0.00017269372693726935,
+      "loss": 0.6275,
+      "step": 156
+    },
+    {
+      "epoch": 0.1741783386492858,
+      "grad_norm": 0.12458593398332596,
+      "learning_rate": 0.00017380073800738006,
+      "loss": 0.4877,
+      "step": 157
+    },
+    {
+      "epoch": 0.17528775481902648,
+      "grad_norm": 0.1632385402917862,
+      "learning_rate": 0.00017490774907749076,
+      "loss": 0.5432,
+      "step": 158
+    },
+    {
+      "epoch": 0.17639717098876717,
+      "grad_norm": 0.12995319068431854,
+      "learning_rate": 0.00017601476014760147,
+      "loss": 0.5964,
+      "step": 159
+    },
+    {
+      "epoch": 0.17750658715850784,
+      "grad_norm": 0.09651859104633331,
+      "learning_rate": 0.00017712177121771217,
+      "loss": 0.5907,
+      "step": 160
+    },
+    {
+      "epoch": 0.1786160033282485,
+      "grad_norm": 0.20671844482421875,
+      "learning_rate": 0.00017822878228782285,
+      "loss": 0.574,
+      "step": 161
+    },
+    {
+      "epoch": 0.17972541949798918,
+      "grad_norm": 0.1272733509540558,
+      "learning_rate": 0.00017933579335793358,
+      "loss": 0.5413,
+      "step": 162
+    },
+    {
+      "epoch": 0.18083483566772987,
+      "grad_norm": 0.12072700262069702,
+      "learning_rate": 0.00018044280442804426,
+      "loss": 0.4555,
+      "step": 163
+    },
+    {
+      "epoch": 0.18194425183747054,
+      "grad_norm": 0.18685540556907654,
+      "learning_rate": 0.00018154981549815496,
+      "loss": 0.7267,
+      "step": 164
+    },
+    {
+      "epoch": 0.1830536680072112,
+      "grad_norm": 0.18687789142131805,
+      "learning_rate": 0.00018265682656826566,
+      "loss": 0.5785,
+      "step": 165
+    },
+    {
+      "epoch": 0.18416308417695187,
+      "grad_norm": 0.2580098807811737,
+      "learning_rate": 0.00018376383763837637,
+      "loss": 0.5662,
+      "step": 166
+    },
+    {
+      "epoch": 0.18527250034669254,
+      "grad_norm": 0.1476382166147232,
+      "learning_rate": 0.00018487084870848707,
+      "loss": 0.5427,
+      "step": 167
+    },
+    {
+      "epoch": 0.18638191651643324,
+      "grad_norm": 0.14074724912643433,
+      "learning_rate": 0.00018597785977859778,
+      "loss": 0.4759,
+      "step": 168
+    },
+    {
+      "epoch": 0.1874913326861739,
+      "grad_norm": 0.2741679847240448,
+      "learning_rate": 0.00018708487084870848,
+      "loss": 0.6973,
+      "step": 169
+    },
+    {
+      "epoch": 0.18860074885591457,
+      "grad_norm": 0.2278490662574768,
+      "learning_rate": 0.00018819188191881918,
+      "loss": 0.586,
+      "step": 170
+    },
+    {
+      "epoch": 0.18971016502565524,
+      "grad_norm": 0.1516968458890915,
+      "learning_rate": 0.00018929889298892986,
+      "loss": 0.378,
+      "step": 171
+    },
+    {
+      "epoch": 0.1908195811953959,
+      "grad_norm": 0.17562097311019897,
+      "learning_rate": 0.00019040590405904056,
+      "loss": 0.6113,
+      "step": 172
+    },
+    {
+      "epoch": 0.1919289973651366,
+      "grad_norm": 0.14954030513763428,
+      "learning_rate": 0.00019151291512915127,
+      "loss": 0.4926,
+      "step": 173
+    },
+    {
+      "epoch": 0.19303841353487727,
+      "grad_norm": 0.1906110644340515,
+      "learning_rate": 0.00019261992619926197,
+      "loss": 0.4884,
+      "step": 174
+    },
+    {
+      "epoch": 0.19414782970461794,
+      "grad_norm": 0.18351547420024872,
+      "learning_rate": 0.00019372693726937268,
+      "loss": 0.5295,
+      "step": 175
+    },
+    {
+      "epoch": 0.1952572458743586,
+      "grad_norm": 0.2049722820520401,
+      "learning_rate": 0.00019483394833948338,
+      "loss": 0.4562,
+      "step": 176
+    },
+    {
+      "epoch": 0.1963666620440993,
+      "grad_norm": 0.21823738515377045,
+      "learning_rate": 0.00019594095940959409,
+      "loss": 0.5329,
+      "step": 177
+    },
+    {
+      "epoch": 0.19747607821383997,
+      "grad_norm": 0.1077185720205307,
+      "learning_rate": 0.00019704797047970476,
+      "loss": 0.4787,
+      "step": 178
+    },
+    {
+      "epoch": 0.19858549438358064,
+      "grad_norm": 0.19151321053504944,
+      "learning_rate": 0.0001981549815498155,
+      "loss": 0.7764,
+      "step": 179
+    },
+    {
+      "epoch": 0.1996949105533213,
+      "grad_norm": 0.16796617209911346,
+      "learning_rate": 0.00019926199261992617,
+      "loss": 0.4962,
+      "step": 180
+    },
+    {
+      "epoch": 0.20080432672306198,
+      "grad_norm": 0.31866180896759033,
+      "learning_rate": 0.00020036900369003687,
+      "loss": 0.8462,
+      "step": 181
+    },
+    {
+      "epoch": 0.20191374289280267,
+      "grad_norm": 0.2093004584312439,
+      "learning_rate": 0.00020147601476014758,
+      "loss": 0.5916,
+      "step": 182
+    },
+    {
+      "epoch": 0.20302315906254334,
+      "grad_norm": 0.2198261171579361,
+      "learning_rate": 0.00020258302583025828,
+      "loss": 0.434,
+      "step": 183
+    },
+    {
+      "epoch": 0.204132575232284,
+      "grad_norm": 0.19753794372081757,
+      "learning_rate": 0.000203690036900369,
+      "loss": 0.6142,
+      "step": 184
+    },
+    {
+      "epoch": 0.20524199140202468,
+      "grad_norm": 0.1537386178970337,
+      "learning_rate": 0.00020479704797047966,
+      "loss": 0.4053,
+      "step": 185
+    },
+    {
+      "epoch": 0.20635140757176537,
+      "grad_norm": 0.194119393825531,
+      "learning_rate": 0.0002059040590405904,
+      "loss": 0.6506,
+      "step": 186
+    },
+    {
+      "epoch": 0.20746082374150604,
+      "grad_norm": 0.16899624466896057,
+      "learning_rate": 0.00020701107011070107,
+      "loss": 0.572,
+      "step": 187
+    },
+    {
+      "epoch": 0.2085702399112467,
+      "grad_norm": 0.14211402833461761,
+      "learning_rate": 0.0002081180811808118,
+      "loss": 0.5999,
+      "step": 188
+    },
+    {
+      "epoch": 0.20967965608098738,
+      "grad_norm": 0.17124126851558685,
+      "learning_rate": 0.00020922509225092248,
+      "loss": 0.4797,
+      "step": 189
+    },
+    {
+      "epoch": 0.21078907225072804,
+      "grad_norm": 0.1637413501739502,
+      "learning_rate": 0.0002103321033210332,
+      "loss": 0.3899,
+      "step": 190
+    },
+    {
+      "epoch": 0.21189848842046874,
+      "grad_norm": 0.2928311228752136,
+      "learning_rate": 0.0002114391143911439,
+      "loss": 0.4297,
+      "step": 191
+    },
+    {
+      "epoch": 0.2130079045902094,
+      "grad_norm": 0.19800159335136414,
+      "learning_rate": 0.0002125461254612546,
+      "loss": 0.6641,
+      "step": 192
+    },
+    {
+      "epoch": 0.21411732075995007,
+      "grad_norm": 0.1297946572303772,
+      "learning_rate": 0.0002136531365313653,
+      "loss": 0.4557,
+      "step": 193
+    },
+    {
+      "epoch": 0.21522673692969074,
+      "grad_norm": 0.24120758473873138,
+      "learning_rate": 0.00021476014760147597,
+      "loss": 0.7058,
+      "step": 194
+    },
+    {
+      "epoch": 0.2163361530994314,
+      "grad_norm": 0.15839487314224243,
+      "learning_rate": 0.0002158671586715867,
+      "loss": 0.5106,
+      "step": 195
+    },
+    {
+      "epoch": 0.2174455692691721,
+      "grad_norm": 0.18157745897769928,
+      "learning_rate": 0.00021697416974169738,
+      "loss": 0.8263,
+      "step": 196
+    },
+    {
+      "epoch": 0.21855498543891277,
+      "grad_norm": 0.2212788164615631,
+      "learning_rate": 0.0002180811808118081,
+      "loss": 0.4721,
+      "step": 197
+    },
+    {
+      "epoch": 0.21966440160865344,
+      "grad_norm": 0.2079571783542633,
+      "learning_rate": 0.0002191881918819188,
+      "loss": 0.5491,
+      "step": 198
+    },
+    {
+      "epoch": 0.2207738177783941,
+      "grad_norm": 0.29977646470069885,
+      "learning_rate": 0.00022029520295202952,
+      "loss": 0.5165,
+      "step": 199
+    },
+    {
+      "epoch": 0.2218832339481348,
+      "grad_norm": 0.15032988786697388,
+      "learning_rate": 0.0002214022140221402,
+      "loss": 0.5833,
+      "step": 200
+    },
+    {
+      "epoch": 0.22299265011787547,
+      "grad_norm": 0.1367197334766388,
+      "learning_rate": 0.00022250922509225088,
+      "loss": 0.4895,
+      "step": 201
+    },
+    {
+      "epoch": 0.22410206628761614,
+      "grad_norm": 0.20162621140480042,
+      "learning_rate": 0.0002236162361623616,
+      "loss": 0.6069,
+      "step": 202
+    },
+    {
+      "epoch": 0.2252114824573568,
+      "grad_norm": 0.1476079821586609,
+      "learning_rate": 0.00022472324723247228,
+      "loss": 0.3195,
+      "step": 203
+    },
+    {
+      "epoch": 0.22632089862709748,
+      "grad_norm": 0.16391322016716003,
+      "learning_rate": 0.00022583025830258302,
+      "loss": 0.5242,
+      "step": 204
+    },
+    {
+      "epoch": 0.22743031479683817,
+      "grad_norm": 0.32637152075767517,
+      "learning_rate": 0.0002269372693726937,
+      "loss": 0.608,
+      "step": 205
+    },
+    {
+      "epoch": 0.22853973096657884,
+      "grad_norm": 0.18414971232414246,
+      "learning_rate": 0.00022804428044280442,
+      "loss": 0.5414,
+      "step": 206
+    },
+    {
+      "epoch": 0.2296491471363195,
+      "grad_norm": 0.13513615727424622,
+      "learning_rate": 0.0002291512915129151,
+      "loss": 0.452,
+      "step": 207
+    },
+    {
+      "epoch": 0.23075856330606018,
+      "grad_norm": 0.1715661883354187,
+      "learning_rate": 0.00023025830258302583,
+      "loss": 0.5552,
+      "step": 208
+    },
+    {
+      "epoch": 0.23186797947580087,
+      "grad_norm": 0.2078085094690323,
+      "learning_rate": 0.0002313653136531365,
+      "loss": 0.6519,
+      "step": 209
+    },
+    {
+      "epoch": 0.23297739564554154,
+      "grad_norm": 0.45334669947624207,
+      "learning_rate": 0.00023247232472324724,
+      "loss": 0.9767,
+      "step": 210
+    },
+    {
+      "epoch": 0.2340868118152822,
+      "grad_norm": 0.17819558084011078,
+      "learning_rate": 0.00023357933579335792,
+      "loss": 0.7568,
+      "step": 211
+    },
+    {
+      "epoch": 0.23519622798502288,
+      "grad_norm": 0.4460853338241577,
+      "learning_rate": 0.0002346863468634686,
+      "loss": 0.4265,
+      "step": 212
+    },
+    {
+      "epoch": 0.23630564415476354,
+      "grad_norm": 0.2014990597963333,
+      "learning_rate": 0.00023579335793357933,
+      "loss": 0.6479,
+      "step": 213
+    },
+    {
+      "epoch": 0.23741506032450424,
+      "grad_norm": 0.17922167479991913,
+      "learning_rate": 0.00023690036900369,
+      "loss": 0.6739,
+      "step": 214
+    },
+    {
+      "epoch": 0.2385244764942449,
+      "grad_norm": 0.14204619824886322,
+      "learning_rate": 0.00023800738007380073,
+      "loss": 0.4641,
+      "step": 215
+    },
+    {
+      "epoch": 0.23963389266398558,
+      "grad_norm": 0.17847496271133423,
+      "learning_rate": 0.0002391143911439114,
+      "loss": 0.6229,
+      "step": 216
+    },
+    {
+      "epoch": 0.24074330883372624,
+      "grad_norm": 0.20707765221595764,
+      "learning_rate": 0.00024022140221402214,
+      "loss": 0.6337,
+      "step": 217
+    },
+    {
+      "epoch": 0.2418527250034669,
+      "grad_norm": 0.1623317301273346,
+      "learning_rate": 0.00024132841328413282,
+      "loss": 0.5256,
+      "step": 218
+    },
+    {
+      "epoch": 0.2429621411732076,
+      "grad_norm": 0.18768182396888733,
+      "learning_rate": 0.00024243542435424352,
+      "loss": 0.5793,
+      "step": 219
+    },
+    {
+      "epoch": 0.24407155734294828,
+      "grad_norm": 0.16495372354984283,
+      "learning_rate": 0.00024354243542435423,
+      "loss": 0.4539,
+      "step": 220
+    },
+    {
+      "epoch": 0.24518097351268894,
+      "grad_norm": 0.14188359677791595,
+      "learning_rate": 0.00024464944649446493,
+      "loss": 0.5089,
+      "step": 221
+    },
+    {
+      "epoch": 0.2462903896824296,
+      "grad_norm": 0.18091897666454315,
+      "learning_rate": 0.0002457564575645756,
+      "loss": 0.5689,
+      "step": 222
+    },
+    {
+      "epoch": 0.2473998058521703,
+      "grad_norm": 0.214166522026062,
+      "learning_rate": 0.00024686346863468634,
+      "loss": 0.6479,
+      "step": 223
+    },
+    {
+      "epoch": 0.24850922202191097,
+      "grad_norm": 0.19823165237903595,
+      "learning_rate": 0.000247970479704797,
+      "loss": 0.4772,
+      "step": 224
+    },
+    {
+      "epoch": 0.24961863819165164,
+      "grad_norm": 0.20246858894824982,
+      "learning_rate": 0.0002490774907749077,
+      "loss": 0.6838,
+      "step": 225
+    },
+    {
+      "epoch": 0.25072805436139234,
+      "grad_norm": 0.17839235067367554,
+      "learning_rate": 0.0002501845018450184,
+      "loss": 0.6234,
+      "step": 226
+    },
+    {
+      "epoch": 0.251837470531133,
+      "grad_norm": 0.15513737499713898,
+      "learning_rate": 0.0002512915129151291,
+      "loss": 0.6836,
+      "step": 227
+    },
+    {
+      "epoch": 0.2529468867008737,
+      "grad_norm": 0.3762029707431793,
+      "learning_rate": 0.00025239852398523983,
+      "loss": 0.6588,
+      "step": 228
+    },
+    {
+      "epoch": 0.2540563028706143,
+      "grad_norm": 0.24218355119228363,
+      "learning_rate": 0.0002535055350553505,
+      "loss": 0.663,
+      "step": 229
+    },
+    {
+      "epoch": 0.255165719040355,
+      "grad_norm": 0.23746894299983978,
+      "learning_rate": 0.00025461254612546124,
+      "loss": 0.8055,
+      "step": 230
+    },
+    {
+      "epoch": 0.2562751352100957,
+      "grad_norm": 0.19097627699375153,
+      "learning_rate": 0.0002557195571955719,
+      "loss": 0.5028,
+      "step": 231
+    },
+    {
+      "epoch": 0.25738455137983635,
+      "grad_norm": 0.16248895227909088,
+      "learning_rate": 0.0002568265682656826,
+      "loss": 0.4644,
+      "step": 232
+    },
+    {
+      "epoch": 0.25849396754957704,
+      "grad_norm": 0.15660254657268524,
+      "learning_rate": 0.0002579335793357933,
+      "loss": 0.4822,
+      "step": 233
+    },
+    {
+      "epoch": 0.2596033837193177,
+      "grad_norm": 0.18094629049301147,
+      "learning_rate": 0.000259040590405904,
+      "loss": 0.6761,
+      "step": 234
+    },
+    {
+      "epoch": 0.2607127998890584,
+      "grad_norm": 0.20552562177181244,
+      "learning_rate": 0.00026014760147601473,
+      "loss": 0.654,
+      "step": 235
+    },
+    {
+      "epoch": 0.2618222160587991,
+      "grad_norm": 0.1620112657546997,
+      "learning_rate": 0.0002612546125461254,
+      "loss": 0.3786,
+      "step": 236
+    },
+    {
+      "epoch": 0.2629316322285397,
+      "grad_norm": 0.25460562109947205,
+      "learning_rate": 0.00026236162361623614,
+      "loss": 0.6258,
+      "step": 237
+    },
+    {
+      "epoch": 0.2640410483982804,
+      "grad_norm": 0.17765319347381592,
+      "learning_rate": 0.0002634686346863468,
+      "loss": 0.3955,
+      "step": 238
+    },
+    {
+      "epoch": 0.2651504645680211,
+      "grad_norm": 0.2620762884616852,
+      "learning_rate": 0.00026457564575645755,
+      "loss": 0.5884,
+      "step": 239
+    },
+    {
+      "epoch": 0.26625988073776174,
+      "grad_norm": 0.14502941071987152,
+      "learning_rate": 0.00026568265682656823,
+      "loss": 0.5391,
+      "step": 240
+    },
+    {
+      "epoch": 0.26736929690750244,
+      "grad_norm": 0.15013466775417328,
+      "learning_rate": 0.0002667896678966789,
+      "loss": 0.5676,
+      "step": 241
+    },
+    {
+      "epoch": 0.2684787130772431,
+      "grad_norm": 0.2311590611934662,
+      "learning_rate": 0.00026789667896678964,
+      "loss": 0.6409,
+      "step": 242
+    },
+    {
+      "epoch": 0.2695881292469838,
+      "grad_norm": 0.16972240805625916,
+      "learning_rate": 0.0002690036900369003,
+      "loss": 0.488,
+      "step": 243
+    },
+    {
+      "epoch": 0.27069754541672447,
+      "grad_norm": 0.13695970177650452,
+      "learning_rate": 0.00027011070110701104,
+      "loss": 0.5997,
+      "step": 244
+    },
+    {
+      "epoch": 0.2718069615864651,
+      "grad_norm": 0.16296394169330597,
+      "learning_rate": 0.0002712177121771217,
+      "loss": 0.5645,
+      "step": 245
+    },
+    {
+      "epoch": 0.2729163777562058,
+      "grad_norm": 0.15607796609401703,
+      "learning_rate": 0.00027232472324723245,
+      "loss": 0.3976,
+      "step": 246
+    },
+    {
+      "epoch": 0.27402579392594645,
+      "grad_norm": 0.17693723738193512,
+      "learning_rate": 0.00027343173431734313,
+      "loss": 0.6873,
+      "step": 247
+    },
+    {
+      "epoch": 0.27513521009568714,
+      "grad_norm": 0.16682426631450653,
+      "learning_rate": 0.00027453874538745386,
+      "loss": 0.5259,
+      "step": 248
+    },
+    {
+      "epoch": 0.27624462626542784,
+      "grad_norm": 0.16855113208293915,
+      "learning_rate": 0.00027564575645756454,
+      "loss": 0.5889,
+      "step": 249
+    },
+    {
+      "epoch": 0.2773540424351685,
+      "grad_norm": 0.160100057721138,
+      "learning_rate": 0.00027675276752767527,
+      "loss": 0.5832,
+      "step": 250
+    },
+    {
+      "epoch": 0.2784634586049092,
+      "grad_norm": 0.24193750321865082,
+      "learning_rate": 0.00027785977859778595,
+      "loss": 0.4325,
+      "step": 251
+    },
+    {
+      "epoch": 0.2795728747746498,
+      "grad_norm": 0.12794721126556396,
+      "learning_rate": 0.0002789667896678966,
+      "loss": 0.7371,
+      "step": 252
+    },
+    {
+      "epoch": 0.2806822909443905,
+      "grad_norm": 0.22222159802913666,
+      "learning_rate": 0.00028007380073800735,
+      "loss": 0.6177,
+      "step": 253
+    },
+    {
+      "epoch": 0.2817917071141312,
+      "grad_norm": 0.24400316178798676,
+      "learning_rate": 0.00028118081180811803,
+      "loss": 0.5169,
+      "step": 254
+    },
+    {
+      "epoch": 0.28290112328387185,
+      "grad_norm": 0.23530688881874084,
+      "learning_rate": 0.00028228782287822876,
+      "loss": 0.4384,
+      "step": 255
+    },
+    {
+      "epoch": 0.28401053945361254,
+      "grad_norm": 0.20001864433288574,
+      "learning_rate": 0.00028339483394833944,
+      "loss": 0.4855,
+      "step": 256
+    },
+    {
+      "epoch": 0.2851199556233532,
+      "grad_norm": 0.1441182792186737,
+      "learning_rate": 0.00028450184501845017,
+      "loss": 0.5467,
+      "step": 257
+    },
+    {
+      "epoch": 0.2862293717930939,
+      "grad_norm": 0.17499203979969025,
+      "learning_rate": 0.00028560885608856085,
+      "loss": 0.6093,
+      "step": 258
+    },
+    {
+      "epoch": 0.2873387879628346,
+      "grad_norm": 0.13634347915649414,
+      "learning_rate": 0.0002867158671586716,
+      "loss": 0.5183,
+      "step": 259
+    },
+    {
+      "epoch": 0.2884482041325752,
+      "grad_norm": 0.1825854778289795,
+      "learning_rate": 0.00028782287822878226,
+      "loss": 0.524,
+      "step": 260
+    },
+    {
+      "epoch": 0.2895576203023159,
+      "grad_norm": 0.1797315776348114,
+      "learning_rate": 0.00028892988929889293,
+      "loss": 0.9663,
+      "step": 261
+    },
+    {
+      "epoch": 0.2906670364720566,
+      "grad_norm": 0.22019031643867493,
+      "learning_rate": 0.00029003690036900366,
+      "loss": 0.8477,
+      "step": 262
+    },
+    {
+      "epoch": 0.29177645264179725,
+      "grad_norm": 0.18698155879974365,
+      "learning_rate": 0.00029114391143911434,
+      "loss": 0.5928,
+      "step": 263
+    },
+    {
+      "epoch": 0.29288586881153794,
+      "grad_norm": 0.18208318948745728,
+      "learning_rate": 0.00029225092250922507,
+      "loss": 0.5586,
+      "step": 264
+    },
+    {
+      "epoch": 0.2939952849812786,
+      "grad_norm": 0.17802134156227112,
+      "learning_rate": 0.00029335793357933575,
+      "loss": 0.499,
+      "step": 265
+    },
+    {
+      "epoch": 0.2951047011510193,
+      "grad_norm": 0.25552839040756226,
+      "learning_rate": 0.0002944649446494465,
+      "loss": 0.6052,
+      "step": 266
+    },
+    {
+      "epoch": 0.29621411732076,
+      "grad_norm": 0.15605288743972778,
+      "learning_rate": 0.00029557195571955716,
+      "loss": 0.4767,
+      "step": 267
+    },
+    {
+      "epoch": 0.2973235334905006,
+      "grad_norm": 0.18597030639648438,
+      "learning_rate": 0.0002966789667896679,
+      "loss": 0.6442,
+      "step": 268
+    },
+    {
+      "epoch": 0.2984329496602413,
+      "grad_norm": 0.2584044933319092,
+      "learning_rate": 0.00029778597785977857,
+      "loss": 0.6259,
+      "step": 269
+    },
+    {
+      "epoch": 0.29954236582998195,
+      "grad_norm": 0.13973256945610046,
+      "learning_rate": 0.0002988929889298893,
+      "loss": 0.6127,
+      "step": 270
+    },
+    {
+      "epoch": 0.30065178199972264,
+      "grad_norm": 0.21028122305870056,
+      "learning_rate": 0.0003,
+      "loss": 0.3969,
+      "step": 271
+    },
+    {
+      "epoch": 0.30176119816946334,
+      "grad_norm": 0.232606902718544,
+      "learning_rate": 0.00029987664473684207,
+      "loss": 0.6801,
+      "step": 272
+    },
+    {
+      "epoch": 0.302870614339204,
+      "grad_norm": 0.24834930896759033,
+      "learning_rate": 0.00029975328947368416,
+      "loss": 0.5395,
+      "step": 273
+    },
+    {
+      "epoch": 0.3039800305089447,
+      "grad_norm": 0.2343815118074417,
+      "learning_rate": 0.00029962993421052625,
+      "loss": 0.6029,
+      "step": 274
+    },
+    {
+      "epoch": 0.3050894466786853,
+      "grad_norm": 0.23860520124435425,
+      "learning_rate": 0.0002995065789473684,
+      "loss": 0.4872,
+      "step": 275
+    },
+    {
+      "epoch": 0.306198862848426,
+      "grad_norm": 0.1793919801712036,
+      "learning_rate": 0.0002993832236842105,
+      "loss": 0.6896,
+      "step": 276
+    },
+    {
+      "epoch": 0.3073082790181667,
+      "grad_norm": 0.2507120668888092,
+      "learning_rate": 0.00029925986842105264,
+      "loss": 0.3792,
+      "step": 277
+    },
+    {
+      "epoch": 0.30841769518790735,
+      "grad_norm": 0.27677059173583984,
+      "learning_rate": 0.00029913651315789473,
+      "loss": 0.7856,
+      "step": 278
+    },
+    {
+      "epoch": 0.30952711135764804,
+      "grad_norm": 0.18887469172477722,
+      "learning_rate": 0.00029901315789473683,
+      "loss": 0.5406,
+      "step": 279
+    },
+    {
+      "epoch": 0.31063652752738874,
+      "grad_norm": 0.23371614515781403,
+      "learning_rate": 0.0002988898026315789,
+      "loss": 0.5748,
+      "step": 280
+    },
+    {
+      "epoch": 0.3117459436971294,
+      "grad_norm": 0.22486557066440582,
+      "learning_rate": 0.000298766447368421,
+      "loss": 0.7443,
+      "step": 281
+    },
+    {
+      "epoch": 0.3128553598668701,
+      "grad_norm": 0.17992804944515228,
+      "learning_rate": 0.0002986430921052631,
+      "loss": 0.5819,
+      "step": 282
+    },
+    {
+      "epoch": 0.3139647760366107,
+      "grad_norm": 0.20137208700180054,
+      "learning_rate": 0.00029851973684210525,
+      "loss": 0.722,
+      "step": 283
+    },
+    {
+      "epoch": 0.3150741922063514,
+      "grad_norm": 0.25975537300109863,
+      "learning_rate": 0.00029839638157894735,
+      "loss": 0.5349,
+      "step": 284
+    },
+    {
+      "epoch": 0.3161836083760921,
+      "grad_norm": 0.2687530219554901,
+      "learning_rate": 0.00029827302631578944,
+      "loss": 0.4361,
+      "step": 285
+    },
+    {
+      "epoch": 0.31729302454583275,
+      "grad_norm": 0.18794257938861847,
+      "learning_rate": 0.0002981496710526316,
+      "loss": 0.5822,
+      "step": 286
+    },
+    {
+      "epoch": 0.31840244071557344,
+      "grad_norm": 0.12230537086725235,
+      "learning_rate": 0.0002980263157894737,
+      "loss": 0.3639,
+      "step": 287
+    },
+    {
+      "epoch": 0.3195118568853141,
+      "grad_norm": 0.20607517659664154,
+      "learning_rate": 0.0002979029605263158,
+      "loss": 0.6476,
+      "step": 288
+    },
+    {
+      "epoch": 0.3206212730550548,
+      "grad_norm": 0.11566425859928131,
+      "learning_rate": 0.00029777960526315787,
+      "loss": 0.5026,
+      "step": 289
+    },
+    {
+      "epoch": 0.3217306892247955,
+      "grad_norm": 0.22872845828533173,
+      "learning_rate": 0.00029765624999999996,
+      "loss": 0.5727,
+      "step": 290
+    },
+    {
+      "epoch": 0.3228401053945361,
+      "grad_norm": 0.17686223983764648,
+      "learning_rate": 0.00029753289473684205,
+      "loss": 0.6867,
+      "step": 291
+    },
+    {
+      "epoch": 0.3239495215642768,
+      "grad_norm": 0.2232068032026291,
+      "learning_rate": 0.0002974095394736842,
+      "loss": 0.4983,
+      "step": 292
+    },
+    {
+      "epoch": 0.32505893773401745,
+      "grad_norm": 0.26865360140800476,
+      "learning_rate": 0.0002972861842105263,
+      "loss": 0.5151,
+      "step": 293
+    },
+    {
+      "epoch": 0.32616835390375815,
+      "grad_norm": 0.14152151346206665,
+      "learning_rate": 0.0002971628289473684,
+      "loss": 0.7316,
+      "step": 294
+    },
+    {
+      "epoch": 0.32727777007349884,
+      "grad_norm": 0.20797161757946014,
+      "learning_rate": 0.00029703947368421054,
+      "loss": 0.4479,
+      "step": 295
+    },
+    {
+      "epoch": 0.3283871862432395,
+      "grad_norm": 0.1608234941959381,
+      "learning_rate": 0.00029691611842105263,
+      "loss": 0.5482,
+      "step": 296
+    },
+    {
+      "epoch": 0.3294966024129802,
+      "grad_norm": 0.2219133973121643,
+      "learning_rate": 0.0002967927631578947,
+      "loss": 0.4474,
+      "step": 297
+    },
+    {
+      "epoch": 0.3306060185827208,
+      "grad_norm": 0.3990642726421356,
+      "learning_rate": 0.0002966694078947368,
+      "loss": 0.6421,
+      "step": 298
+    },
+    {
+      "epoch": 0.3317154347524615,
+      "grad_norm": 0.20786860585212708,
+      "learning_rate": 0.0002965460526315789,
+      "loss": 0.5569,
+      "step": 299
+    },
+    {
+      "epoch": 0.3328248509222022,
+      "grad_norm": 0.1967337280511856,
+      "learning_rate": 0.000296422697368421,
+      "loss": 0.6013,
+      "step": 300
+    },
+    {
+      "epoch": 0.33393426709194285,
+      "grad_norm": 0.19602453708648682,
+      "learning_rate": 0.00029629934210526315,
+      "loss": 0.5516,
+      "step": 301
+    },
+    {
+      "epoch": 0.33504368326168354,
+      "grad_norm": 0.23872393369674683,
+      "learning_rate": 0.00029617598684210524,
+      "loss": 0.5124,
+      "step": 302
+    },
+    {
+      "epoch": 0.33615309943142424,
+      "grad_norm": 0.20506146550178528,
+      "learning_rate": 0.00029605263157894733,
+      "loss": 0.6206,
+      "step": 303
+    },
+    {
+      "epoch": 0.3372625156011649,
+      "grad_norm": 0.20456762611865997,
+      "learning_rate": 0.00029592927631578943,
+      "loss": 0.4852,
+      "step": 304
+    },
+    {
+      "epoch": 0.3383719317709056,
+      "grad_norm": 0.23360048234462738,
+      "learning_rate": 0.0002958059210526316,
+      "loss": 0.4821,
+      "step": 305
+    },
+    {
+      "epoch": 0.3394813479406462,
+      "grad_norm": 0.16443900763988495,
+      "learning_rate": 0.00029568256578947367,
+      "loss": 0.4089,
+      "step": 306
+    },
+    {
+      "epoch": 0.3405907641103869,
+      "grad_norm": 0.15696674585342407,
+      "learning_rate": 0.00029555921052631576,
+      "loss": 0.5479,
+      "step": 307
+    },
+    {
+      "epoch": 0.3417001802801276,
+      "grad_norm": 0.20780624449253082,
+      "learning_rate": 0.00029543585526315785,
+      "loss": 0.4644,
+      "step": 308
+    },
+    {
+      "epoch": 0.34280959644986825,
+      "grad_norm": 0.185526043176651,
+      "learning_rate": 0.00029531249999999995,
+      "loss": 0.7259,
+      "step": 309
+    },
+    {
+      "epoch": 0.34391901261960894,
+      "grad_norm": 0.1540479063987732,
+      "learning_rate": 0.0002951891447368421,
+      "loss": 0.6167,
+      "step": 310
+    },
+    {
+      "epoch": 0.3450284287893496,
+      "grad_norm": 0.26093733310699463,
+      "learning_rate": 0.0002950657894736842,
+      "loss": 0.6427,
+      "step": 311
+    },
+    {
+      "epoch": 0.3461378449590903,
+      "grad_norm": 0.21596834063529968,
+      "learning_rate": 0.0002949424342105263,
+      "loss": 0.5916,
+      "step": 312
+    },
+    {
+      "epoch": 0.347247261128831,
+      "grad_norm": 0.20977520942687988,
+      "learning_rate": 0.0002948190789473684,
+      "loss": 0.6572,
+      "step": 313
+    },
+    {
+      "epoch": 0.3483566772985716,
+      "grad_norm": 0.1886155605316162,
+      "learning_rate": 0.0002946957236842105,
+      "loss": 0.5399,
+      "step": 314
+    },
+    {
+      "epoch": 0.3494660934683123,
+      "grad_norm": 0.18215329945087433,
+      "learning_rate": 0.0002945723684210526,
+      "loss": 0.5724,
+      "step": 315
+    },
+    {
+      "epoch": 0.35057550963805295,
+      "grad_norm": 0.2055482119321823,
+      "learning_rate": 0.0002944490131578947,
+      "loss": 0.5834,
+      "step": 316
+    },
+    {
+      "epoch": 0.35168492580779365,
+      "grad_norm": 0.15489786863327026,
+      "learning_rate": 0.0002943256578947368,
+      "loss": 0.5003,
+      "step": 317
+    },
+    {
+      "epoch": 0.35279434197753434,
+      "grad_norm": 0.17881543934345245,
+      "learning_rate": 0.00029420230263157895,
+      "loss": 0.5105,
+      "step": 318
+    },
+    {
+      "epoch": 0.353903758147275,
+      "grad_norm": 0.1768861562013626,
+      "learning_rate": 0.00029407894736842104,
+      "loss": 0.5363,
+      "step": 319
+    },
+    {
+      "epoch": 0.3550131743170157,
+      "grad_norm": 0.2283925861120224,
+      "learning_rate": 0.00029395559210526314,
+      "loss": 0.5452,
+      "step": 320
+    },
+    {
+      "epoch": 0.3561225904867563,
+      "grad_norm": 0.19439998269081116,
+      "learning_rate": 0.00029383223684210523,
+      "loss": 0.5603,
+      "step": 321
+    },
+    {
+      "epoch": 0.357232006656497,
+      "grad_norm": 0.2097710222005844,
+      "learning_rate": 0.0002937088815789473,
+      "loss": 0.625,
+      "step": 322
+    },
+    {
+      "epoch": 0.3583414228262377,
+      "grad_norm": 0.186342254281044,
+      "learning_rate": 0.0002935855263157894,
+      "loss": 0.3553,
+      "step": 323
+    },
+    {
+      "epoch": 0.35945083899597835,
+      "grad_norm": 0.275612473487854,
+      "learning_rate": 0.00029346217105263156,
+      "loss": 0.6274,
+      "step": 324
+    },
+    {
+      "epoch": 0.36056025516571905,
+      "grad_norm": 0.19332240521907806,
+      "learning_rate": 0.00029333881578947366,
+      "loss": 0.5827,
+      "step": 325
+    },
+    {
+      "epoch": 0.36166967133545974,
+      "grad_norm": 0.18259958922863007,
+      "learning_rate": 0.00029321546052631575,
+      "loss": 0.4502,
+      "step": 326
+    },
+    {
+      "epoch": 0.3627790875052004,
+      "grad_norm": 0.25983044505119324,
+      "learning_rate": 0.0002930921052631579,
+      "loss": 0.8896,
+      "step": 327
+    },
+    {
+      "epoch": 0.3638885036749411,
+      "grad_norm": 0.1671958863735199,
+      "learning_rate": 0.00029296875,
+      "loss": 0.5232,
+      "step": 328
+    },
+    {
+      "epoch": 0.3649979198446817,
+      "grad_norm": 0.18751101195812225,
+      "learning_rate": 0.0002928453947368421,
+      "loss": 0.4556,
+      "step": 329
+    },
+    {
+      "epoch": 0.3661073360144224,
+      "grad_norm": 0.19776886701583862,
+      "learning_rate": 0.0002927220394736842,
+      "loss": 0.8467,
+      "step": 330
+    },
+    {
+      "epoch": 0.3672167521841631,
+      "grad_norm": 0.2738226354122162,
+      "learning_rate": 0.00029259868421052627,
+      "loss": 0.7323,
+      "step": 331
+    },
+    {
+      "epoch": 0.36832616835390375,
+      "grad_norm": 0.16836367547512054,
+      "learning_rate": 0.00029247532894736836,
+      "loss": 0.43,
+      "step": 332
+    },
+    {
+      "epoch": 0.36943558452364444,
+      "grad_norm": 0.30866721272468567,
+      "learning_rate": 0.0002923519736842105,
+      "loss": 0.3759,
+      "step": 333
+    },
+    {
+      "epoch": 0.3705450006933851,
+      "grad_norm": 0.20256434381008148,
+      "learning_rate": 0.0002922286184210526,
+      "loss": 0.6727,
+      "step": 334
+    },
+    {
+      "epoch": 0.3716544168631258,
+      "grad_norm": 0.18781127035617828,
+      "learning_rate": 0.0002921052631578947,
+      "loss": 0.5784,
+      "step": 335
+    },
+    {
+      "epoch": 0.3727638330328665,
+      "grad_norm": 0.3620914816856384,
+      "learning_rate": 0.00029198190789473684,
+      "loss": 0.8316,
+      "step": 336
+    },
+    {
+      "epoch": 0.3738732492026071,
+      "grad_norm": 0.1937960535287857,
+      "learning_rate": 0.00029185855263157894,
+      "loss": 0.5956,
+      "step": 337
+    },
+    {
+      "epoch": 0.3749826653723478,
+      "grad_norm": 0.21955540776252747,
+      "learning_rate": 0.00029173519736842103,
+      "loss": 0.6847,
+      "step": 338
+    },
+    {
+      "epoch": 0.37609208154208845,
+      "grad_norm": 0.22091850638389587,
+      "learning_rate": 0.0002916118421052631,
+      "loss": 0.7571,
+      "step": 339
+    },
+    {
+      "epoch": 0.37720149771182915,
+      "grad_norm": 0.17953120172023773,
+      "learning_rate": 0.0002914884868421052,
+      "loss": 0.596,
+      "step": 340
+    },
+    {
+      "epoch": 0.37831091388156984,
+      "grad_norm": 0.2171243131160736,
+      "learning_rate": 0.0002913651315789473,
+      "loss": 0.5556,
+      "step": 341
+    },
+    {
+      "epoch": 0.3794203300513105,
+      "grad_norm": 0.21151772141456604,
+      "learning_rate": 0.00029124177631578946,
+      "loss": 0.8455,
+      "step": 342
+    },
+    {
+      "epoch": 0.3805297462210512,
+      "grad_norm": 0.21299928426742554,
+      "learning_rate": 0.00029111842105263155,
+      "loss": 0.4473,
+      "step": 343
+    },
+    {
+      "epoch": 0.3816391623907918,
+      "grad_norm": 0.20761217176914215,
+      "learning_rate": 0.00029099506578947364,
+      "loss": 0.6742,
+      "step": 344
+    },
+    {
+      "epoch": 0.3827485785605325,
+      "grad_norm": 0.21071919798851013,
+      "learning_rate": 0.0002908717105263158,
+      "loss": 0.5404,
+      "step": 345
+    },
+    {
+      "epoch": 0.3838579947302732,
+      "grad_norm": 0.3248625099658966,
+      "learning_rate": 0.0002907483552631579,
+      "loss": 0.7006,
+      "step": 346
+    },
+    {
+      "epoch": 0.38496741090001385,
+      "grad_norm": 0.3169274628162384,
+      "learning_rate": 0.000290625,
+      "loss": 0.5953,
+      "step": 347
+    },
+    {
+      "epoch": 0.38607682706975455,
+      "grad_norm": 0.16656579077243805,
+      "learning_rate": 0.00029050164473684207,
+      "loss": 0.5392,
+      "step": 348
+    },
+    {
+      "epoch": 0.38718624323949524,
+      "grad_norm": 0.1759122908115387,
+      "learning_rate": 0.00029037828947368416,
+      "loss": 0.6298,
+      "step": 349
+    },
+    {
+      "epoch": 0.3882956594092359,
+      "grad_norm": 0.17183220386505127,
+      "learning_rate": 0.00029025493421052626,
+      "loss": 0.49,
+      "step": 350
+    },
+    {
+      "epoch": 0.3894050755789766,
+      "grad_norm": 0.2180389016866684,
+      "learning_rate": 0.0002901315789473684,
+      "loss": 0.4447,
+      "step": 351
+    },
+    {
+      "epoch": 0.3905144917487172,
+      "grad_norm": 0.3889177143573761,
+      "learning_rate": 0.0002900082236842105,
+      "loss": 0.48,
+      "step": 352
+    },
+    {
+      "epoch": 0.3916239079184579,
+      "grad_norm": 0.22550411522388458,
+      "learning_rate": 0.00028988486842105264,
+      "loss": 0.5976,
+      "step": 353
+    },
+    {
+      "epoch": 0.3927333240881986,
+      "grad_norm": 0.20059050619602203,
+      "learning_rate": 0.00028976151315789474,
+      "loss": 0.578,
+      "step": 354
+    },
+    {
+      "epoch": 0.39384274025793925,
+      "grad_norm": 0.2586449086666107,
+      "learning_rate": 0.00028963815789473683,
+      "loss": 0.6453,
+      "step": 355
+    },
+    {
+      "epoch": 0.39495215642767995,
+      "grad_norm": 0.24946491420269012,
+      "learning_rate": 0.0002895148026315789,
+      "loss": 0.649,
+      "step": 356
+    },
+    {
+      "epoch": 0.3960615725974206,
+      "grad_norm": 0.1763986498117447,
+      "learning_rate": 0.000289391447368421,
+      "loss": 0.6183,
+      "step": 357
+    },
+    {
+      "epoch": 0.3971709887671613,
+      "grad_norm": 0.1732664704322815,
+      "learning_rate": 0.0002892680921052631,
+      "loss": 0.8931,
+      "step": 358
+    },
+    {
+      "epoch": 0.398280404936902,
+      "grad_norm": 0.22204923629760742,
+      "learning_rate": 0.0002891447368421052,
+      "loss": 0.5979,
+      "step": 359
+    },
+    {
+      "epoch": 0.3993898211066426,
+      "grad_norm": 0.1942061334848404,
+      "learning_rate": 0.00028902138157894735,
+      "loss": 0.5693,
+      "step": 360
+    },
+    {
+      "epoch": 0.4004992372763833,
+      "grad_norm": 0.2240975797176361,
+      "learning_rate": 0.00028889802631578944,
+      "loss": 0.6672,
+      "step": 361
+    },
+    {
+      "epoch": 0.40160865344612395,
+      "grad_norm": 0.19991931319236755,
+      "learning_rate": 0.0002887746710526316,
+      "loss": 0.7095,
+      "step": 362
+    },
+    {
+      "epoch": 0.40271806961586465,
+      "grad_norm": 0.1844676285982132,
+      "learning_rate": 0.0002886513157894737,
+      "loss": 0.5337,
+      "step": 363
+    },
+    {
+      "epoch": 0.40382748578560534,
+      "grad_norm": 0.2115306556224823,
+      "learning_rate": 0.0002885279605263158,
+      "loss": 0.6032,
+      "step": 364
+    },
+    {
+      "epoch": 0.404936901955346,
+      "grad_norm": 0.2993430495262146,
+      "learning_rate": 0.00028840460526315787,
+      "loss": 0.8077,
+      "step": 365
+    },
+    {
+      "epoch": 0.4060463181250867,
+      "grad_norm": 0.41001948714256287,
+      "learning_rate": 0.00028828124999999996,
+      "loss": 0.7461,
+      "step": 366
+    },
+    {
+      "epoch": 0.4071557342948273,
+      "grad_norm": 0.1562204509973526,
+      "learning_rate": 0.00028815789473684206,
+      "loss": 0.4517,
+      "step": 367
+    },
+    {
+      "epoch": 0.408265150464568,
+      "grad_norm": 0.27741947770118713,
+      "learning_rate": 0.0002880345394736842,
+      "loss": 0.6076,
+      "step": 368
+    },
+    {
+      "epoch": 0.4093745666343087,
+      "grad_norm": 0.205497607588768,
+      "learning_rate": 0.0002879111842105263,
+      "loss": 0.6499,
+      "step": 369
+    },
+    {
+      "epoch": 0.41048398280404935,
+      "grad_norm": 0.412622332572937,
+      "learning_rate": 0.0002877878289473684,
+      "loss": 0.6487,
+      "step": 370
+    },
+    {
+      "epoch": 0.41159339897379005,
+      "grad_norm": 0.19786472618579865,
+      "learning_rate": 0.00028766447368421054,
+      "loss": 0.4202,
+      "step": 371
+    },
+    {
+      "epoch": 0.41270281514353074,
+      "grad_norm": 0.1875920295715332,
+      "learning_rate": 0.00028754111842105263,
+      "loss": 0.5377,
+      "step": 372
+    },
+    {
+      "epoch": 0.4138122313132714,
+      "grad_norm": 0.22071506083011627,
+      "learning_rate": 0.0002874177631578947,
+      "loss": 0.4671,
+      "step": 373
+    },
+    {
+      "epoch": 0.4149216474830121,
+      "grad_norm": 0.22277134656906128,
+      "learning_rate": 0.0002872944078947368,
+      "loss": 0.6436,
+      "step": 374
+    },
+    {
+      "epoch": 0.4160310636527527,
+      "grad_norm": 0.13663825392723083,
+      "learning_rate": 0.0002871710526315789,
+      "loss": 0.3595,
+      "step": 375
+    },
+    {
+      "epoch": 0.4171404798224934,
+      "grad_norm": 0.25280505418777466,
+      "learning_rate": 0.000287047697368421,
+      "loss": 0.7154,
+      "step": 376
+    },
+    {
+      "epoch": 0.4182498959922341,
+      "grad_norm": 0.2542460262775421,
+      "learning_rate": 0.00028692434210526315,
+      "loss": 0.5229,
+      "step": 377
+    },
+    {
+      "epoch": 0.41935931216197475,
+      "grad_norm": 0.20687228441238403,
+      "learning_rate": 0.00028680098684210524,
+      "loss": 0.4229,
+      "step": 378
+    },
+    {
+      "epoch": 0.42046872833171545,
+      "grad_norm": 0.1827574074268341,
+      "learning_rate": 0.00028667763157894734,
+      "loss": 0.504,
+      "step": 379
+    },
+    {
+      "epoch": 0.4215781445014561,
+      "grad_norm": 0.22924618422985077,
+      "learning_rate": 0.00028655427631578943,
+      "loss": 0.4645,
+      "step": 380
+    },
+    {
+      "epoch": 0.4226875606711968,
+      "grad_norm": 0.1500721126794815,
+      "learning_rate": 0.0002864309210526316,
+      "loss": 0.2866,
+      "step": 381
+    },
+    {
+      "epoch": 0.4237969768409375,
+      "grad_norm": 0.19137370586395264,
+      "learning_rate": 0.00028630756578947367,
+      "loss": 0.4739,
+      "step": 382
+    },
+    {
+      "epoch": 0.4249063930106781,
+      "grad_norm": 0.1940913051366806,
+      "learning_rate": 0.00028618421052631576,
+      "loss": 0.6433,
+      "step": 383
+    },
+    {
+      "epoch": 0.4260158091804188,
+      "grad_norm": 0.17999312281608582,
+      "learning_rate": 0.00028606085526315786,
+      "loss": 0.6118,
+      "step": 384
+    },
+    {
+      "epoch": 0.42712522535015945,
+      "grad_norm": 0.21522557735443115,
+      "learning_rate": 0.00028593749999999995,
+      "loss": 0.5608,
+      "step": 385
+    },
+    {
+      "epoch": 0.42823464151990015,
+      "grad_norm": 0.23753724992275238,
+      "learning_rate": 0.0002858141447368421,
+      "loss": 0.4775,
+      "step": 386
+    },
+    {
+      "epoch": 0.42934405768964085,
+      "grad_norm": 0.28104132413864136,
+      "learning_rate": 0.0002856907894736842,
+      "loss": 0.7179,
+      "step": 387
+    },
+    {
+      "epoch": 0.4304534738593815,
+      "grad_norm": 0.16530390083789825,
+      "learning_rate": 0.0002855674342105263,
+      "loss": 0.5765,
+      "step": 388
+    },
+    {
+      "epoch": 0.4315628900291222,
+      "grad_norm": 0.20358699560165405,
+      "learning_rate": 0.0002854440789473684,
+      "loss": 0.6325,
+      "step": 389
+    },
+    {
+      "epoch": 0.4326723061988628,
+      "grad_norm": 0.17629845440387726,
+      "learning_rate": 0.0002853207236842105,
+      "loss": 0.5528,
+      "step": 390
+    },
+    {
+      "epoch": 0.4337817223686035,
+      "grad_norm": 0.25051596760749817,
+      "learning_rate": 0.0002851973684210526,
+      "loss": 0.6712,
+      "step": 391
+    },
+    {
+      "epoch": 0.4348911385383442,
+      "grad_norm": 0.19358691573143005,
+      "learning_rate": 0.0002850740131578947,
+      "loss": 0.5572,
+      "step": 392
+    },
+    {
+      "epoch": 0.43600055470808485,
+      "grad_norm": 0.13769972324371338,
+      "learning_rate": 0.0002849506578947368,
+      "loss": 0.5055,
+      "step": 393
+    },
+    {
+      "epoch": 0.43710997087782555,
+      "grad_norm": 0.14807964861392975,
+      "learning_rate": 0.0002848273026315789,
+      "loss": 0.5963,
+      "step": 394
+    },
+    {
+      "epoch": 0.43821938704756624,
+      "grad_norm": 0.16840098798274994,
+      "learning_rate": 0.00028470394736842104,
+      "loss": 0.4203,
+      "step": 395
+    },
+    {
+      "epoch": 0.4393288032173069,
+      "grad_norm": 0.16224174201488495,
+      "learning_rate": 0.00028458059210526314,
+      "loss": 0.7647,
+      "step": 396
+    },
+    {
+      "epoch": 0.4404382193870476,
+      "grad_norm": 0.2029973566532135,
+      "learning_rate": 0.00028445723684210523,
+      "loss": 0.6427,
+      "step": 397
+    },
+    {
+      "epoch": 0.4415476355567882,
+      "grad_norm": 0.23025457561016083,
+      "learning_rate": 0.0002843338815789473,
+      "loss": 0.5838,
+      "step": 398
+    },
+    {
+      "epoch": 0.4426570517265289,
+      "grad_norm": 0.2005675584077835,
+      "learning_rate": 0.0002842105263157894,
+      "loss": 0.6441,
+      "step": 399
+    },
+    {
+      "epoch": 0.4437664678962696,
+      "grad_norm": 0.2208050787448883,
+      "learning_rate": 0.00028408717105263156,
+      "loss": 0.656,
+      "step": 400
+    },
+    {
+      "epoch": 0.44487588406601025,
+      "grad_norm": 0.28568586707115173,
+      "learning_rate": 0.00028396381578947366,
+      "loss": 0.6623,
+      "step": 401
+    },
+    {
+      "epoch": 0.44598530023575095,
+      "grad_norm": 0.22206373512744904,
+      "learning_rate": 0.00028384046052631575,
+      "loss": 0.5584,
+      "step": 402
+    },
+    {
+      "epoch": 0.4470947164054916,
+      "grad_norm": 0.20780105888843536,
+      "learning_rate": 0.0002837171052631579,
+      "loss": 0.5801,
+      "step": 403
+    },
+    {
+      "epoch": 0.4482041325752323,
+      "grad_norm": 0.20285826921463013,
+      "learning_rate": 0.00028359375,
+      "loss": 0.6751,
+      "step": 404
+    },
+    {
+      "epoch": 0.449313548744973,
+      "grad_norm": 0.15252311527729034,
+      "learning_rate": 0.0002834703947368421,
+      "loss": 1.1206,
+      "step": 405
+    },
+    {
+      "epoch": 0.4504229649147136,
+      "grad_norm": 0.23378504812717438,
+      "learning_rate": 0.0002833470394736842,
+      "loss": 0.6577,
+      "step": 406
+    },
+    {
+      "epoch": 0.4515323810844543,
+      "grad_norm": 0.1787406951189041,
+      "learning_rate": 0.00028322368421052627,
+      "loss": 0.7497,
+      "step": 407
+    },
+    {
+      "epoch": 0.45264179725419496,
+      "grad_norm": 0.48000597953796387,
+      "learning_rate": 0.00028310032894736836,
+      "loss": 0.752,
+      "step": 408
+    },
+    {
+      "epoch": 0.45375121342393565,
+      "grad_norm": 0.2794741094112396,
+      "learning_rate": 0.0002829769736842105,
+      "loss": 0.5731,
+      "step": 409
+    },
+    {
+      "epoch": 0.45486062959367635,
+      "grad_norm": 0.2653048634529114,
+      "learning_rate": 0.0002828536184210526,
+      "loss": 0.4867,
+      "step": 410
+    },
+    {
+      "epoch": 0.455970045763417,
+      "grad_norm": 0.19287265837192535,
+      "learning_rate": 0.0002827302631578947,
+      "loss": 0.643,
+      "step": 411
+    },
+    {
+      "epoch": 0.4570794619331577,
+      "grad_norm": 0.2332431823015213,
+      "learning_rate": 0.00028260690789473685,
+      "loss": 0.4386,
+      "step": 412
+    },
+    {
+      "epoch": 0.4581888781028983,
+      "grad_norm": 0.21826831996440887,
+      "learning_rate": 0.00028248355263157894,
+      "loss": 0.548,
+      "step": 413
+    },
+    {
+      "epoch": 0.459298294272639,
+      "grad_norm": 0.23967108130455017,
+      "learning_rate": 0.00028236019736842103,
+      "loss": 0.5739,
+      "step": 414
+    },
+    {
+      "epoch": 0.4604077104423797,
+      "grad_norm": 0.18406794965267181,
+      "learning_rate": 0.0002822368421052631,
+      "loss": 0.4547,
+      "step": 415
+    },
+    {
+      "epoch": 0.46151712661212035,
+      "grad_norm": 0.18926838040351868,
+      "learning_rate": 0.0002821134868421052,
+      "loss": 0.5208,
+      "step": 416
+    },
+    {
+      "epoch": 0.46262654278186105,
+      "grad_norm": 0.16055501997470856,
+      "learning_rate": 0.0002819901315789473,
+      "loss": 0.4273,
+      "step": 417
+    },
+    {
+      "epoch": 0.46373595895160175,
+      "grad_norm": 0.18228095769882202,
+      "learning_rate": 0.00028186677631578946,
+      "loss": 0.6194,
+      "step": 418
+    },
+    {
+      "epoch": 0.4648453751213424,
+      "grad_norm": 0.1525285542011261,
+      "learning_rate": 0.00028174342105263155,
+      "loss": 0.5515,
+      "step": 419
+    },
+    {
+      "epoch": 0.4659547912910831,
+      "grad_norm": 0.24812228977680206,
+      "learning_rate": 0.00028162006578947365,
+      "loss": 0.4666,
+      "step": 420
+    },
+    {
+      "epoch": 0.4670642074608237,
+      "grad_norm": 0.3236676752567291,
+      "learning_rate": 0.0002814967105263158,
+      "loss": 0.6294,
+      "step": 421
+    },
+    {
+      "epoch": 0.4681736236305644,
+      "grad_norm": 0.15236404538154602,
+      "learning_rate": 0.0002813733552631579,
+      "loss": 0.4148,
+      "step": 422
+    },
+    {
+      "epoch": 0.4692830398003051,
+      "grad_norm": 0.20661397278308868,
+      "learning_rate": 0.00028125,
+      "loss": 0.4696,
+      "step": 423
+    },
+    {
+      "epoch": 0.47039245597004575,
+      "grad_norm": 0.3039199113845825,
+      "learning_rate": 0.00028112664473684207,
+      "loss": 0.7764,
+      "step": 424
+    },
+    {
+      "epoch": 0.47150187213978645,
+      "grad_norm": 0.2243487536907196,
+      "learning_rate": 0.00028100328947368417,
+      "loss": 0.575,
+      "step": 425
+    },
+    {
+      "epoch": 0.4726112883095271,
+      "grad_norm": 0.15252256393432617,
+      "learning_rate": 0.00028087993421052626,
+      "loss": 0.6139,
+      "step": 426
+    },
+    {
+      "epoch": 0.4737207044792678,
+      "grad_norm": 0.2288883924484253,
+      "learning_rate": 0.0002807565789473684,
+      "loss": 0.5199,
+      "step": 427
+    },
+    {
+      "epoch": 0.4748301206490085,
+      "grad_norm": 0.1678112894296646,
+      "learning_rate": 0.0002806332236842105,
+      "loss": 0.699,
+      "step": 428
+    },
+    {
+      "epoch": 0.4759395368187491,
+      "grad_norm": 0.27088475227355957,
+      "learning_rate": 0.00028050986842105265,
+      "loss": 0.5913,
+      "step": 429
+    },
+    {
+      "epoch": 0.4770489529884898,
+      "grad_norm": 0.19018007814884186,
+      "learning_rate": 0.00028038651315789474,
+      "loss": 0.5341,
+      "step": 430
+    },
+    {
+      "epoch": 0.47815836915823046,
+      "grad_norm": 0.19086478650569916,
+      "learning_rate": 0.00028026315789473683,
+      "loss": 0.565,
+      "step": 431
+    },
+    {
+      "epoch": 0.47926778532797115,
+      "grad_norm": 0.26616349816322327,
+      "learning_rate": 0.0002801398026315789,
+      "loss": 0.7576,
+      "step": 432
+    },
+    {
+      "epoch": 0.48037720149771185,
+      "grad_norm": 0.16013029217720032,
+      "learning_rate": 0.000280016447368421,
+      "loss": 0.5242,
+      "step": 433
+    },
+    {
+      "epoch": 0.4814866176674525,
+      "grad_norm": 0.16926300525665283,
+      "learning_rate": 0.0002798930921052631,
+      "loss": 0.4289,
+      "step": 434
+    },
+    {
+      "epoch": 0.4825960338371932,
+      "grad_norm": 0.2056371569633484,
+      "learning_rate": 0.0002797697368421052,
+      "loss": 0.5971,
+      "step": 435
+    },
+    {
+      "epoch": 0.4837054500069338,
+      "grad_norm": 0.1635441929101944,
+      "learning_rate": 0.00027964638157894735,
+      "loss": 0.4277,
+      "step": 436
+    },
+    {
+      "epoch": 0.4848148661766745,
+      "grad_norm": 0.24154643714427948,
+      "learning_rate": 0.00027952302631578945,
+      "loss": 0.6508,
+      "step": 437
+    },
+    {
+      "epoch": 0.4859242823464152,
+      "grad_norm": 0.2069096565246582,
+      "learning_rate": 0.0002793996710526316,
+      "loss": 0.5105,
+      "step": 438
+    },
+    {
+      "epoch": 0.48703369851615586,
+      "grad_norm": 0.23238608241081238,
+      "learning_rate": 0.0002792763157894737,
+      "loss": 0.7069,
+      "step": 439
+    },
+    {
+      "epoch": 0.48814311468589655,
+      "grad_norm": 0.18234537541866302,
+      "learning_rate": 0.0002791529605263158,
+      "loss": 0.5141,
+      "step": 440
+    },
+    {
+      "epoch": 0.48925253085563725,
+      "grad_norm": 0.1497894525527954,
+      "learning_rate": 0.00027902960526315787,
+      "loss": 0.5206,
+      "step": 441
+    },
+    {
+      "epoch": 0.4903619470253779,
+      "grad_norm": 0.2433656007051468,
+      "learning_rate": 0.00027890624999999997,
+      "loss": 0.5697,
+      "step": 442
+    },
+    {
+      "epoch": 0.4914713631951186,
+      "grad_norm": 0.1533818542957306,
+      "learning_rate": 0.00027878289473684206,
+      "loss": 0.359,
+      "step": 443
+    },
+    {
+      "epoch": 0.4925807793648592,
+      "grad_norm": 0.2589110732078552,
+      "learning_rate": 0.00027865953947368415,
+      "loss": 0.6381,
+      "step": 444
+    },
+    {
+      "epoch": 0.4936901955345999,
+      "grad_norm": 0.2857501208782196,
+      "learning_rate": 0.0002785361842105263,
+      "loss": 0.7064,
+      "step": 445
+    },
+    {
+      "epoch": 0.4947996117043406,
+      "grad_norm": 0.21398474276065826,
+      "learning_rate": 0.0002784128289473684,
+      "loss": 0.5152,
+      "step": 446
+    },
+    {
+      "epoch": 0.49590902787408125,
+      "grad_norm": 0.23514775931835175,
+      "learning_rate": 0.00027828947368421054,
+      "loss": 0.8143,
+      "step": 447
+    },
+    {
+      "epoch": 0.49701844404382195,
+      "grad_norm": 0.18278779089450836,
+      "learning_rate": 0.00027816611842105263,
+      "loss": 0.4275,
+      "step": 448
+    },
+    {
+      "epoch": 0.4981278602135626,
+      "grad_norm": 0.16910268366336823,
+      "learning_rate": 0.0002780427631578947,
+      "loss": 0.5477,
+      "step": 449
+    },
+    {
+      "epoch": 0.4992372763833033,
+      "grad_norm": 0.17349810898303986,
+      "learning_rate": 0.0002779194078947368,
+      "loss": 0.5224,
+      "step": 450
+    },
+    {
+      "epoch": 0.500346692553044,
+      "grad_norm": 0.2979370653629303,
+      "learning_rate": 0.0002777960526315789,
+      "loss": 0.6723,
+      "step": 451
+    },
+    {
+      "epoch": 0.5014561087227847,
+      "grad_norm": 0.2329479455947876,
+      "learning_rate": 0.000277672697368421,
+      "loss": 0.5004,
+      "step": 452
+    },
+    {
+      "epoch": 0.5025655248925253,
+      "grad_norm": 0.18267230689525604,
+      "learning_rate": 0.00027754934210526315,
+      "loss": 0.5729,
+      "step": 453
+    },
+    {
+      "epoch": 0.503674941062266,
+      "grad_norm": 0.21524755656719208,
+      "learning_rate": 0.00027742598684210525,
+      "loss": 0.4295,
+      "step": 454
+    },
+    {
+      "epoch": 0.5047843572320067,
+      "grad_norm": 0.18233224749565125,
+      "learning_rate": 0.00027730263157894734,
+      "loss": 0.4668,
+      "step": 455
+    },
+    {
+      "epoch": 0.5058937734017473,
+      "grad_norm": 0.19789119064807892,
+      "learning_rate": 0.00027717927631578943,
+      "loss": 0.6149,
+      "step": 456
+    },
+    {
+      "epoch": 0.507003189571488,
+      "grad_norm": 0.16243144869804382,
+      "learning_rate": 0.0002770559210526316,
+      "loss": 0.415,
+      "step": 457
+    },
+    {
+      "epoch": 0.5081126057412286,
+      "grad_norm": 0.22004704177379608,
+      "learning_rate": 0.0002769325657894737,
+      "loss": 0.6058,
+      "step": 458
+    },
+    {
+      "epoch": 0.5092220219109693,
+      "grad_norm": 0.1755845546722412,
+      "learning_rate": 0.00027680921052631577,
+      "loss": 0.659,
+      "step": 459
+    },
+    {
+      "epoch": 0.51033143808071,
+      "grad_norm": 0.16181863844394684,
+      "learning_rate": 0.00027668585526315786,
+      "loss": 0.5118,
+      "step": 460
+    },
+    {
+      "epoch": 0.5114408542504507,
+      "grad_norm": 0.30584779381752014,
+      "learning_rate": 0.00027656249999999995,
+      "loss": 0.5185,
+      "step": 461
+    },
+    {
+      "epoch": 0.5125502704201914,
+      "grad_norm": 0.2442709058523178,
+      "learning_rate": 0.0002764391447368421,
+      "loss": 0.6267,
+      "step": 462
+    },
+    {
+      "epoch": 0.513659686589932,
+      "grad_norm": 0.17913980782032013,
+      "learning_rate": 0.0002763157894736842,
+      "loss": 0.4123,
+      "step": 463
+    },
+    {
+      "epoch": 0.5147691027596727,
+      "grad_norm": 0.16953568160533905,
+      "learning_rate": 0.0002761924342105263,
+      "loss": 0.4336,
+      "step": 464
+    },
+    {
+      "epoch": 0.5158785189294134,
+      "grad_norm": 0.2636931538581848,
+      "learning_rate": 0.0002760690789473684,
+      "loss": 0.8255,
+      "step": 465
+    },
+    {
+      "epoch": 0.5169879350991541,
+      "grad_norm": 0.1953415870666504,
+      "learning_rate": 0.00027594572368421053,
+      "loss": 0.3879,
+      "step": 466
+    },
+    {
+      "epoch": 0.5180973512688948,
+      "grad_norm": 0.23631513118743896,
+      "learning_rate": 0.0002758223684210526,
+      "loss": 0.4591,
+      "step": 467
+    },
+    {
+      "epoch": 0.5192067674386354,
+      "grad_norm": 0.25506916642189026,
+      "learning_rate": 0.0002756990131578947,
+      "loss": 0.6347,
+      "step": 468
+    },
+    {
+      "epoch": 0.5203161836083761,
+      "grad_norm": 0.1907813549041748,
+      "learning_rate": 0.0002755756578947368,
+      "loss": 0.6049,
+      "step": 469
+    },
+    {
+      "epoch": 0.5214255997781168,
+      "grad_norm": 0.26337459683418274,
+      "learning_rate": 0.0002754523026315789,
+      "loss": 0.9162,
+      "step": 470
+    },
+    {
+      "epoch": 0.5225350159478575,
+      "grad_norm": 0.2560301721096039,
+      "learning_rate": 0.00027532894736842105,
+      "loss": 0.5243,
+      "step": 471
+    },
+    {
+      "epoch": 0.5236444321175981,
+      "grad_norm": 0.21078939735889435,
+      "learning_rate": 0.00027520559210526314,
+      "loss": 0.528,
+      "step": 472
+    },
+    {
+      "epoch": 0.5247538482873388,
+      "grad_norm": 0.19924308359622955,
+      "learning_rate": 0.00027508223684210523,
+      "loss": 0.484,
+      "step": 473
+    },
+    {
+      "epoch": 0.5258632644570794,
+      "grad_norm": 0.1312379091978073,
+      "learning_rate": 0.0002749588815789473,
+      "loss": 0.5321,
+      "step": 474
+    },
+    {
+      "epoch": 0.5269726806268201,
+      "grad_norm": 0.32874926924705505,
+      "learning_rate": 0.0002748355263157894,
+      "loss": 0.5229,
+      "step": 475
+    },
+    {
+      "epoch": 0.5280820967965608,
+      "grad_norm": 0.2065068781375885,
+      "learning_rate": 0.00027471217105263157,
+      "loss": 0.5401,
+      "step": 476
+    },
+    {
+      "epoch": 0.5291915129663015,
+      "grad_norm": 0.18827693164348602,
+      "learning_rate": 0.00027458881578947366,
+      "loss": 0.3993,
+      "step": 477
+    },
+    {
+      "epoch": 0.5303009291360422,
+      "grad_norm": 0.20950929820537567,
+      "learning_rate": 0.00027446546052631575,
+      "loss": 0.6372,
+      "step": 478
+    },
+    {
+      "epoch": 0.5314103453057828,
+      "grad_norm": 0.20649929344654083,
+      "learning_rate": 0.0002743421052631579,
+      "loss": 0.647,
+      "step": 479
+    },
+    {
+      "epoch": 0.5325197614755235,
+      "grad_norm": 0.14667537808418274,
+      "learning_rate": 0.00027421875,
+      "loss": 0.5186,
+      "step": 480
+    },
+    {
+      "epoch": 0.5336291776452642,
+      "grad_norm": 0.19212156534194946,
+      "learning_rate": 0.0002740953947368421,
+      "loss": 0.8352,
+      "step": 481
+    },
+    {
+      "epoch": 0.5347385938150049,
+      "grad_norm": 0.2529224753379822,
+      "learning_rate": 0.0002739720394736842,
+      "loss": 0.5348,
+      "step": 482
+    },
+    {
+      "epoch": 0.5358480099847456,
+      "grad_norm": 0.2153153419494629,
+      "learning_rate": 0.0002738486842105263,
+      "loss": 0.6096,
+      "step": 483
+    },
+    {
+      "epoch": 0.5369574261544862,
+      "grad_norm": 0.2908189296722412,
+      "learning_rate": 0.00027372532894736837,
+      "loss": 0.6725,
+      "step": 484
+    },
+    {
+      "epoch": 0.5380668423242269,
+      "grad_norm": 0.2697788178920746,
+      "learning_rate": 0.0002736019736842105,
+      "loss": 0.4649,
+      "step": 485
+    },
+    {
+      "epoch": 0.5391762584939676,
+      "grad_norm": 0.20288242399692535,
+      "learning_rate": 0.0002734786184210526,
+      "loss": 0.5319,
+      "step": 486
+    },
+    {
+      "epoch": 0.5402856746637082,
+      "grad_norm": 0.25883370637893677,
+      "learning_rate": 0.0002733552631578947,
+      "loss": 0.3363,
+      "step": 487
+    },
+    {
+      "epoch": 0.5413950908334489,
+      "grad_norm": 0.15326879918575287,
+      "learning_rate": 0.00027323190789473685,
+      "loss": 0.4186,
+      "step": 488
+    },
+    {
+      "epoch": 0.5425045070031895,
+      "grad_norm": 0.20244112610816956,
+      "learning_rate": 0.00027310855263157894,
+      "loss": 0.5039,
+      "step": 489
+    },
+    {
+      "epoch": 0.5436139231729302,
+      "grad_norm": 0.21093213558197021,
+      "learning_rate": 0.00027298519736842103,
+      "loss": 0.7453,
+      "step": 490
+    },
+    {
+      "epoch": 0.5447233393426709,
+      "grad_norm": 0.22425080835819244,
+      "learning_rate": 0.00027286184210526313,
+      "loss": 0.5743,
+      "step": 491
+    },
+    {
+      "epoch": 0.5458327555124116,
+      "grad_norm": 0.2680664360523224,
+      "learning_rate": 0.0002727384868421052,
+      "loss": 0.5446,
+      "step": 492
+    },
+    {
+      "epoch": 0.5469421716821523,
+      "grad_norm": 0.24040096998214722,
+      "learning_rate": 0.0002726151315789473,
+      "loss": 0.4714,
+      "step": 493
+    },
+    {
+      "epoch": 0.5480515878518929,
+      "grad_norm": 0.30965113639831543,
+      "learning_rate": 0.0002724917763157894,
+      "loss": 0.8154,
+      "step": 494
+    },
+    {
+      "epoch": 0.5491610040216336,
+      "grad_norm": 0.16201867163181305,
+      "learning_rate": 0.00027236842105263155,
+      "loss": 0.4898,
+      "step": 495
+    },
+    {
+      "epoch": 0.5502704201913743,
+      "grad_norm": 0.3444017767906189,
+      "learning_rate": 0.00027224506578947365,
+      "loss": 0.5012,
+      "step": 496
+    },
+    {
+      "epoch": 0.551379836361115,
+      "grad_norm": 0.5116562247276306,
+      "learning_rate": 0.0002721217105263158,
+      "loss": 0.4425,
+      "step": 497
+    },
+    {
+      "epoch": 0.5524892525308557,
+      "grad_norm": 0.17541073262691498,
+      "learning_rate": 0.0002719983552631579,
+      "loss": 0.6107,
+      "step": 498
+    },
+    {
+      "epoch": 0.5535986687005963,
+      "grad_norm": 0.4572921097278595,
+      "learning_rate": 0.000271875,
+      "loss": 0.5645,
+      "step": 499
+    },
+    {
+      "epoch": 0.554708084870337,
+      "grad_norm": 0.35991188883781433,
+      "learning_rate": 0.0002717516447368421,
+      "loss": 0.4648,
+      "step": 500
+    },
+    {
+      "epoch": 0.5558175010400777,
+      "grad_norm": 0.2327331304550171,
+      "learning_rate": 0.00027162828947368417,
+      "loss": 0.5978,
+      "step": 501
+    },
+    {
+      "epoch": 0.5569269172098184,
+      "grad_norm": 0.25607866048812866,
+      "learning_rate": 0.00027150493421052626,
+      "loss": 0.7341,
+      "step": 502
+    },
+    {
+      "epoch": 0.558036333379559,
+      "grad_norm": 0.26063939929008484,
+      "learning_rate": 0.0002713815789473684,
+      "loss": 0.6693,
+      "step": 503
+    },
+    {
+      "epoch": 0.5591457495492996,
+      "grad_norm": 0.23963363468647003,
+      "learning_rate": 0.0002712582236842105,
+      "loss": 0.4704,
+      "step": 504
+    },
+    {
+      "epoch": 0.5602551657190403,
+      "grad_norm": 0.21853481233119965,
+      "learning_rate": 0.0002711348684210526,
+      "loss": 0.827,
+      "step": 505
+    },
+    {
+      "epoch": 0.561364581888781,
+      "grad_norm": 0.2731577754020691,
+      "learning_rate": 0.00027101151315789474,
+      "loss": 0.5781,
+      "step": 506
+    },
+    {
+      "epoch": 0.5624739980585217,
+      "grad_norm": 0.1824404001235962,
+      "learning_rate": 0.00027088815789473684,
+      "loss": 0.5114,
+      "step": 507
+    },
+    {
+      "epoch": 0.5635834142282624,
+      "grad_norm": 0.26666054129600525,
+      "learning_rate": 0.00027076480263157893,
+      "loss": 0.5057,
+      "step": 508
+    },
+    {
+      "epoch": 0.564692830398003,
+      "grad_norm": 0.22783181071281433,
+      "learning_rate": 0.000270641447368421,
+      "loss": 0.6204,
+      "step": 509
+    },
+    {
+      "epoch": 0.5658022465677437,
+      "grad_norm": 0.20822562277317047,
+      "learning_rate": 0.0002705180921052631,
+      "loss": 0.3751,
+      "step": 510
+    },
+    {
+      "epoch": 0.5669116627374844,
+      "grad_norm": 0.1988370418548584,
+      "learning_rate": 0.0002703947368421052,
+      "loss": 0.5364,
+      "step": 511
+    },
+    {
+      "epoch": 0.5680210789072251,
+      "grad_norm": 0.26172971725463867,
+      "learning_rate": 0.00027027138157894736,
+      "loss": 0.7982,
+      "step": 512
+    },
+    {
+      "epoch": 0.5691304950769658,
+      "grad_norm": 0.25788214802742004,
+      "learning_rate": 0.00027014802631578945,
+      "loss": 0.5565,
+      "step": 513
+    },
+    {
+      "epoch": 0.5702399112467064,
+      "grad_norm": 0.23320072889328003,
+      "learning_rate": 0.0002700246710526316,
+      "loss": 0.586,
+      "step": 514
+    },
+    {
+      "epoch": 0.5713493274164471,
+      "grad_norm": 0.1965775191783905,
+      "learning_rate": 0.0002699013157894737,
+      "loss": 0.681,
+      "step": 515
+    },
+    {
+      "epoch": 0.5724587435861878,
+      "grad_norm": 0.4177470803260803,
+      "learning_rate": 0.0002697779605263158,
+      "loss": 0.5532,
+      "step": 516
+    },
+    {
+      "epoch": 0.5735681597559285,
+      "grad_norm": 0.16181616485118866,
+      "learning_rate": 0.0002696546052631579,
+      "loss": 0.5,
+      "step": 517
+    },
+    {
+      "epoch": 0.5746775759256691,
+      "grad_norm": 0.20417065918445587,
+      "learning_rate": 0.00026953124999999997,
+      "loss": 0.5589,
+      "step": 518
+    },
+    {
+      "epoch": 0.5757869920954098,
+      "grad_norm": 0.2022491842508316,
+      "learning_rate": 0.00026940789473684206,
+      "loss": 0.5517,
+      "step": 519
+    },
+    {
+      "epoch": 0.5768964082651504,
+      "grad_norm": 0.3004019558429718,
+      "learning_rate": 0.00026928453947368415,
+      "loss": 0.46,
+      "step": 520
+    },
+    {
+      "epoch": 0.5780058244348911,
+      "grad_norm": 0.2016931027173996,
+      "learning_rate": 0.0002691611842105263,
+      "loss": 0.3938,
+      "step": 521
+    },
+    {
+      "epoch": 0.5791152406046318,
+      "grad_norm": 0.22006861865520477,
+      "learning_rate": 0.0002690378289473684,
+      "loss": 0.6507,
+      "step": 522
+    },
+    {
+      "epoch": 0.5802246567743725,
+      "grad_norm": 0.2743866741657257,
+      "learning_rate": 0.00026891447368421054,
+      "loss": 0.6233,
+      "step": 523
+    },
+    {
+      "epoch": 0.5813340729441132,
+      "grad_norm": 0.2324676811695099,
+      "learning_rate": 0.00026879111842105264,
+      "loss": 0.7202,
+      "step": 524
+    },
+    {
+      "epoch": 0.5824434891138538,
+      "grad_norm": 0.2942185401916504,
+      "learning_rate": 0.00026866776315789473,
+      "loss": 0.577,
+      "step": 525
+    },
+    {
+      "epoch": 0.5835529052835945,
+      "grad_norm": 0.20303772389888763,
+      "learning_rate": 0.0002685444078947368,
+      "loss": 0.5867,
+      "step": 526
+    },
+    {
+      "epoch": 0.5846623214533352,
+      "grad_norm": 0.3175172507762909,
+      "learning_rate": 0.0002684210526315789,
+      "loss": 0.5936,
+      "step": 527
+    },
+    {
+      "epoch": 0.5857717376230759,
+      "grad_norm": 0.26434624195098877,
+      "learning_rate": 0.000268297697368421,
+      "loss": 0.8496,
+      "step": 528
+    },
+    {
+      "epoch": 0.5868811537928166,
+      "grad_norm": 0.20476919412612915,
+      "learning_rate": 0.00026817434210526316,
+      "loss": 0.5323,
+      "step": 529
+    },
+    {
+      "epoch": 0.5879905699625572,
+      "grad_norm": 0.17890197038650513,
+      "learning_rate": 0.00026805098684210525,
+      "loss": 0.618,
+      "step": 530
+    },
+    {
+      "epoch": 0.5890999861322979,
+      "grad_norm": 0.12501509487628937,
+      "learning_rate": 0.00026792763157894734,
+      "loss": 0.3985,
+      "step": 531
+    },
+    {
+      "epoch": 0.5902094023020386,
+      "grad_norm": 0.14128711819648743,
+      "learning_rate": 0.00026780427631578944,
+      "loss": 0.3521,
+      "step": 532
+    },
+    {
+      "epoch": 0.5913188184717793,
+      "grad_norm": 0.24814924597740173,
+      "learning_rate": 0.0002676809210526316,
+      "loss": 0.5524,
+      "step": 533
+    },
+    {
+      "epoch": 0.59242823464152,
+      "grad_norm": 0.1946108043193817,
+      "learning_rate": 0.0002675575657894737,
+      "loss": 0.6405,
+      "step": 534
+    },
+    {
+      "epoch": 0.5935376508112605,
+      "grad_norm": 0.20254820585250854,
+      "learning_rate": 0.00026743421052631577,
+      "loss": 0.5621,
+      "step": 535
+    },
+    {
+      "epoch": 0.5946470669810012,
+      "grad_norm": 0.19255991280078888,
+      "learning_rate": 0.00026731085526315786,
+      "loss": 0.7144,
+      "step": 536
+    },
+    {
+      "epoch": 0.5957564831507419,
+      "grad_norm": 0.20394358038902283,
+      "learning_rate": 0.00026718749999999996,
+      "loss": 0.4172,
+      "step": 537
+    },
+    {
+      "epoch": 0.5968658993204826,
+      "grad_norm": 0.2700938582420349,
+      "learning_rate": 0.0002670641447368421,
+      "loss": 0.6688,
+      "step": 538
+    },
+    {
+      "epoch": 0.5979753154902233,
+      "grad_norm": 0.3247049152851105,
+      "learning_rate": 0.0002669407894736842,
+      "loss": 0.632,
+      "step": 539
+    },
+    {
+      "epoch": 0.5990847316599639,
+      "grad_norm": 0.2315102368593216,
+      "learning_rate": 0.0002668174342105263,
+      "loss": 0.5881,
+      "step": 540
+    },
+    {
+      "epoch": 0.6001941478297046,
+      "grad_norm": 0.28765103220939636,
+      "learning_rate": 0.0002666940789473684,
+      "loss": 0.6717,
+      "step": 541
+    },
+    {
+      "epoch": 0.6013035639994453,
+      "grad_norm": 0.24762357771396637,
+      "learning_rate": 0.00026657072368421053,
+      "loss": 0.692,
+      "step": 542
+    },
+    {
+      "epoch": 0.602412980169186,
+      "grad_norm": 0.25794705748558044,
+      "learning_rate": 0.0002664473684210526,
+      "loss": 0.5335,
+      "step": 543
+    },
+    {
+      "epoch": 0.6035223963389267,
+      "grad_norm": 0.2661387622356415,
+      "learning_rate": 0.0002663240131578947,
+      "loss": 0.7539,
+      "step": 544
+    },
+    {
+      "epoch": 0.6046318125086673,
+      "grad_norm": 0.2539893388748169,
+      "learning_rate": 0.0002662006578947368,
+      "loss": 0.4355,
+      "step": 545
+    },
+    {
+      "epoch": 0.605741228678408,
+      "grad_norm": 0.23424486815929413,
+      "learning_rate": 0.0002660773026315789,
+      "loss": 0.5143,
+      "step": 546
+    },
+    {
+      "epoch": 0.6068506448481487,
+      "grad_norm": 0.1962471306324005,
+      "learning_rate": 0.00026595394736842105,
+      "loss": 0.4722,
+      "step": 547
+    },
+    {
+      "epoch": 0.6079600610178894,
+      "grad_norm": 0.1904420107603073,
+      "learning_rate": 0.00026583059210526314,
+      "loss": 0.5893,
+      "step": 548
+    },
+    {
+      "epoch": 0.60906947718763,
+      "grad_norm": 0.3046864867210388,
+      "learning_rate": 0.00026570723684210524,
+      "loss": 0.8071,
+      "step": 549
+    },
+    {
+      "epoch": 0.6101788933573706,
+      "grad_norm": 0.1840696930885315,
+      "learning_rate": 0.00026558388157894733,
+      "loss": 0.4925,
+      "step": 550
+    },
+    {
+      "epoch": 0.6112883095271113,
+      "grad_norm": 0.33538711071014404,
+      "learning_rate": 0.0002654605263157894,
+      "loss": 0.6222,
+      "step": 551
+    },
+    {
+      "epoch": 0.612397725696852,
+      "grad_norm": 0.22219829261302948,
+      "learning_rate": 0.00026533717105263157,
+      "loss": 0.4808,
+      "step": 552
+    },
+    {
+      "epoch": 0.6135071418665927,
+      "grad_norm": 0.3794260621070862,
+      "learning_rate": 0.00026521381578947366,
+      "loss": 0.4464,
+      "step": 553
+    },
+    {
+      "epoch": 0.6146165580363334,
+      "grad_norm": 0.21123401820659637,
+      "learning_rate": 0.00026509046052631576,
+      "loss": 0.6366,
+      "step": 554
+    },
+    {
+      "epoch": 0.615725974206074,
+      "grad_norm": 0.19497540593147278,
+      "learning_rate": 0.00026496710526315785,
+      "loss": 0.5467,
+      "step": 555
+    },
+    {
+      "epoch": 0.6168353903758147,
+      "grad_norm": 0.18902145326137543,
+      "learning_rate": 0.00026484375,
+      "loss": 0.4334,
+      "step": 556
+    },
+    {
+      "epoch": 0.6179448065455554,
+      "grad_norm": 0.24114537239074707,
+      "learning_rate": 0.0002647203947368421,
+      "loss": 0.5268,
+      "step": 557
+    },
+    {
+      "epoch": 0.6190542227152961,
+      "grad_norm": 0.18477365374565125,
+      "learning_rate": 0.0002645970394736842,
+      "loss": 0.6423,
+      "step": 558
+    },
+    {
+      "epoch": 0.6201636388850368,
+      "grad_norm": 0.21324200928211212,
+      "learning_rate": 0.0002644736842105263,
+      "loss": 0.6366,
+      "step": 559
+    },
+    {
+      "epoch": 0.6212730550547775,
+      "grad_norm": 0.38751551508903503,
+      "learning_rate": 0.00026435032894736837,
+      "loss": 0.6193,
+      "step": 560
+    },
+    {
+      "epoch": 0.6223824712245181,
+      "grad_norm": 0.22451990842819214,
+      "learning_rate": 0.0002642269736842105,
+      "loss": 0.6747,
+      "step": 561
+    },
+    {
+      "epoch": 0.6234918873942588,
+      "grad_norm": 0.2260679006576538,
+      "learning_rate": 0.0002641036184210526,
+      "loss": 0.8031,
+      "step": 562
+    },
+    {
+      "epoch": 0.6246013035639995,
+      "grad_norm": 0.22427742183208466,
+      "learning_rate": 0.0002639802631578947,
+      "loss": 0.7171,
+      "step": 563
+    },
+    {
+      "epoch": 0.6257107197337402,
+      "grad_norm": 0.20108933746814728,
+      "learning_rate": 0.00026385690789473685,
+      "loss": 0.4898,
+      "step": 564
+    },
+    {
+      "epoch": 0.6268201359034808,
+      "grad_norm": 0.3800278604030609,
+      "learning_rate": 0.00026373355263157894,
+      "loss": 0.6466,
+      "step": 565
+    },
+    {
+      "epoch": 0.6279295520732214,
+      "grad_norm": 0.22784464061260223,
+      "learning_rate": 0.00026361019736842104,
+      "loss": 0.5532,
+      "step": 566
+    },
+    {
+      "epoch": 0.6290389682429621,
+      "grad_norm": 0.23498325049877167,
+      "learning_rate": 0.00026348684210526313,
+      "loss": 0.8687,
+      "step": 567
+    },
+    {
+      "epoch": 0.6301483844127028,
+      "grad_norm": 0.1839025616645813,
+      "learning_rate": 0.0002633634868421052,
+      "loss": 0.6382,
+      "step": 568
+    },
+    {
+      "epoch": 0.6312578005824435,
+      "grad_norm": 0.22980616986751556,
+      "learning_rate": 0.0002632401315789473,
+      "loss": 0.6109,
+      "step": 569
+    },
+    {
+      "epoch": 0.6323672167521842,
+      "grad_norm": 0.17458495497703552,
+      "learning_rate": 0.0002631167763157894,
+      "loss": 0.5498,
+      "step": 570
+    },
+    {
+      "epoch": 0.6334766329219248,
+      "grad_norm": 0.22085556387901306,
+      "learning_rate": 0.00026299342105263156,
+      "loss": 0.5102,
+      "step": 571
+    },
+    {
+      "epoch": 0.6345860490916655,
+      "grad_norm": 0.3213456869125366,
+      "learning_rate": 0.00026287006578947365,
+      "loss": 0.7377,
+      "step": 572
+    },
+    {
+      "epoch": 0.6356954652614062,
+      "grad_norm": 0.2649673819541931,
+      "learning_rate": 0.0002627467105263158,
+      "loss": 0.6664,
+      "step": 573
+    },
+    {
+      "epoch": 0.6368048814311469,
+      "grad_norm": 0.3784686326980591,
+      "learning_rate": 0.0002626233552631579,
+      "loss": 0.9214,
+      "step": 574
+    },
+    {
+      "epoch": 0.6379142976008876,
+      "grad_norm": 0.1708430051803589,
+      "learning_rate": 0.0002625,
+      "loss": 0.4383,
+      "step": 575
+    },
+    {
+      "epoch": 0.6390237137706282,
+      "grad_norm": 0.26163679361343384,
+      "learning_rate": 0.0002623766447368421,
+      "loss": 0.8272,
+      "step": 576
+    },
+    {
+      "epoch": 0.6401331299403689,
+      "grad_norm": 0.16542355716228485,
+      "learning_rate": 0.00026225328947368417,
+      "loss": 0.4739,
+      "step": 577
+    },
+    {
+      "epoch": 0.6412425461101096,
+      "grad_norm": 0.2180267572402954,
+      "learning_rate": 0.00026212993421052626,
+      "loss": 0.6221,
+      "step": 578
+    },
+    {
+      "epoch": 0.6423519622798503,
+      "grad_norm": 0.21957628428936005,
+      "learning_rate": 0.0002620065789473684,
+      "loss": 0.6062,
+      "step": 579
+    },
+    {
+      "epoch": 0.643461378449591,
+      "grad_norm": 0.20948325097560883,
+      "learning_rate": 0.0002618832236842105,
+      "loss": 0.5542,
+      "step": 580
+    },
+    {
+      "epoch": 0.6445707946193315,
+      "grad_norm": 0.28024452924728394,
+      "learning_rate": 0.0002617598684210526,
+      "loss": 0.718,
+      "step": 581
+    },
+    {
+      "epoch": 0.6456802107890722,
+      "grad_norm": 0.1376865804195404,
+      "learning_rate": 0.00026163651315789474,
+      "loss": 0.4168,
+      "step": 582
+    },
+    {
+      "epoch": 0.6467896269588129,
+      "grad_norm": 0.15050861239433289,
+      "learning_rate": 0.00026151315789473684,
+      "loss": 0.5308,
+      "step": 583
+    },
+    {
+      "epoch": 0.6478990431285536,
+      "grad_norm": 0.21709243953227997,
+      "learning_rate": 0.00026138980263157893,
+      "loss": 0.677,
+      "step": 584
+    },
+    {
+      "epoch": 0.6490084592982943,
+      "grad_norm": 0.37020203471183777,
+      "learning_rate": 0.000261266447368421,
+      "loss": 0.6208,
+      "step": 585
+    },
+    {
+      "epoch": 0.6501178754680349,
+      "grad_norm": 0.2256883978843689,
+      "learning_rate": 0.0002611430921052631,
+      "loss": 0.7711,
+      "step": 586
+    },
+    {
+      "epoch": 0.6512272916377756,
+      "grad_norm": 0.22099201381206512,
+      "learning_rate": 0.0002610197368421052,
+      "loss": 0.6091,
+      "step": 587
+    },
+    {
+      "epoch": 0.6523367078075163,
+      "grad_norm": 0.23496872186660767,
+      "learning_rate": 0.00026089638157894736,
+      "loss": 0.2471,
+      "step": 588
+    },
+    {
+      "epoch": 0.653446123977257,
+      "grad_norm": 0.2800827622413635,
+      "learning_rate": 0.00026077302631578945,
+      "loss": 0.5343,
+      "step": 589
+    },
+    {
+      "epoch": 0.6545555401469977,
+      "grad_norm": 0.22502388060092926,
+      "learning_rate": 0.00026064967105263154,
+      "loss": 0.5117,
+      "step": 590
+    },
+    {
+      "epoch": 0.6556649563167383,
+      "grad_norm": 0.1460188329219818,
+      "learning_rate": 0.0002605263157894737,
+      "loss": 0.4772,
+      "step": 591
+    },
+    {
+      "epoch": 0.656774372486479,
+      "grad_norm": 0.16591776907444,
+      "learning_rate": 0.0002604029605263158,
+      "loss": 0.6087,
+      "step": 592
+    },
+    {
+      "epoch": 0.6578837886562197,
+      "grad_norm": 0.13937248289585114,
+      "learning_rate": 0.0002602796052631579,
+      "loss": 0.5587,
+      "step": 593
+    },
+    {
+      "epoch": 0.6589932048259604,
+      "grad_norm": 0.2956066131591797,
+      "learning_rate": 0.00026015624999999997,
+      "loss": 0.7117,
+      "step": 594
+    },
+    {
+      "epoch": 0.660102620995701,
+      "grad_norm": 0.495911568403244,
+      "learning_rate": 0.00026003289473684206,
+      "loss": 0.8364,
+      "step": 595
+    },
+    {
+      "epoch": 0.6612120371654416,
+      "grad_norm": 0.1725756675004959,
+      "learning_rate": 0.00025990953947368416,
+      "loss": 0.4979,
+      "step": 596
+    },
+    {
+      "epoch": 0.6623214533351823,
+      "grad_norm": 0.25791987776756287,
+      "learning_rate": 0.0002597861842105263,
+      "loss": 0.7304,
+      "step": 597
+    },
+    {
+      "epoch": 0.663430869504923,
+      "grad_norm": 0.18017613887786865,
+      "learning_rate": 0.0002596628289473684,
+      "loss": 0.5706,
+      "step": 598
+    },
+    {
+      "epoch": 0.6645402856746637,
+      "grad_norm": 0.25128671526908875,
+      "learning_rate": 0.00025953947368421055,
+      "loss": 0.5846,
+      "step": 599
+    },
+    {
+      "epoch": 0.6656497018444044,
+      "grad_norm": 0.34931543469429016,
+      "learning_rate": 0.00025941611842105264,
+      "loss": 0.7017,
+      "step": 600
+    },
+    {
+      "epoch": 0.666759118014145,
+      "grad_norm": 0.25030258297920227,
+      "learning_rate": 0.00025929276315789473,
+      "loss": 0.5193,
+      "step": 601
+    },
+    {
+      "epoch": 0.6678685341838857,
+      "grad_norm": 0.236861452460289,
+      "learning_rate": 0.0002591694078947368,
+      "loss": 0.6901,
+      "step": 602
+    },
+    {
+      "epoch": 0.6689779503536264,
+      "grad_norm": 0.308292418718338,
+      "learning_rate": 0.0002590460526315789,
+      "loss": 0.4285,
+      "step": 603
+    },
+    {
+      "epoch": 0.6700873665233671,
+      "grad_norm": 0.2141687422990799,
+      "learning_rate": 0.000258922697368421,
+      "loss": 0.3857,
+      "step": 604
+    },
+    {
+      "epoch": 0.6711967826931078,
+      "grad_norm": 0.164393350481987,
+      "learning_rate": 0.0002587993421052631,
+      "loss": 0.5326,
+      "step": 605
+    },
+    {
+      "epoch": 0.6723061988628485,
+      "grad_norm": 0.30191662907600403,
+      "learning_rate": 0.00025867598684210525,
+      "loss": 0.4978,
+      "step": 606
+    },
+    {
+      "epoch": 0.6734156150325891,
+      "grad_norm": 0.2955259382724762,
+      "learning_rate": 0.00025855263157894734,
+      "loss": 0.5253,
+      "step": 607
+    },
+    {
+      "epoch": 0.6745250312023298,
+      "grad_norm": 0.22022663056850433,
+      "learning_rate": 0.00025842927631578944,
+      "loss": 0.7104,
+      "step": 608
+    },
+    {
+      "epoch": 0.6756344473720705,
+      "grad_norm": 0.21236523985862732,
+      "learning_rate": 0.0002583059210526316,
+      "loss": 0.5324,
+      "step": 609
+    },
+    {
+      "epoch": 0.6767438635418112,
+      "grad_norm": 0.1945660263299942,
+      "learning_rate": 0.0002581825657894737,
+      "loss": 0.5858,
+      "step": 610
+    },
+    {
+      "epoch": 0.6778532797115518,
+      "grad_norm": 0.31970614194869995,
+      "learning_rate": 0.00025805921052631577,
+      "loss": 0.7485,
+      "step": 611
+    },
+    {
+      "epoch": 0.6789626958812924,
+      "grad_norm": 0.3211202621459961,
+      "learning_rate": 0.00025793585526315786,
+      "loss": 0.5554,
+      "step": 612
+    },
+    {
+      "epoch": 0.6800721120510331,
+      "grad_norm": 0.17990931868553162,
+      "learning_rate": 0.00025781249999999996,
+      "loss": 0.6221,
+      "step": 613
+    },
+    {
+      "epoch": 0.6811815282207738,
+      "grad_norm": 0.3069283366203308,
+      "learning_rate": 0.0002576891447368421,
+      "loss": 0.7163,
+      "step": 614
+    },
+    {
+      "epoch": 0.6822909443905145,
+      "grad_norm": 0.19691799581050873,
+      "learning_rate": 0.0002575657894736842,
+      "loss": 0.4445,
+      "step": 615
+    },
+    {
+      "epoch": 0.6834003605602552,
+      "grad_norm": 0.18806682527065277,
+      "learning_rate": 0.0002574424342105263,
+      "loss": 0.5781,
+      "step": 616
+    },
+    {
+      "epoch": 0.6845097767299958,
+      "grad_norm": 0.24056103825569153,
+      "learning_rate": 0.0002573190789473684,
+      "loss": 0.5351,
+      "step": 617
+    },
+    {
+      "epoch": 0.6856191928997365,
+      "grad_norm": 0.2140192836523056,
+      "learning_rate": 0.00025719572368421053,
+      "loss": 0.6222,
+      "step": 618
+    },
+    {
+      "epoch": 0.6867286090694772,
+      "grad_norm": 0.227885901927948,
+      "learning_rate": 0.0002570723684210526,
+      "loss": 0.6394,
+      "step": 619
+    },
+    {
+      "epoch": 0.6878380252392179,
+      "grad_norm": 0.37848934531211853,
+      "learning_rate": 0.0002569490131578947,
+      "loss": 0.6231,
+      "step": 620
+    },
+    {
+      "epoch": 0.6889474414089586,
+      "grad_norm": 0.290159672498703,
+      "learning_rate": 0.0002568256578947368,
+      "loss": 0.4614,
+      "step": 621
+    },
+    {
+      "epoch": 0.6900568575786992,
+      "grad_norm": 0.18309064209461212,
+      "learning_rate": 0.0002567023026315789,
+      "loss": 0.7418,
+      "step": 622
+    },
+    {
+      "epoch": 0.6911662737484399,
+      "grad_norm": 0.20930887758731842,
+      "learning_rate": 0.00025657894736842105,
+      "loss": 0.6572,
+      "step": 623
+    },
+    {
+      "epoch": 0.6922756899181806,
+      "grad_norm": 0.24094976484775543,
+      "learning_rate": 0.00025645559210526315,
+      "loss": 0.5702,
+      "step": 624
+    },
+    {
+      "epoch": 0.6933851060879213,
+      "grad_norm": 0.22989119589328766,
+      "learning_rate": 0.00025633223684210524,
+      "loss": 0.4517,
+      "step": 625
+    },
+    {
+      "epoch": 0.694494522257662,
+      "grad_norm": 0.2922836244106293,
+      "learning_rate": 0.00025620888157894733,
+      "loss": 0.6436,
+      "step": 626
+    },
+    {
+      "epoch": 0.6956039384274025,
+      "grad_norm": 0.2564910650253296,
+      "learning_rate": 0.0002560855263157894,
+      "loss": 0.7321,
+      "step": 627
+    },
+    {
+      "epoch": 0.6967133545971432,
+      "grad_norm": 0.26571327447891235,
+      "learning_rate": 0.00025596217105263157,
+      "loss": 0.4964,
+      "step": 628
+    },
+    {
+      "epoch": 0.6978227707668839,
+      "grad_norm": 0.5190631151199341,
+      "learning_rate": 0.00025583881578947367,
+      "loss": 0.7778,
+      "step": 629
+    },
+    {
+      "epoch": 0.6989321869366246,
+      "grad_norm": 0.17522084712982178,
+      "learning_rate": 0.00025571546052631576,
+      "loss": 0.4768,
+      "step": 630
+    },
+    {
+      "epoch": 0.7000416031063653,
+      "grad_norm": 0.2567191421985626,
+      "learning_rate": 0.00025559210526315785,
+      "loss": 0.5936,
+      "step": 631
+    },
+    {
+      "epoch": 0.7011510192761059,
+      "grad_norm": 0.46300792694091797,
+      "learning_rate": 0.00025546875,
+      "loss": 0.6074,
+      "step": 632
+    },
+    {
+      "epoch": 0.7022604354458466,
+      "grad_norm": 0.1528376042842865,
+      "learning_rate": 0.0002553453947368421,
+      "loss": 0.4779,
+      "step": 633
+    },
+    {
+      "epoch": 0.7033698516155873,
+      "grad_norm": 0.3135516941547394,
+      "learning_rate": 0.0002552220394736842,
+      "loss": 0.8674,
+      "step": 634
+    },
+    {
+      "epoch": 0.704479267785328,
+      "grad_norm": 0.22676752507686615,
+      "learning_rate": 0.0002550986842105263,
+      "loss": 0.5182,
+      "step": 635
+    },
+    {
+      "epoch": 0.7055886839550687,
+      "grad_norm": 0.21783347427845,
+      "learning_rate": 0.00025497532894736837,
+      "loss": 0.4951,
+      "step": 636
+    },
+    {
+      "epoch": 0.7066981001248093,
+      "grad_norm": 0.2860846221446991,
+      "learning_rate": 0.0002548519736842105,
+      "loss": 0.4363,
+      "step": 637
+    },
+    {
+      "epoch": 0.70780751629455,
+      "grad_norm": 0.281086266040802,
+      "learning_rate": 0.0002547286184210526,
+      "loss": 0.5309,
+      "step": 638
+    },
+    {
+      "epoch": 0.7089169324642907,
+      "grad_norm": 0.30188825726509094,
+      "learning_rate": 0.0002546052631578947,
+      "loss": 0.5422,
+      "step": 639
+    },
+    {
+      "epoch": 0.7100263486340314,
+      "grad_norm": 0.26086941361427307,
+      "learning_rate": 0.0002544819078947368,
+      "loss": 0.4605,
+      "step": 640
+    },
+    {
+      "epoch": 0.711135764803772,
+      "grad_norm": 0.3494928777217865,
+      "learning_rate": 0.00025435855263157895,
+      "loss": 0.4638,
+      "step": 641
+    },
+    {
+      "epoch": 0.7122451809735126,
+      "grad_norm": 0.282701313495636,
+      "learning_rate": 0.00025423519736842104,
+      "loss": 0.6283,
+      "step": 642
+    },
+    {
+      "epoch": 0.7133545971432533,
+      "grad_norm": 0.16015778481960297,
+      "learning_rate": 0.00025411184210526313,
+      "loss": 0.4514,
+      "step": 643
+    },
+    {
+      "epoch": 0.714464013312994,
+      "grad_norm": 0.2207580804824829,
+      "learning_rate": 0.0002539884868421052,
+      "loss": 0.7127,
+      "step": 644
+    },
+    {
+      "epoch": 0.7155734294827347,
+      "grad_norm": 0.2917775511741638,
+      "learning_rate": 0.0002538651315789473,
+      "loss": 0.7254,
+      "step": 645
+    },
+    {
+      "epoch": 0.7166828456524754,
+      "grad_norm": 0.26504382491111755,
+      "learning_rate": 0.0002537417763157894,
+      "loss": 0.5335,
+      "step": 646
+    },
+    {
+      "epoch": 0.717792261822216,
+      "grad_norm": 0.3495447635650635,
+      "learning_rate": 0.00025361842105263156,
+      "loss": 0.8425,
+      "step": 647
+    },
+    {
+      "epoch": 0.7189016779919567,
+      "grad_norm": 0.18636609613895416,
+      "learning_rate": 0.00025349506578947365,
+      "loss": 0.5437,
+      "step": 648
+    },
+    {
+      "epoch": 0.7200110941616974,
+      "grad_norm": 0.18877021968364716,
+      "learning_rate": 0.0002533717105263158,
+      "loss": 0.5643,
+      "step": 649
+    },
+    {
+      "epoch": 0.7211205103314381,
+      "grad_norm": 0.15186652541160583,
+      "learning_rate": 0.0002532483552631579,
+      "loss": 0.532,
+      "step": 650
+    },
+    {
+      "epoch": 0.7222299265011788,
+      "grad_norm": 0.14779016375541687,
+      "learning_rate": 0.000253125,
+      "loss": 0.566,
+      "step": 651
+    },
+    {
+      "epoch": 0.7233393426709195,
+      "grad_norm": 0.19088061153888702,
+      "learning_rate": 0.0002530016447368421,
+      "loss": 0.5351,
+      "step": 652
+    },
+    {
+      "epoch": 0.7244487588406601,
+      "grad_norm": 0.17555399239063263,
+      "learning_rate": 0.00025287828947368417,
+      "loss": 0.4758,
+      "step": 653
+    },
+    {
+      "epoch": 0.7255581750104008,
+      "grad_norm": 0.2535383999347687,
+      "learning_rate": 0.00025275493421052627,
+      "loss": 0.6948,
+      "step": 654
+    },
+    {
+      "epoch": 0.7266675911801415,
+      "grad_norm": 0.2938152551651001,
+      "learning_rate": 0.00025263157894736836,
+      "loss": 0.626,
+      "step": 655
+    },
+    {
+      "epoch": 0.7277770073498822,
+      "grad_norm": 0.2159254252910614,
+      "learning_rate": 0.0002525082236842105,
+      "loss": 0.471,
+      "step": 656
+    },
+    {
+      "epoch": 0.7288864235196229,
+      "grad_norm": 0.20253120362758636,
+      "learning_rate": 0.0002523848684210526,
+      "loss": 0.4693,
+      "step": 657
+    },
+    {
+      "epoch": 0.7299958396893634,
+      "grad_norm": 0.49963316321372986,
+      "learning_rate": 0.00025226151315789475,
+      "loss": 0.5185,
+      "step": 658
+    },
+    {
+      "epoch": 0.7311052558591041,
+      "grad_norm": 0.2259654551744461,
+      "learning_rate": 0.00025213815789473684,
+      "loss": 0.7249,
+      "step": 659
+    },
+    {
+      "epoch": 0.7322146720288448,
+      "grad_norm": 0.21344606578350067,
+      "learning_rate": 0.00025201480263157893,
+      "loss": 0.4977,
+      "step": 660
+    },
+    {
+      "epoch": 0.7333240881985855,
+      "grad_norm": 0.2689608037471771,
+      "learning_rate": 0.000251891447368421,
+      "loss": 0.3908,
+      "step": 661
+    },
+    {
+      "epoch": 0.7344335043683262,
+      "grad_norm": 0.18120594322681427,
+      "learning_rate": 0.0002517680921052631,
+      "loss": 0.4518,
+      "step": 662
+    },
+    {
+      "epoch": 0.7355429205380668,
+      "grad_norm": 0.3393332064151764,
+      "learning_rate": 0.0002516447368421052,
+      "loss": 0.6439,
+      "step": 663
+    },
+    {
+      "epoch": 0.7366523367078075,
+      "grad_norm": 0.21560847759246826,
+      "learning_rate": 0.00025152138157894736,
+      "loss": 0.6041,
+      "step": 664
+    },
+    {
+      "epoch": 0.7377617528775482,
+      "grad_norm": 0.4047819972038269,
+      "learning_rate": 0.00025139802631578945,
+      "loss": 0.7444,
+      "step": 665
+    },
+    {
+      "epoch": 0.7388711690472889,
+      "grad_norm": 0.17038540542125702,
+      "learning_rate": 0.00025127467105263155,
+      "loss": 0.4562,
+      "step": 666
+    },
+    {
+      "epoch": 0.7399805852170296,
+      "grad_norm": 0.1971050649881363,
+      "learning_rate": 0.0002511513157894737,
+      "loss": 0.4497,
+      "step": 667
+    },
+    {
+      "epoch": 0.7410900013867702,
+      "grad_norm": 0.233141228556633,
+      "learning_rate": 0.0002510279605263158,
+      "loss": 0.6446,
+      "step": 668
+    },
+    {
+      "epoch": 0.7421994175565109,
+      "grad_norm": 0.22134460508823395,
+      "learning_rate": 0.0002509046052631579,
+      "loss": 0.4393,
+      "step": 669
+    },
+    {
+      "epoch": 0.7433088337262516,
+      "grad_norm": 0.18341104686260223,
+      "learning_rate": 0.00025078125,
+      "loss": 0.6756,
+      "step": 670
+    },
+    {
+      "epoch": 0.7444182498959923,
+      "grad_norm": 0.2109827697277069,
+      "learning_rate": 0.00025065789473684207,
+      "loss": 0.5686,
+      "step": 671
+    },
+    {
+      "epoch": 0.745527666065733,
+      "grad_norm": 0.249485045671463,
+      "learning_rate": 0.00025053453947368416,
+      "loss": 0.5849,
+      "step": 672
+    },
+    {
+      "epoch": 0.7466370822354735,
+      "grad_norm": 0.36929988861083984,
+      "learning_rate": 0.0002504111842105263,
+      "loss": 0.5873,
+      "step": 673
+    },
+    {
+      "epoch": 0.7477464984052142,
+      "grad_norm": 0.2066950798034668,
+      "learning_rate": 0.0002502878289473684,
+      "loss": 0.4152,
+      "step": 674
+    },
+    {
+      "epoch": 0.7488559145749549,
+      "grad_norm": 0.14233893156051636,
+      "learning_rate": 0.0002501644736842105,
+      "loss": 0.4972,
+      "step": 675
+    },
+    {
+      "epoch": 0.7499653307446956,
+      "grad_norm": 0.24324600398540497,
+      "learning_rate": 0.00025004111842105264,
+      "loss": 0.4693,
+      "step": 676
+    },
+    {
+      "epoch": 0.7510747469144363,
+      "grad_norm": 0.18247400224208832,
+      "learning_rate": 0.00024991776315789473,
+      "loss": 0.6579,
+      "step": 677
+    },
+    {
+      "epoch": 0.7521841630841769,
+      "grad_norm": 0.20285794138908386,
+      "learning_rate": 0.00024979440789473683,
+      "loss": 0.3657,
+      "step": 678
+    },
+    {
+      "epoch": 0.7532935792539176,
+      "grad_norm": 0.2386598140001297,
+      "learning_rate": 0.0002496710526315789,
+      "loss": 0.5693,
+      "step": 679
+    },
+    {
+      "epoch": 0.7544029954236583,
+      "grad_norm": 0.28417715430259705,
+      "learning_rate": 0.000249547697368421,
+      "loss": 0.5416,
+      "step": 680
+    },
+    {
+      "epoch": 0.755512411593399,
+      "grad_norm": 0.4812435805797577,
+      "learning_rate": 0.0002494243421052631,
+      "loss": 0.6869,
+      "step": 681
+    },
+    {
+      "epoch": 0.7566218277631397,
+      "grad_norm": 0.1943156123161316,
+      "learning_rate": 0.00024930098684210525,
+      "loss": 0.4625,
+      "step": 682
+    },
+    {
+      "epoch": 0.7577312439328803,
+      "grad_norm": 0.22697441279888153,
+      "learning_rate": 0.00024917763157894735,
+      "loss": 0.4708,
+      "step": 683
+    },
+    {
+      "epoch": 0.758840660102621,
+      "grad_norm": 0.32567355036735535,
+      "learning_rate": 0.00024905427631578944,
+      "loss": 0.4607,
+      "step": 684
+    },
+    {
+      "epoch": 0.7599500762723617,
+      "grad_norm": 0.2505040168762207,
+      "learning_rate": 0.0002489309210526316,
+      "loss": 0.5327,
+      "step": 685
+    },
+    {
+      "epoch": 0.7610594924421024,
+      "grad_norm": 0.19925042986869812,
+      "learning_rate": 0.0002488075657894737,
+      "loss": 0.7429,
+      "step": 686
+    },
+    {
+      "epoch": 0.762168908611843,
+      "grad_norm": 0.19456201791763306,
+      "learning_rate": 0.0002486842105263158,
+      "loss": 0.6922,
+      "step": 687
+    },
+    {
+      "epoch": 0.7632783247815836,
+      "grad_norm": 0.2642272412776947,
+      "learning_rate": 0.00024856085526315787,
+      "loss": 0.5104,
+      "step": 688
+    },
+    {
+      "epoch": 0.7643877409513243,
+      "grad_norm": 0.22339658439159393,
+      "learning_rate": 0.00024843749999999996,
+      "loss": 0.5151,
+      "step": 689
+    },
+    {
+      "epoch": 0.765497157121065,
+      "grad_norm": 0.23145624995231628,
+      "learning_rate": 0.00024831414473684205,
+      "loss": 0.4034,
+      "step": 690
+    },
+    {
+      "epoch": 0.7666065732908057,
+      "grad_norm": 0.20969951152801514,
+      "learning_rate": 0.0002481907894736842,
+      "loss": 0.5701,
+      "step": 691
+    },
+    {
+      "epoch": 0.7677159894605464,
+      "grad_norm": 0.2449328750371933,
+      "learning_rate": 0.0002480674342105263,
+      "loss": 0.4631,
+      "step": 692
+    },
+    {
+      "epoch": 0.768825405630287,
+      "grad_norm": 0.2677520513534546,
+      "learning_rate": 0.0002479440789473684,
+      "loss": 0.5099,
+      "step": 693
+    },
+    {
+      "epoch": 0.7699348218000277,
+      "grad_norm": 0.21626238524913788,
+      "learning_rate": 0.00024782072368421053,
+      "loss": 0.5724,
+      "step": 694
+    },
+    {
+      "epoch": 0.7710442379697684,
+      "grad_norm": 0.2530820071697235,
+      "learning_rate": 0.00024769736842105263,
+      "loss": 0.4764,
+      "step": 695
+    },
+    {
+      "epoch": 0.7721536541395091,
+      "grad_norm": 0.14730204641819,
+      "learning_rate": 0.0002475740131578947,
+      "loss": 0.4657,
+      "step": 696
+    },
+    {
+      "epoch": 0.7732630703092498,
+      "grad_norm": 0.21123374998569489,
+      "learning_rate": 0.0002474506578947368,
+      "loss": 0.6438,
+      "step": 697
+    },
+    {
+      "epoch": 0.7743724864789905,
+      "grad_norm": 0.2234024703502655,
+      "learning_rate": 0.0002473273026315789,
+      "loss": 0.7459,
+      "step": 698
+    },
+    {
+      "epoch": 0.7754819026487311,
+      "grad_norm": 0.25458112359046936,
+      "learning_rate": 0.00024720394736842105,
+      "loss": 0.5767,
+      "step": 699
+    },
+    {
+      "epoch": 0.7765913188184718,
+      "grad_norm": 0.19247955083847046,
+      "learning_rate": 0.00024708059210526315,
+      "loss": 0.4738,
+      "step": 700
+    },
+    {
+      "epoch": 0.7777007349882125,
+      "grad_norm": 0.3082413673400879,
+      "learning_rate": 0.00024695723684210524,
+      "loss": 0.5998,
+      "step": 701
+    },
+    {
+      "epoch": 0.7788101511579532,
+      "grad_norm": 0.21955102682113647,
+      "learning_rate": 0.00024683388157894733,
+      "loss": 0.6763,
+      "step": 702
+    },
+    {
+      "epoch": 0.7799195673276939,
+      "grad_norm": 0.21807517111301422,
+      "learning_rate": 0.00024671052631578943,
+      "loss": 0.6072,
+      "step": 703
+    },
+    {
+      "epoch": 0.7810289834974344,
+      "grad_norm": 0.2064640372991562,
+      "learning_rate": 0.0002465871710526316,
+      "loss": 0.5148,
+      "step": 704
+    },
+    {
+      "epoch": 0.7821383996671751,
+      "grad_norm": 0.19100461900234222,
+      "learning_rate": 0.00024646381578947367,
+      "loss": 0.3789,
+      "step": 705
+    },
+    {
+      "epoch": 0.7832478158369158,
+      "grad_norm": 0.4141659140586853,
+      "learning_rate": 0.00024634046052631576,
+      "loss": 0.5838,
+      "step": 706
+    },
+    {
+      "epoch": 0.7843572320066565,
+      "grad_norm": 0.260050892829895,
+      "learning_rate": 0.00024621710526315785,
+      "loss": 0.6696,
+      "step": 707
+    },
+    {
+      "epoch": 0.7854666481763972,
+      "grad_norm": 0.22340181469917297,
+      "learning_rate": 0.00024609375,
+      "loss": 0.7745,
+      "step": 708
+    },
+    {
+      "epoch": 0.7865760643461378,
+      "grad_norm": 0.23725625872612,
+      "learning_rate": 0.0002459703947368421,
+      "loss": 0.5473,
+      "step": 709
+    },
+    {
+      "epoch": 0.7876854805158785,
+      "grad_norm": 0.2455572783946991,
+      "learning_rate": 0.0002458470394736842,
+      "loss": 0.5893,
+      "step": 710
+    },
+    {
+      "epoch": 0.7887948966856192,
+      "grad_norm": 0.2484428882598877,
+      "learning_rate": 0.0002457236842105263,
+      "loss": 0.6448,
+      "step": 711
+    },
+    {
+      "epoch": 0.7899043128553599,
+      "grad_norm": 0.1755114644765854,
+      "learning_rate": 0.0002456003289473684,
+      "loss": 0.5489,
+      "step": 712
+    },
+    {
+      "epoch": 0.7910137290251006,
+      "grad_norm": 0.18849492073059082,
+      "learning_rate": 0.0002454769736842105,
+      "loss": 0.5535,
+      "step": 713
+    },
+    {
+      "epoch": 0.7921231451948412,
+      "grad_norm": 0.2751491665840149,
+      "learning_rate": 0.0002453536184210526,
+      "loss": 0.6029,
+      "step": 714
+    },
+    {
+      "epoch": 0.7932325613645819,
+      "grad_norm": 0.2292255163192749,
+      "learning_rate": 0.0002452302631578947,
+      "loss": 0.452,
+      "step": 715
+    },
+    {
+      "epoch": 0.7943419775343226,
+      "grad_norm": 0.2448405623435974,
+      "learning_rate": 0.0002451069078947368,
+      "loss": 0.6581,
+      "step": 716
+    },
+    {
+      "epoch": 0.7954513937040633,
+      "grad_norm": 0.22489365935325623,
+      "learning_rate": 0.00024498355263157895,
+      "loss": 0.3962,
+      "step": 717
+    },
+    {
+      "epoch": 0.796560809873804,
+      "grad_norm": 0.31602340936660767,
+      "learning_rate": 0.00024486019736842104,
+      "loss": 0.6768,
+      "step": 718
+    },
+    {
+      "epoch": 0.7976702260435445,
+      "grad_norm": 0.3364429175853729,
+      "learning_rate": 0.00024473684210526314,
+      "loss": 0.7009,
+      "step": 719
+    },
+    {
+      "epoch": 0.7987796422132852,
+      "grad_norm": 0.2554149031639099,
+      "learning_rate": 0.00024461348684210523,
+      "loss": 0.7788,
+      "step": 720
+    },
+    {
+      "epoch": 0.7998890583830259,
+      "grad_norm": 0.3130899667739868,
+      "learning_rate": 0.0002444901315789473,
+      "loss": 0.698,
+      "step": 721
+    },
+    {
+      "epoch": 0.8009984745527666,
+      "grad_norm": 0.17528441548347473,
+      "learning_rate": 0.0002443667763157894,
+      "loss": 0.5581,
+      "step": 722
+    },
+    {
+      "epoch": 0.8021078907225073,
+      "grad_norm": 0.18897710740566254,
+      "learning_rate": 0.00024424342105263156,
+      "loss": 0.4447,
+      "step": 723
+    },
+    {
+      "epoch": 0.8032173068922479,
+      "grad_norm": 0.2809373438358307,
+      "learning_rate": 0.00024412006578947368,
+      "loss": 0.4655,
+      "step": 724
+    },
+    {
+      "epoch": 0.8043267230619886,
+      "grad_norm": 0.16998441517353058,
+      "learning_rate": 0.00024399671052631578,
+      "loss": 0.568,
+      "step": 725
+    },
+    {
+      "epoch": 0.8054361392317293,
+      "grad_norm": 0.21042829751968384,
+      "learning_rate": 0.00024387335526315787,
+      "loss": 0.581,
+      "step": 726
+    },
+    {
+      "epoch": 0.80654555540147,
+      "grad_norm": 0.18260133266448975,
+      "learning_rate": 0.00024375,
+      "loss": 0.4479,
+      "step": 727
+    },
+    {
+      "epoch": 0.8076549715712107,
+      "grad_norm": 0.323022723197937,
+      "learning_rate": 0.00024362664473684208,
+      "loss": 0.6448,
+      "step": 728
+    },
+    {
+      "epoch": 0.8087643877409513,
+      "grad_norm": 0.2532772123813629,
+      "learning_rate": 0.00024350328947368418,
+      "loss": 0.6414,
+      "step": 729
+    },
+    {
+      "epoch": 0.809873803910692,
+      "grad_norm": 0.186369851231575,
+      "learning_rate": 0.0002433799342105263,
+      "loss": 0.5935,
+      "step": 730
+    },
+    {
+      "epoch": 0.8109832200804327,
+      "grad_norm": 0.22981056571006775,
+      "learning_rate": 0.0002432565789473684,
+      "loss": 0.4538,
+      "step": 731
+    },
+    {
+      "epoch": 0.8120926362501734,
+      "grad_norm": 0.2305302619934082,
+      "learning_rate": 0.0002431332236842105,
+      "loss": 0.3573,
+      "step": 732
+    },
+    {
+      "epoch": 0.8132020524199141,
+      "grad_norm": 0.19394156336784363,
+      "learning_rate": 0.00024300986842105263,
+      "loss": 0.7218,
+      "step": 733
+    },
+    {
+      "epoch": 0.8143114685896546,
+      "grad_norm": 0.26196351647377014,
+      "learning_rate": 0.00024288651315789472,
+      "loss": 0.5631,
+      "step": 734
+    },
+    {
+      "epoch": 0.8154208847593953,
+      "grad_norm": 0.21519336104393005,
+      "learning_rate": 0.00024276315789473682,
+      "loss": 0.7168,
+      "step": 735
+    },
+    {
+      "epoch": 0.816530300929136,
+      "grad_norm": 0.14451566338539124,
+      "learning_rate": 0.00024263980263157894,
+      "loss": 0.4952,
+      "step": 736
+    },
+    {
+      "epoch": 0.8176397170988767,
+      "grad_norm": 0.19086521863937378,
+      "learning_rate": 0.00024251644736842103,
+      "loss": 0.441,
+      "step": 737
+    },
+    {
+      "epoch": 0.8187491332686174,
+      "grad_norm": 0.2369484156370163,
+      "learning_rate": 0.00024239309210526312,
+      "loss": 0.6517,
+      "step": 738
+    },
+    {
+      "epoch": 0.819858549438358,
+      "grad_norm": 0.22705571353435516,
+      "learning_rate": 0.00024226973684210524,
+      "loss": 0.5155,
+      "step": 739
+    },
+    {
+      "epoch": 0.8209679656080987,
+      "grad_norm": 0.16014251112937927,
+      "learning_rate": 0.00024214638157894734,
+      "loss": 0.6391,
+      "step": 740
+    },
+    {
+      "epoch": 0.8220773817778394,
+      "grad_norm": 0.24363452196121216,
+      "learning_rate": 0.00024202302631578943,
+      "loss": 0.6092,
+      "step": 741
+    },
+    {
+      "epoch": 0.8231867979475801,
+      "grad_norm": 0.16901741921901703,
+      "learning_rate": 0.00024189967105263158,
+      "loss": 0.394,
+      "step": 742
+    },
+    {
+      "epoch": 0.8242962141173208,
+      "grad_norm": 0.15532980859279633,
+      "learning_rate": 0.00024177631578947367,
+      "loss": 0.516,
+      "step": 743
+    },
+    {
+      "epoch": 0.8254056302870615,
+      "grad_norm": 0.1833130270242691,
+      "learning_rate": 0.00024165296052631576,
+      "loss": 0.4056,
+      "step": 744
+    },
+    {
+      "epoch": 0.8265150464568021,
+      "grad_norm": 0.23910647630691528,
+      "learning_rate": 0.00024152960526315788,
+      "loss": 0.5056,
+      "step": 745
+    },
+    {
+      "epoch": 0.8276244626265428,
+      "grad_norm": 0.4546511769294739,
+      "learning_rate": 0.00024140624999999998,
+      "loss": 0.701,
+      "step": 746
+    },
+    {
+      "epoch": 0.8287338787962835,
+      "grad_norm": 0.20680895447731018,
+      "learning_rate": 0.00024128289473684207,
+      "loss": 0.5841,
+      "step": 747
+    },
+    {
+      "epoch": 0.8298432949660242,
+      "grad_norm": 0.29819783568382263,
+      "learning_rate": 0.0002411595394736842,
+      "loss": 0.5503,
+      "step": 748
+    },
+    {
+      "epoch": 0.8309527111357649,
+      "grad_norm": 0.18847742676734924,
+      "learning_rate": 0.00024103618421052628,
+      "loss": 0.5588,
+      "step": 749
+    },
+    {
+      "epoch": 0.8320621273055054,
+      "grad_norm": 0.2954421937465668,
+      "learning_rate": 0.00024091282894736838,
+      "loss": 0.5611,
+      "step": 750
+    },
+    {
+      "epoch": 0.8331715434752461,
+      "grad_norm": 0.2604624927043915,
+      "learning_rate": 0.00024078947368421052,
+      "loss": 0.6853,
+      "step": 751
+    },
+    {
+      "epoch": 0.8342809596449868,
+      "grad_norm": 0.31594741344451904,
+      "learning_rate": 0.00024066611842105262,
+      "loss": 0.383,
+      "step": 752
+    },
+    {
+      "epoch": 0.8353903758147275,
+      "grad_norm": 0.2093072086572647,
+      "learning_rate": 0.00024054276315789474,
+      "loss": 0.6134,
+      "step": 753
+    },
+    {
+      "epoch": 0.8364997919844682,
+      "grad_norm": 0.22096338868141174,
+      "learning_rate": 0.00024041940789473683,
+      "loss": 0.4454,
+      "step": 754
+    },
+    {
+      "epoch": 0.8376092081542088,
+      "grad_norm": 0.31209510564804077,
+      "learning_rate": 0.00024029605263157892,
+      "loss": 0.8408,
+      "step": 755
+    },
+    {
+      "epoch": 0.8387186243239495,
+      "grad_norm": 0.36502930521965027,
+      "learning_rate": 0.00024017269736842104,
+      "loss": 0.5053,
+      "step": 756
+    },
+    {
+      "epoch": 0.8398280404936902,
+      "grad_norm": 0.24059328436851501,
+      "learning_rate": 0.00024004934210526314,
+      "loss": 0.6181,
+      "step": 757
+    },
+    {
+      "epoch": 0.8409374566634309,
+      "grad_norm": 0.202326238155365,
+      "learning_rate": 0.00023992598684210523,
+      "loss": 0.4569,
+      "step": 758
+    },
+    {
+      "epoch": 0.8420468728331716,
+      "grad_norm": 0.14637093245983124,
+      "learning_rate": 0.00023980263157894732,
+      "loss": 0.438,
+      "step": 759
+    },
+    {
+      "epoch": 0.8431562890029122,
+      "grad_norm": 0.15636491775512695,
+      "learning_rate": 0.00023967927631578944,
+      "loss": 0.6549,
+      "step": 760
+    },
+    {
+      "epoch": 0.8442657051726529,
+      "grad_norm": 0.25059443712234497,
+      "learning_rate": 0.00023955592105263156,
+      "loss": 0.5497,
+      "step": 761
+    },
+    {
+      "epoch": 0.8453751213423936,
+      "grad_norm": 0.1603458672761917,
+      "learning_rate": 0.00023943256578947368,
+      "loss": 0.5023,
+      "step": 762
+    },
+    {
+      "epoch": 0.8464845375121343,
+      "grad_norm": 0.202356219291687,
+      "learning_rate": 0.00023930921052631578,
+      "loss": 0.6076,
+      "step": 763
+    },
+    {
+      "epoch": 0.847593953681875,
+      "grad_norm": 0.2810531258583069,
+      "learning_rate": 0.00023918585526315787,
+      "loss": 0.4029,
+      "step": 764
+    },
+    {
+      "epoch": 0.8487033698516155,
+      "grad_norm": 0.21425937116146088,
+      "learning_rate": 0.0002390625,
+      "loss": 0.4594,
+      "step": 765
+    },
+    {
+      "epoch": 0.8498127860213562,
+      "grad_norm": 0.29210686683654785,
+      "learning_rate": 0.00023893914473684208,
+      "loss": 0.4246,
+      "step": 766
+    },
+    {
+      "epoch": 0.8509222021910969,
+      "grad_norm": 0.483568400144577,
+      "learning_rate": 0.00023881578947368418,
+      "loss": 0.6333,
+      "step": 767
+    },
+    {
+      "epoch": 0.8520316183608376,
+      "grad_norm": 0.22136186063289642,
+      "learning_rate": 0.0002386924342105263,
+      "loss": 0.6211,
+      "step": 768
+    },
+    {
+      "epoch": 0.8531410345305783,
+      "grad_norm": 0.29133930802345276,
+      "learning_rate": 0.0002385690789473684,
+      "loss": 0.5317,
+      "step": 769
+    },
+    {
+      "epoch": 0.8542504507003189,
+      "grad_norm": 0.20380742847919464,
+      "learning_rate": 0.0002384457236842105,
+      "loss": 0.6699,
+      "step": 770
+    },
+    {
+      "epoch": 0.8553598668700596,
+      "grad_norm": 0.16621272265911102,
+      "learning_rate": 0.00023832236842105263,
+      "loss": 0.8654,
+      "step": 771
+    },
+    {
+      "epoch": 0.8564692830398003,
+      "grad_norm": 0.19278384745121002,
+      "learning_rate": 0.00023819901315789472,
+      "loss": 0.5077,
+      "step": 772
+    },
+    {
+      "epoch": 0.857578699209541,
+      "grad_norm": 0.18274825811386108,
+      "learning_rate": 0.00023807565789473682,
+      "loss": 0.5807,
+      "step": 773
+    },
+    {
+      "epoch": 0.8586881153792817,
+      "grad_norm": 0.36240458488464355,
+      "learning_rate": 0.00023795230263157894,
+      "loss": 0.5612,
+      "step": 774
+    },
+    {
+      "epoch": 0.8597975315490223,
+      "grad_norm": 0.2410973459482193,
+      "learning_rate": 0.00023782894736842103,
+      "loss": 0.6891,
+      "step": 775
+    },
+    {
+      "epoch": 0.860906947718763,
+      "grad_norm": 0.2783324718475342,
+      "learning_rate": 0.00023770559210526312,
+      "loss": 0.4547,
+      "step": 776
+    },
+    {
+      "epoch": 0.8620163638885037,
+      "grad_norm": 0.7285773754119873,
+      "learning_rate": 0.00023758223684210524,
+      "loss": 0.4894,
+      "step": 777
+    },
+    {
+      "epoch": 0.8631257800582444,
+      "grad_norm": 0.16931070387363434,
+      "learning_rate": 0.00023745888157894734,
+      "loss": 0.5592,
+      "step": 778
+    },
+    {
+      "epoch": 0.8642351962279851,
+      "grad_norm": 0.24053402245044708,
+      "learning_rate": 0.00023733552631578943,
+      "loss": 0.4662,
+      "step": 779
+    },
+    {
+      "epoch": 0.8653446123977256,
+      "grad_norm": 0.42136144638061523,
+      "learning_rate": 0.00023721217105263158,
+      "loss": 0.617,
+      "step": 780
+    },
+    {
+      "epoch": 0.8664540285674663,
+      "grad_norm": 0.16937246918678284,
+      "learning_rate": 0.00023708881578947367,
+      "loss": 0.4531,
+      "step": 781
+    },
+    {
+      "epoch": 0.867563444737207,
+      "grad_norm": 0.25334227085113525,
+      "learning_rate": 0.00023696546052631576,
+      "loss": 0.4013,
+      "step": 782
+    },
+    {
+      "epoch": 0.8686728609069477,
+      "grad_norm": 0.18785522878170013,
+      "learning_rate": 0.00023684210526315788,
+      "loss": 0.5154,
+      "step": 783
+    },
+    {
+      "epoch": 0.8697822770766884,
+      "grad_norm": 0.21429309248924255,
+      "learning_rate": 0.00023671874999999998,
+      "loss": 0.4044,
+      "step": 784
+    },
+    {
+      "epoch": 0.8708916932464291,
+      "grad_norm": 0.1965511292219162,
+      "learning_rate": 0.00023659539473684207,
+      "loss": 0.6695,
+      "step": 785
+    },
+    {
+      "epoch": 0.8720011094161697,
+      "grad_norm": 0.2309243083000183,
+      "learning_rate": 0.0002364720394736842,
+      "loss": 0.5465,
+      "step": 786
+    },
+    {
+      "epoch": 0.8731105255859104,
+      "grad_norm": 0.2777848243713379,
+      "learning_rate": 0.00023634868421052628,
+      "loss": 0.6245,
+      "step": 787
+    },
+    {
+      "epoch": 0.8742199417556511,
+      "grad_norm": 0.34535712003707886,
+      "learning_rate": 0.00023622532894736838,
+      "loss": 0.6035,
+      "step": 788
+    },
+    {
+      "epoch": 0.8753293579253918,
+      "grad_norm": 0.24912825226783752,
+      "learning_rate": 0.00023610197368421052,
+      "loss": 0.5369,
+      "step": 789
+    },
+    {
+      "epoch": 0.8764387740951325,
+      "grad_norm": 0.23429974913597107,
+      "learning_rate": 0.00023597861842105262,
+      "loss": 0.3689,
+      "step": 790
+    },
+    {
+      "epoch": 0.8775481902648731,
+      "grad_norm": 0.25908464193344116,
+      "learning_rate": 0.00023585526315789474,
+      "loss": 0.6301,
+      "step": 791
+    },
+    {
+      "epoch": 0.8786576064346138,
+      "grad_norm": 0.3178803324699402,
+      "learning_rate": 0.00023573190789473683,
+      "loss": 0.707,
+      "step": 792
+    },
+    {
+      "epoch": 0.8797670226043545,
+      "grad_norm": 0.23064696788787842,
+      "learning_rate": 0.00023560855263157892,
+      "loss": 0.5725,
+      "step": 793
+    },
+    {
+      "epoch": 0.8808764387740952,
+      "grad_norm": 0.2530830502510071,
+      "learning_rate": 0.00023548519736842102,
+      "loss": 0.5131,
+      "step": 794
+    },
+    {
+      "epoch": 0.8819858549438359,
+      "grad_norm": 0.21092426776885986,
+      "learning_rate": 0.00023536184210526314,
+      "loss": 0.6452,
+      "step": 795
+    },
+    {
+      "epoch": 0.8830952711135764,
+      "grad_norm": 0.21597221493721008,
+      "learning_rate": 0.00023523848684210523,
+      "loss": 0.5438,
+      "step": 796
+    },
+    {
+      "epoch": 0.8842046872833171,
+      "grad_norm": 0.21937009692192078,
+      "learning_rate": 0.00023511513157894732,
+      "loss": 0.4943,
+      "step": 797
+    },
+    {
+      "epoch": 0.8853141034530578,
+      "grad_norm": 0.2742394506931305,
+      "learning_rate": 0.00023499177631578944,
+      "loss": 0.7567,
+      "step": 798
+    },
+    {
+      "epoch": 0.8864235196227985,
+      "grad_norm": 0.3151918053627014,
+      "learning_rate": 0.00023486842105263156,
+      "loss": 0.7817,
+      "step": 799
+    },
+    {
+      "epoch": 0.8875329357925392,
+      "grad_norm": 0.23402948677539825,
+      "learning_rate": 0.00023474506578947368,
+      "loss": 0.3204,
+      "step": 800
+    },
+    {
+      "epoch": 0.8886423519622798,
+      "grad_norm": 0.25392022728919983,
+      "learning_rate": 0.00023462171052631578,
+      "loss": 0.6914,
+      "step": 801
+    },
+    {
+      "epoch": 0.8897517681320205,
+      "grad_norm": 0.22734297811985016,
+      "learning_rate": 0.00023449835526315787,
+      "loss": 0.5779,
+      "step": 802
+    },
+    {
+      "epoch": 0.8908611843017612,
+      "grad_norm": 0.2071351408958435,
+      "learning_rate": 0.000234375,
+      "loss": 0.6024,
+      "step": 803
+    },
+    {
+      "epoch": 0.8919706004715019,
+      "grad_norm": 0.2632406949996948,
+      "learning_rate": 0.00023425164473684208,
+      "loss": 0.5668,
+      "step": 804
+    },
+    {
+      "epoch": 0.8930800166412426,
+      "grad_norm": 0.27089810371398926,
+      "learning_rate": 0.00023412828947368418,
+      "loss": 0.6043,
+      "step": 805
+    },
+    {
+      "epoch": 0.8941894328109832,
+      "grad_norm": 0.1918395459651947,
+      "learning_rate": 0.0002340049342105263,
+      "loss": 0.3305,
+      "step": 806
+    },
+    {
+      "epoch": 0.8952988489807239,
+      "grad_norm": 0.183834046125412,
+      "learning_rate": 0.0002338815789473684,
+      "loss": 0.603,
+      "step": 807
+    },
+    {
+      "epoch": 0.8964082651504646,
+      "grad_norm": 0.29509904980659485,
+      "learning_rate": 0.0002337582236842105,
+      "loss": 0.6681,
+      "step": 808
+    },
+    {
+      "epoch": 0.8975176813202053,
+      "grad_norm": 0.24960756301879883,
+      "learning_rate": 0.00023363486842105263,
+      "loss": 0.5601,
+      "step": 809
+    },
+    {
+      "epoch": 0.898627097489946,
+      "grad_norm": 0.5940669775009155,
+      "learning_rate": 0.00023351151315789472,
+      "loss": 0.5074,
+      "step": 810
+    },
+    {
+      "epoch": 0.8997365136596865,
+      "grad_norm": 0.21877194941043854,
+      "learning_rate": 0.00023338815789473682,
+      "loss": 0.6044,
+      "step": 811
+    },
+    {
+      "epoch": 0.9008459298294272,
+      "grad_norm": 0.24642789363861084,
+      "learning_rate": 0.00023326480263157894,
+      "loss": 0.5249,
+      "step": 812
+    },
+    {
+      "epoch": 0.9019553459991679,
+      "grad_norm": 0.21799951791763306,
+      "learning_rate": 0.00023314144736842103,
+      "loss": 0.5438,
+      "step": 813
+    },
+    {
+      "epoch": 0.9030647621689086,
+      "grad_norm": 0.2310633808374405,
+      "learning_rate": 0.00023301809210526312,
+      "loss": 0.5451,
+      "step": 814
+    },
+    {
+      "epoch": 0.9041741783386493,
+      "grad_norm": 0.1848413199186325,
+      "learning_rate": 0.00023289473684210524,
+      "loss": 0.4699,
+      "step": 815
+    },
+    {
+      "epoch": 0.9052835945083899,
+      "grad_norm": 0.282272607088089,
+      "learning_rate": 0.00023277138157894734,
+      "loss": 0.5288,
+      "step": 816
+    },
+    {
+      "epoch": 0.9063930106781306,
+      "grad_norm": 0.19741901755332947,
+      "learning_rate": 0.00023264802631578943,
+      "loss": 0.4977,
+      "step": 817
+    },
+    {
+      "epoch": 0.9075024268478713,
+      "grad_norm": 0.2287929505109787,
+      "learning_rate": 0.00023252467105263158,
+      "loss": 0.5804,
+      "step": 818
+    },
+    {
+      "epoch": 0.908611843017612,
+      "grad_norm": 0.2509765326976776,
+      "learning_rate": 0.00023240131578947367,
+      "loss": 0.4381,
+      "step": 819
+    },
+    {
+      "epoch": 0.9097212591873527,
+      "grad_norm": 0.2717498540878296,
+      "learning_rate": 0.00023227796052631576,
+      "loss": 0.6095,
+      "step": 820
+    },
+    {
+      "epoch": 0.9108306753570933,
+      "grad_norm": 0.38568127155303955,
+      "learning_rate": 0.00023215460526315789,
+      "loss": 0.512,
+      "step": 821
+    },
+    {
+      "epoch": 0.911940091526834,
+      "grad_norm": 0.22532710433006287,
+      "learning_rate": 0.00023203124999999998,
+      "loss": 0.6333,
+      "step": 822
+    },
+    {
+      "epoch": 0.9130495076965747,
+      "grad_norm": 0.2616422176361084,
+      "learning_rate": 0.00023190789473684207,
+      "loss": 0.4955,
+      "step": 823
+    },
+    {
+      "epoch": 0.9141589238663154,
+      "grad_norm": 0.30875271558761597,
+      "learning_rate": 0.0002317845394736842,
+      "loss": 0.6645,
+      "step": 824
+    },
+    {
+      "epoch": 0.9152683400360561,
+      "grad_norm": 0.22438234090805054,
+      "learning_rate": 0.00023166118421052629,
+      "loss": 0.6209,
+      "step": 825
+    },
+    {
+      "epoch": 0.9163777562057966,
+      "grad_norm": 0.30306366086006165,
+      "learning_rate": 0.00023153782894736838,
+      "loss": 0.6832,
+      "step": 826
+    },
+    {
+      "epoch": 0.9174871723755373,
+      "grad_norm": 0.19736707210540771,
+      "learning_rate": 0.00023141447368421053,
+      "loss": 0.5076,
+      "step": 827
+    },
+    {
+      "epoch": 0.918596588545278,
+      "grad_norm": 0.37344890832901,
+      "learning_rate": 0.00023129111842105262,
+      "loss": 0.5337,
+      "step": 828
+    },
+    {
+      "epoch": 0.9197060047150187,
+      "grad_norm": 0.22570458054542542,
+      "learning_rate": 0.0002311677631578947,
+      "loss": 0.4477,
+      "step": 829
+    },
+    {
+      "epoch": 0.9208154208847594,
+      "grad_norm": 0.2158828228712082,
+      "learning_rate": 0.00023104440789473683,
+      "loss": 0.5738,
+      "step": 830
+    },
+    {
+      "epoch": 0.9219248370545001,
+      "grad_norm": 0.17967133224010468,
+      "learning_rate": 0.00023092105263157893,
+      "loss": 0.6245,
+      "step": 831
+    },
+    {
+      "epoch": 0.9230342532242407,
+      "grad_norm": 0.23400144279003143,
+      "learning_rate": 0.00023079769736842102,
+      "loss": 0.5328,
+      "step": 832
+    },
+    {
+      "epoch": 0.9241436693939814,
+      "grad_norm": 0.20684117078781128,
+      "learning_rate": 0.00023067434210526314,
+      "loss": 0.4556,
+      "step": 833
+    },
+    {
+      "epoch": 0.9252530855637221,
+      "grad_norm": 0.17495577037334442,
+      "learning_rate": 0.00023055098684210523,
+      "loss": 0.4022,
+      "step": 834
+    },
+    {
+      "epoch": 0.9263625017334628,
+      "grad_norm": 0.3661905527114868,
+      "learning_rate": 0.00023042763157894733,
+      "loss": 0.8714,
+      "step": 835
+    },
+    {
+      "epoch": 0.9274719179032035,
+      "grad_norm": 0.20915554463863373,
+      "learning_rate": 0.00023030427631578945,
+      "loss": 0.5164,
+      "step": 836
+    },
+    {
+      "epoch": 0.9285813340729441,
+      "grad_norm": 0.2413186877965927,
+      "learning_rate": 0.00023018092105263157,
+      "loss": 0.515,
+      "step": 837
+    },
+    {
+      "epoch": 0.9296907502426848,
+      "grad_norm": 0.23484086990356445,
+      "learning_rate": 0.00023005756578947369,
+      "loss": 0.5425,
+      "step": 838
+    },
+    {
+      "epoch": 0.9308001664124255,
+      "grad_norm": 0.269280344247818,
+      "learning_rate": 0.00022993421052631578,
+      "loss": 0.5335,
+      "step": 839
+    },
+    {
+      "epoch": 0.9319095825821662,
+      "grad_norm": 0.3550933003425598,
+      "learning_rate": 0.00022981085526315787,
+      "loss": 0.4588,
+      "step": 840
+    },
+    {
+      "epoch": 0.9330189987519069,
+      "grad_norm": 0.33913251757621765,
+      "learning_rate": 0.0002296875,
+      "loss": 0.6615,
+      "step": 841
+    },
+    {
+      "epoch": 0.9341284149216474,
+      "grad_norm": 0.31911739706993103,
+      "learning_rate": 0.00022956414473684209,
+      "loss": 0.6322,
+      "step": 842
+    },
+    {
+      "epoch": 0.9352378310913881,
+      "grad_norm": 0.33628326654434204,
+      "learning_rate": 0.00022944078947368418,
+      "loss": 0.5983,
+      "step": 843
+    },
+    {
+      "epoch": 0.9363472472611288,
+      "grad_norm": 0.1979479342699051,
+      "learning_rate": 0.00022931743421052627,
+      "loss": 0.4327,
+      "step": 844
+    },
+    {
+      "epoch": 0.9374566634308695,
+      "grad_norm": 0.23927690088748932,
+      "learning_rate": 0.0002291940789473684,
+      "loss": 0.5675,
+      "step": 845
+    },
+    {
+      "epoch": 0.9385660796006102,
+      "grad_norm": 0.23125141859054565,
+      "learning_rate": 0.0002290707236842105,
+      "loss": 0.5135,
+      "step": 846
+    },
+    {
+      "epoch": 0.9396754957703508,
+      "grad_norm": 0.2613430917263031,
+      "learning_rate": 0.00022894736842105263,
+      "loss": 0.4361,
+      "step": 847
+    },
+    {
+      "epoch": 0.9407849119400915,
+      "grad_norm": 0.28109273314476013,
+      "learning_rate": 0.00022882401315789473,
+      "loss": 0.5697,
+      "step": 848
+    },
+    {
+      "epoch": 0.9418943281098322,
+      "grad_norm": 0.25478866696357727,
+      "learning_rate": 0.00022870065789473682,
+      "loss": 0.5453,
+      "step": 849
+    },
+    {
+      "epoch": 0.9430037442795729,
+      "grad_norm": 0.2179301530122757,
+      "learning_rate": 0.00022857730263157894,
+      "loss": 0.5588,
+      "step": 850
+    },
+    {
+      "epoch": 0.9441131604493136,
+      "grad_norm": 0.20109961926937103,
+      "learning_rate": 0.00022845394736842103,
+      "loss": 0.5843,
+      "step": 851
+    },
+    {
+      "epoch": 0.9452225766190542,
+      "grad_norm": 0.18201051652431488,
+      "learning_rate": 0.00022833059210526313,
+      "loss": 0.5031,
+      "step": 852
+    },
+    {
+      "epoch": 0.9463319927887949,
+      "grad_norm": 0.25168418884277344,
+      "learning_rate": 0.00022820723684210525,
+      "loss": 0.7,
+      "step": 853
+    },
+    {
+      "epoch": 0.9474414089585356,
+      "grad_norm": 0.2502383291721344,
+      "learning_rate": 0.00022808388157894734,
+      "loss": 0.6002,
+      "step": 854
+    },
+    {
+      "epoch": 0.9485508251282763,
+      "grad_norm": 0.2077435404062271,
+      "learning_rate": 0.00022796052631578943,
+      "loss": 0.632,
+      "step": 855
+    },
+    {
+      "epoch": 0.949660241298017,
+      "grad_norm": 0.28856661915779114,
+      "learning_rate": 0.00022783717105263158,
+      "loss": 0.5973,
+      "step": 856
+    },
+    {
+      "epoch": 0.9507696574677575,
+      "grad_norm": 0.18906134366989136,
+      "learning_rate": 0.00022771381578947367,
+      "loss": 0.5271,
+      "step": 857
+    },
+    {
+      "epoch": 0.9518790736374982,
+      "grad_norm": 0.2617158591747284,
+      "learning_rate": 0.00022759046052631577,
+      "loss": 0.5909,
+      "step": 858
+    },
+    {
+      "epoch": 0.9529884898072389,
+      "grad_norm": 0.42665448784828186,
+      "learning_rate": 0.0002274671052631579,
+      "loss": 0.5476,
+      "step": 859
+    },
+    {
+      "epoch": 0.9540979059769796,
+      "grad_norm": 0.2901977598667145,
+      "learning_rate": 0.00022734374999999998,
+      "loss": 0.6033,
+      "step": 860
+    },
+    {
+      "epoch": 0.9552073221467203,
+      "grad_norm": 0.26080241799354553,
+      "learning_rate": 0.00022722039473684207,
+      "loss": 0.6149,
+      "step": 861
+    },
+    {
+      "epoch": 0.9563167383164609,
+      "grad_norm": 0.2692500948905945,
+      "learning_rate": 0.0002270970394736842,
+      "loss": 0.6035,
+      "step": 862
+    },
+    {
+      "epoch": 0.9574261544862016,
+      "grad_norm": 0.17691655457019806,
+      "learning_rate": 0.00022697368421052629,
+      "loss": 0.5072,
+      "step": 863
+    },
+    {
+      "epoch": 0.9585355706559423,
+      "grad_norm": 0.4164085388183594,
+      "learning_rate": 0.00022685032894736838,
+      "loss": 0.5342,
+      "step": 864
+    },
+    {
+      "epoch": 0.959644986825683,
+      "grad_norm": 0.18131154775619507,
+      "learning_rate": 0.00022672697368421053,
+      "loss": 0.5121,
+      "step": 865
+    },
+    {
+      "epoch": 0.9607544029954237,
+      "grad_norm": 0.22192832827568054,
+      "learning_rate": 0.00022660361842105262,
+      "loss": 0.5307,
+      "step": 866
+    },
+    {
+      "epoch": 0.9618638191651643,
+      "grad_norm": 0.350583016872406,
+      "learning_rate": 0.0002264802631578947,
+      "loss": 0.4176,
+      "step": 867
+    },
+    {
+      "epoch": 0.962973235334905,
+      "grad_norm": 0.2533394396305084,
+      "learning_rate": 0.00022635690789473683,
+      "loss": 0.5289,
+      "step": 868
+    },
+    {
+      "epoch": 0.9640826515046457,
+      "grad_norm": 0.19999092817306519,
+      "learning_rate": 0.00022623355263157893,
+      "loss": 0.5624,
+      "step": 869
+    },
+    {
+      "epoch": 0.9651920676743864,
+      "grad_norm": 0.17564158141613007,
+      "learning_rate": 0.00022611019736842102,
+      "loss": 0.4738,
+      "step": 870
+    },
+    {
+      "epoch": 0.9663014838441271,
+      "grad_norm": 0.35863691568374634,
+      "learning_rate": 0.00022598684210526314,
+      "loss": 0.7332,
+      "step": 871
+    },
+    {
+      "epoch": 0.9674109000138676,
+      "grad_norm": 0.20389032363891602,
+      "learning_rate": 0.00022586348684210523,
+      "loss": 0.4931,
+      "step": 872
+    },
+    {
+      "epoch": 0.9685203161836083,
+      "grad_norm": 0.19918256998062134,
+      "learning_rate": 0.00022574013157894733,
+      "loss": 0.5961,
+      "step": 873
+    },
+    {
+      "epoch": 0.969629732353349,
+      "grad_norm": 0.25909268856048584,
+      "learning_rate": 0.00022561677631578945,
+      "loss": 0.5575,
+      "step": 874
+    },
+    {
+      "epoch": 0.9707391485230897,
+      "grad_norm": 0.18549631536006927,
+      "learning_rate": 0.00022549342105263157,
+      "loss": 0.5181,
+      "step": 875
+    },
+    {
+      "epoch": 0.9718485646928304,
+      "grad_norm": 0.2645319998264313,
+      "learning_rate": 0.0002253700657894737,
+      "loss": 0.6302,
+      "step": 876
+    },
+    {
+      "epoch": 0.9729579808625711,
+      "grad_norm": 0.36148592829704285,
+      "learning_rate": 0.00022524671052631578,
+      "loss": 0.4178,
+      "step": 877
+    },
+    {
+      "epoch": 0.9740673970323117,
+      "grad_norm": 0.2523496150970459,
+      "learning_rate": 0.00022512335526315787,
+      "loss": 0.5333,
+      "step": 878
+    },
+    {
+      "epoch": 0.9751768132020524,
+      "grad_norm": 0.20175378024578094,
+      "learning_rate": 0.000225,
+      "loss": 0.4555,
+      "step": 879
+    },
+    {
+      "epoch": 0.9762862293717931,
+      "grad_norm": 0.7862651348114014,
+      "learning_rate": 0.0002248766447368421,
+      "loss": 0.492,
+      "step": 880
+    },
+    {
+      "epoch": 0.9773956455415338,
+      "grad_norm": 0.27353614568710327,
+      "learning_rate": 0.00022475328947368418,
+      "loss": 0.6768,
+      "step": 881
+    },
+    {
+      "epoch": 0.9785050617112745,
+      "grad_norm": 0.27447402477264404,
+      "learning_rate": 0.00022462993421052627,
+      "loss": 0.7058,
+      "step": 882
+    },
+    {
+      "epoch": 0.9796144778810151,
+      "grad_norm": 0.1933760941028595,
+      "learning_rate": 0.0002245065789473684,
+      "loss": 0.6976,
+      "step": 883
+    },
+    {
+      "epoch": 0.9807238940507558,
+      "grad_norm": 0.18439841270446777,
+      "learning_rate": 0.00022438322368421051,
+      "loss": 0.403,
+      "step": 884
+    },
+    {
+      "epoch": 0.9818333102204965,
+      "grad_norm": 0.27743586897850037,
+      "learning_rate": 0.00022425986842105263,
+      "loss": 0.7012,
+      "step": 885
+    },
+    {
+      "epoch": 0.9829427263902372,
+      "grad_norm": 0.26009905338287354,
+      "learning_rate": 0.00022413651315789473,
+      "loss": 0.5342,
+      "step": 886
+    },
+    {
+      "epoch": 0.9840521425599779,
+      "grad_norm": 0.30932557582855225,
+      "learning_rate": 0.00022401315789473682,
+      "loss": 0.6027,
+      "step": 887
+    },
+    {
+      "epoch": 0.9851615587297184,
+      "grad_norm": 0.21047377586364746,
+      "learning_rate": 0.00022388980263157894,
+      "loss": 0.4078,
+      "step": 888
+    },
+    {
+      "epoch": 0.9862709748994591,
+      "grad_norm": 0.17896795272827148,
+      "learning_rate": 0.00022376644736842103,
+      "loss": 0.5645,
+      "step": 889
+    },
+    {
+      "epoch": 0.9873803910691998,
+      "grad_norm": 0.31928160786628723,
+      "learning_rate": 0.00022364309210526313,
+      "loss": 0.6017,
+      "step": 890
+    },
+    {
+      "epoch": 0.9884898072389405,
+      "grad_norm": 0.2661576271057129,
+      "learning_rate": 0.00022351973684210525,
+      "loss": 0.4079,
+      "step": 891
+    },
+    {
+      "epoch": 0.9895992234086812,
+      "grad_norm": 0.17466966807842255,
+      "learning_rate": 0.00022339638157894734,
+      "loss": 0.4499,
+      "step": 892
+    },
+    {
+      "epoch": 0.9907086395784218,
+      "grad_norm": 0.29928284883499146,
+      "learning_rate": 0.00022327302631578943,
+      "loss": 0.6738,
+      "step": 893
+    },
+    {
+      "epoch": 0.9918180557481625,
+      "grad_norm": 0.22055590152740479,
+      "learning_rate": 0.00022314967105263158,
+      "loss": 0.4979,
+      "step": 894
+    },
+    {
+      "epoch": 0.9929274719179032,
+      "grad_norm": 0.20703110098838806,
+      "learning_rate": 0.00022302631578947367,
+      "loss": 0.34,
+      "step": 895
+    },
+    {
+      "epoch": 0.9940368880876439,
+      "grad_norm": 0.23895259201526642,
+      "learning_rate": 0.00022290296052631577,
+      "loss": 0.6151,
+      "step": 896
+    },
+    {
+      "epoch": 0.9951463042573846,
+      "grad_norm": 0.23790138959884644,
+      "learning_rate": 0.0002227796052631579,
+      "loss": 0.5713,
+      "step": 897
+    },
+    {
+      "epoch": 0.9962557204271252,
+      "grad_norm": 0.18377721309661865,
+      "learning_rate": 0.00022265624999999998,
+      "loss": 0.5082,
+      "step": 898
+    },
+    {
+      "epoch": 0.9973651365968659,
+      "grad_norm": 0.23333214223384857,
+      "learning_rate": 0.00022253289473684207,
+      "loss": 0.4654,
+      "step": 899
+    },
+    {
+      "epoch": 0.9984745527666066,
+      "grad_norm": 0.22151075303554535,
+      "learning_rate": 0.0002224095394736842,
+      "loss": 0.5104,
+      "step": 900
+    },
+    {
+      "epoch": 0.9995839689363473,
+      "grad_norm": 0.2722238302230835,
+      "learning_rate": 0.0002222861842105263,
+      "loss": 0.5298,
+      "step": 901
+    },
+    {
+      "epoch": 1.000693385106088,
+      "grad_norm": 0.3227219879627228,
+      "learning_rate": 0.00022216282894736838,
+      "loss": 0.6207,
+      "step": 902
+    },
+    {
+      "epoch": 1.0018028012758287,
+      "grad_norm": 0.21921642124652863,
+      "learning_rate": 0.00022203947368421053,
+      "loss": 0.4741,
+      "step": 903
+    },
+    {
+      "epoch": 1.0029122174455694,
+      "grad_norm": 0.18044739961624146,
+      "learning_rate": 0.00022191611842105262,
+      "loss": 0.2975,
+      "step": 904
+    },
+    {
+      "epoch": 1.00402163361531,
+      "grad_norm": 0.20503199100494385,
+      "learning_rate": 0.00022179276315789471,
+      "loss": 0.4205,
+      "step": 905
+    },
+    {
+      "epoch": 1.0051310497850505,
+      "grad_norm": 0.18091997504234314,
+      "learning_rate": 0.00022166940789473683,
+      "loss": 0.4393,
+      "step": 906
+    },
+    {
+      "epoch": 1.0062404659547912,
+      "grad_norm": 0.2026747614145279,
+      "learning_rate": 0.00022154605263157893,
+      "loss": 0.6194,
+      "step": 907
+    },
+    {
+      "epoch": 1.007349882124532,
+      "grad_norm": 0.20425230264663696,
+      "learning_rate": 0.00022142269736842102,
+      "loss": 0.7142,
+      "step": 908
+    },
+    {
+      "epoch": 1.0084592982942726,
+      "grad_norm": 0.22876769304275513,
+      "learning_rate": 0.00022129934210526314,
+      "loss": 0.5316,
+      "step": 909
+    },
+    {
+      "epoch": 1.0095687144640133,
+      "grad_norm": 0.3051469624042511,
+      "learning_rate": 0.00022117598684210523,
+      "loss": 0.526,
+      "step": 910
+    },
+    {
+      "epoch": 1.010678130633754,
+      "grad_norm": 0.31023165583610535,
+      "learning_rate": 0.00022105263157894733,
+      "loss": 0.6281,
+      "step": 911
+    },
+    {
+      "epoch": 1.0117875468034947,
+      "grad_norm": 0.2776643633842468,
+      "learning_rate": 0.00022092927631578945,
+      "loss": 0.6222,
+      "step": 912
+    },
+    {
+      "epoch": 1.0128969629732354,
+      "grad_norm": 0.17718107998371124,
+      "learning_rate": 0.00022080592105263157,
+      "loss": 0.497,
+      "step": 913
+    },
+    {
+      "epoch": 1.014006379142976,
+      "grad_norm": 0.18926644325256348,
+      "learning_rate": 0.0002206825657894737,
+      "loss": 0.3907,
+      "step": 914
+    },
+    {
+      "epoch": 1.0151157953127168,
+      "grad_norm": 0.21219785511493683,
+      "learning_rate": 0.00022055921052631578,
+      "loss": 0.7063,
+      "step": 915
+    },
+    {
+      "epoch": 1.0162252114824573,
+      "grad_norm": 0.30529090762138367,
+      "learning_rate": 0.00022043585526315787,
+      "loss": 0.5273,
+      "step": 916
+    },
+    {
+      "epoch": 1.017334627652198,
+      "grad_norm": 0.2160457968711853,
+      "learning_rate": 0.00022031249999999997,
+      "loss": 0.6503,
+      "step": 917
+    },
+    {
+      "epoch": 1.0184440438219386,
+      "grad_norm": 0.19193992018699646,
+      "learning_rate": 0.0002201891447368421,
+      "loss": 0.5096,
+      "step": 918
+    },
+    {
+      "epoch": 1.0195534599916793,
+      "grad_norm": 0.1873706579208374,
+      "learning_rate": 0.00022006578947368418,
+      "loss": 0.5354,
+      "step": 919
+    },
+    {
+      "epoch": 1.02066287616142,
+      "grad_norm": 0.183074951171875,
+      "learning_rate": 0.00021994243421052627,
+      "loss": 0.4163,
+      "step": 920
+    },
+    {
+      "epoch": 1.0217722923311607,
+      "grad_norm": 0.15584836900234222,
+      "learning_rate": 0.0002198190789473684,
+      "loss": 0.3087,
+      "step": 921
+    },
+    {
+      "epoch": 1.0228817085009014,
+      "grad_norm": 0.1514248251914978,
+      "learning_rate": 0.00021969572368421052,
+      "loss": 0.492,
+      "step": 922
+    },
+    {
+      "epoch": 1.0239911246706421,
+      "grad_norm": 0.11994735896587372,
+      "learning_rate": 0.00021957236842105264,
+      "loss": 0.4445,
+      "step": 923
+    },
+    {
+      "epoch": 1.0251005408403828,
+      "grad_norm": 0.23181642591953278,
+      "learning_rate": 0.00021944901315789473,
+      "loss": 0.5837,
+      "step": 924
+    },
+    {
+      "epoch": 1.0262099570101235,
+      "grad_norm": 0.24097567796707153,
+      "learning_rate": 0.00021932565789473682,
+      "loss": 0.5709,
+      "step": 925
+    },
+    {
+      "epoch": 1.027319373179864,
+      "grad_norm": 0.2901909649372101,
+      "learning_rate": 0.00021920230263157894,
+      "loss": 0.6171,
+      "step": 926
+    },
+    {
+      "epoch": 1.0284287893496047,
+      "grad_norm": 0.2848079800605774,
+      "learning_rate": 0.00021907894736842104,
+      "loss": 0.5066,
+      "step": 927
+    },
+    {
+      "epoch": 1.0295382055193454,
+      "grad_norm": 0.26711946725845337,
+      "learning_rate": 0.00021895559210526313,
+      "loss": 0.4874,
+      "step": 928
+    },
+    {
+      "epoch": 1.030647621689086,
+      "grad_norm": 0.20230190455913544,
+      "learning_rate": 0.00021883223684210525,
+      "loss": 0.4349,
+      "step": 929
+    },
+    {
+      "epoch": 1.0317570378588268,
+      "grad_norm": 0.2531988322734833,
+      "learning_rate": 0.00021870888157894734,
+      "loss": 0.5349,
+      "step": 930
+    },
+    {
+      "epoch": 1.0328664540285675,
+      "grad_norm": 0.19993358850479126,
+      "learning_rate": 0.00021858552631578944,
+      "loss": 0.4689,
+      "step": 931
+    },
+    {
+      "epoch": 1.0339758701983082,
+      "grad_norm": 0.36045175790786743,
+      "learning_rate": 0.00021846217105263158,
+      "loss": 0.5803,
+      "step": 932
+    },
+    {
+      "epoch": 1.0350852863680489,
+      "grad_norm": 0.2449137419462204,
+      "learning_rate": 0.00021833881578947368,
+      "loss": 0.5587,
+      "step": 933
+    },
+    {
+      "epoch": 1.0361947025377896,
+      "grad_norm": 0.21992360055446625,
+      "learning_rate": 0.00021821546052631577,
+      "loss": 0.4949,
+      "step": 934
+    },
+    {
+      "epoch": 1.0373041187075303,
+      "grad_norm": 0.22803427278995514,
+      "learning_rate": 0.0002180921052631579,
+      "loss": 0.3884,
+      "step": 935
+    },
+    {
+      "epoch": 1.0384135348772707,
+      "grad_norm": 0.31868723034858704,
+      "learning_rate": 0.00021796874999999998,
+      "loss": 0.4716,
+      "step": 936
+    },
+    {
+      "epoch": 1.0395229510470114,
+      "grad_norm": 0.2443029284477234,
+      "learning_rate": 0.00021784539473684208,
+      "loss": 0.4942,
+      "step": 937
+    },
+    {
+      "epoch": 1.0406323672167521,
+      "grad_norm": 0.16727162897586823,
+      "learning_rate": 0.0002177220394736842,
+      "loss": 0.5104,
+      "step": 938
+    },
+    {
+      "epoch": 1.0417417833864928,
+      "grad_norm": 0.2707812488079071,
+      "learning_rate": 0.0002175986842105263,
+      "loss": 0.7545,
+      "step": 939
+    },
+    {
+      "epoch": 1.0428511995562335,
+      "grad_norm": 0.20524704456329346,
+      "learning_rate": 0.00021747532894736838,
+      "loss": 0.4782,
+      "step": 940
+    },
+    {
+      "epoch": 1.0439606157259742,
+      "grad_norm": 0.2047165483236313,
+      "learning_rate": 0.00021735197368421053,
+      "loss": 0.6799,
+      "step": 941
+    },
+    {
+      "epoch": 1.045070031895715,
+      "grad_norm": 0.21171070635318756,
+      "learning_rate": 0.00021722861842105262,
+      "loss": 0.6063,
+      "step": 942
+    },
+    {
+      "epoch": 1.0461794480654556,
+      "grad_norm": 0.2965310215950012,
+      "learning_rate": 0.00021710526315789472,
+      "loss": 0.8489,
+      "step": 943
+    },
+    {
+      "epoch": 1.0472888642351963,
+      "grad_norm": 0.18332551419734955,
+      "learning_rate": 0.00021698190789473684,
+      "loss": 0.5516,
+      "step": 944
+    },
+    {
+      "epoch": 1.048398280404937,
+      "grad_norm": 0.21715614199638367,
+      "learning_rate": 0.00021685855263157893,
+      "loss": 0.5219,
+      "step": 945
+    },
+    {
+      "epoch": 1.0495076965746777,
+      "grad_norm": 0.2923019230365753,
+      "learning_rate": 0.00021673519736842102,
+      "loss": 0.7848,
+      "step": 946
+    },
+    {
+      "epoch": 1.0506171127444182,
+      "grad_norm": 0.259674072265625,
+      "learning_rate": 0.00021661184210526314,
+      "loss": 0.3492,
+      "step": 947
+    },
+    {
+      "epoch": 1.0517265289141589,
+      "grad_norm": 0.158514603972435,
+      "learning_rate": 0.00021648848684210524,
+      "loss": 0.5425,
+      "step": 948
+    },
+    {
+      "epoch": 1.0528359450838995,
+      "grad_norm": 0.1703483611345291,
+      "learning_rate": 0.00021636513157894733,
+      "loss": 0.4018,
+      "step": 949
+    },
+    {
+      "epoch": 1.0539453612536402,
+      "grad_norm": 0.1840825378894806,
+      "learning_rate": 0.00021624177631578945,
+      "loss": 0.3122,
+      "step": 950
+    },
+    {
+      "epoch": 1.055054777423381,
+      "grad_norm": 0.3050348162651062,
+      "learning_rate": 0.00021611842105263157,
+      "loss": 0.6332,
+      "step": 951
+    },
+    {
+      "epoch": 1.0561641935931216,
+      "grad_norm": 0.2472129911184311,
+      "learning_rate": 0.00021599506578947366,
+      "loss": 0.4852,
+      "step": 952
+    },
+    {
+      "epoch": 1.0572736097628623,
+      "grad_norm": 0.18661095201969147,
+      "learning_rate": 0.00021587171052631578,
+      "loss": 0.3293,
+      "step": 953
+    },
+    {
+      "epoch": 1.058383025932603,
+      "grad_norm": 0.3015456795692444,
+      "learning_rate": 0.00021574835526315788,
+      "loss": 0.5981,
+      "step": 954
+    },
+    {
+      "epoch": 1.0594924421023437,
+      "grad_norm": 0.24642759561538696,
+      "learning_rate": 0.00021562499999999997,
+      "loss": 0.4401,
+      "step": 955
+    },
+    {
+      "epoch": 1.0606018582720844,
+      "grad_norm": 0.23688514530658722,
+      "learning_rate": 0.0002155016447368421,
+      "loss": 0.6831,
+      "step": 956
+    },
+    {
+      "epoch": 1.061711274441825,
+      "grad_norm": 0.20036476850509644,
+      "learning_rate": 0.00021537828947368418,
+      "loss": 0.4985,
+      "step": 957
+    },
+    {
+      "epoch": 1.0628206906115656,
+      "grad_norm": 0.21736058592796326,
+      "learning_rate": 0.00021525493421052628,
+      "loss": 0.7388,
+      "step": 958
+    },
+    {
+      "epoch": 1.0639301067813063,
+      "grad_norm": 0.16539451479911804,
+      "learning_rate": 0.0002151315789473684,
+      "loss": 0.6139,
+      "step": 959
+    },
+    {
+      "epoch": 1.065039522951047,
+      "grad_norm": 0.2896292209625244,
+      "learning_rate": 0.00021500822368421052,
+      "loss": 0.6531,
+      "step": 960
+    },
+    {
+      "epoch": 1.0661489391207877,
+      "grad_norm": 0.2285917103290558,
+      "learning_rate": 0.00021488486842105264,
+      "loss": 0.4059,
+      "step": 961
+    },
+    {
+      "epoch": 1.0672583552905284,
+      "grad_norm": 0.2950262725353241,
+      "learning_rate": 0.00021476151315789473,
+      "loss": 0.6509,
+      "step": 962
+    },
+    {
+      "epoch": 1.068367771460269,
+      "grad_norm": 0.30050045251846313,
+      "learning_rate": 0.00021463815789473682,
+      "loss": 0.513,
+      "step": 963
+    },
+    {
+      "epoch": 1.0694771876300098,
+      "grad_norm": 0.22528204321861267,
+      "learning_rate": 0.00021451480263157894,
+      "loss": 0.4885,
+      "step": 964
+    },
+    {
+      "epoch": 1.0705866037997505,
+      "grad_norm": 0.18981274962425232,
+      "learning_rate": 0.00021439144736842104,
+      "loss": 0.4872,
+      "step": 965
+    },
+    {
+      "epoch": 1.0716960199694912,
+      "grad_norm": 0.17064198851585388,
+      "learning_rate": 0.00021426809210526313,
+      "loss": 0.4601,
+      "step": 966
+    },
+    {
+      "epoch": 1.0728054361392316,
+      "grad_norm": 0.2478228509426117,
+      "learning_rate": 0.00021414473684210522,
+      "loss": 0.4115,
+      "step": 967
+    },
+    {
+      "epoch": 1.0739148523089723,
+      "grad_norm": 0.22317509353160858,
+      "learning_rate": 0.00021402138157894734,
+      "loss": 0.4975,
+      "step": 968
+    },
+    {
+      "epoch": 1.075024268478713,
+      "grad_norm": 0.26215845346450806,
+      "learning_rate": 0.00021389802631578944,
+      "loss": 0.5119,
+      "step": 969
+    },
+    {
+      "epoch": 1.0761336846484537,
+      "grad_norm": 0.2661817967891693,
+      "learning_rate": 0.00021377467105263158,
+      "loss": 0.4886,
+      "step": 970
+    },
+    {
+      "epoch": 1.0772431008181944,
+      "grad_norm": 0.3064764738082886,
+      "learning_rate": 0.00021365131578947368,
+      "loss": 0.4684,
+      "step": 971
+    },
+    {
+      "epoch": 1.078352516987935,
+      "grad_norm": 0.1887666881084442,
+      "learning_rate": 0.00021352796052631577,
+      "loss": 0.5324,
+      "step": 972
+    },
+    {
+      "epoch": 1.0794619331576758,
+      "grad_norm": 0.446283221244812,
+      "learning_rate": 0.0002134046052631579,
+      "loss": 0.4313,
+      "step": 973
+    },
+    {
+      "epoch": 1.0805713493274165,
+      "grad_norm": 0.33540961146354675,
+      "learning_rate": 0.00021328124999999998,
+      "loss": 0.6241,
+      "step": 974
+    },
+    {
+      "epoch": 1.0816807654971572,
+      "grad_norm": 0.234646275639534,
+      "learning_rate": 0.00021315789473684208,
+      "loss": 0.3787,
+      "step": 975
+    },
+    {
+      "epoch": 1.0827901816668979,
+      "grad_norm": 0.21899907290935516,
+      "learning_rate": 0.0002130345394736842,
+      "loss": 0.5201,
+      "step": 976
+    },
+    {
+      "epoch": 1.0838995978366386,
+      "grad_norm": 0.18094147741794586,
+      "learning_rate": 0.0002129111842105263,
+      "loss": 0.44,
+      "step": 977
+    },
+    {
+      "epoch": 1.085009014006379,
+      "grad_norm": 0.1795743703842163,
+      "learning_rate": 0.00021278782894736838,
+      "loss": 0.4502,
+      "step": 978
+    },
+    {
+      "epoch": 1.0861184301761198,
+      "grad_norm": 0.24622996151447296,
+      "learning_rate": 0.00021266447368421053,
+      "loss": 0.4467,
+      "step": 979
+    },
+    {
+      "epoch": 1.0872278463458604,
+      "grad_norm": 0.32457393407821655,
+      "learning_rate": 0.00021254111842105262,
+      "loss": 0.5973,
+      "step": 980
+    },
+    {
+      "epoch": 1.0883372625156011,
+      "grad_norm": 0.18922576308250427,
+      "learning_rate": 0.00021241776315789472,
+      "loss": 0.2973,
+      "step": 981
+    },
+    {
+      "epoch": 1.0894466786853418,
+      "grad_norm": 0.2579767405986786,
+      "learning_rate": 0.00021229440789473684,
+      "loss": 0.4825,
+      "step": 982
+    },
+    {
+      "epoch": 1.0905560948550825,
+      "grad_norm": 0.26494085788726807,
+      "learning_rate": 0.00021217105263157893,
+      "loss": 0.7183,
+      "step": 983
+    },
+    {
+      "epoch": 1.0916655110248232,
+      "grad_norm": 0.26830682158470154,
+      "learning_rate": 0.00021204769736842102,
+      "loss": 0.6421,
+      "step": 984
+    },
+    {
+      "epoch": 1.092774927194564,
+      "grad_norm": 0.21053458750247955,
+      "learning_rate": 0.00021192434210526314,
+      "loss": 0.3367,
+      "step": 985
+    },
+    {
+      "epoch": 1.0938843433643046,
+      "grad_norm": 0.27601686120033264,
+      "learning_rate": 0.00021180098684210524,
+      "loss": 0.5145,
+      "step": 986
+    },
+    {
+      "epoch": 1.094993759534045,
+      "grad_norm": 0.28960034251213074,
+      "learning_rate": 0.00021167763157894733,
+      "loss": 0.548,
+      "step": 987
+    },
+    {
+      "epoch": 1.0961031757037858,
+      "grad_norm": 0.1937156766653061,
+      "learning_rate": 0.00021155427631578945,
+      "loss": 0.3746,
+      "step": 988
+    },
+    {
+      "epoch": 1.0972125918735265,
+      "grad_norm": 0.22715970873832703,
+      "learning_rate": 0.00021143092105263157,
+      "loss": 0.6459,
+      "step": 989
+    },
+    {
+      "epoch": 1.0983220080432672,
+      "grad_norm": 0.2094552218914032,
+      "learning_rate": 0.00021130756578947366,
+      "loss": 0.4155,
+      "step": 990
+    },
+    {
+      "epoch": 1.0994314242130079,
+      "grad_norm": 0.23469318449497223,
+      "learning_rate": 0.00021118421052631578,
+      "loss": 0.6588,
+      "step": 991
+    },
+    {
+      "epoch": 1.1005408403827486,
+      "grad_norm": 0.18322822451591492,
+      "learning_rate": 0.00021106085526315788,
+      "loss": 0.4636,
+      "step": 992
+    },
+    {
+      "epoch": 1.1016502565524893,
+      "grad_norm": 0.2863262891769409,
+      "learning_rate": 0.00021093749999999997,
+      "loss": 0.4921,
+      "step": 993
+    },
+    {
+      "epoch": 1.10275967272223,
+      "grad_norm": 0.2982548177242279,
+      "learning_rate": 0.0002108141447368421,
+      "loss": 0.6896,
+      "step": 994
+    },
+    {
+      "epoch": 1.1038690888919707,
+      "grad_norm": 0.35214439034461975,
+      "learning_rate": 0.00021069078947368418,
+      "loss": 0.4343,
+      "step": 995
+    },
+    {
+      "epoch": 1.1049785050617114,
+      "grad_norm": 0.19073866307735443,
+      "learning_rate": 0.00021056743421052628,
+      "loss": 0.4724,
+      "step": 996
+    },
+    {
+      "epoch": 1.106087921231452,
+      "grad_norm": 0.24945704638957977,
+      "learning_rate": 0.0002104440789473684,
+      "loss": 0.3815,
+      "step": 997
+    },
+    {
+      "epoch": 1.1071973374011925,
+      "grad_norm": 0.27915990352630615,
+      "learning_rate": 0.00021032072368421052,
+      "loss": 0.5862,
+      "step": 998
+    },
+    {
+      "epoch": 1.1083067535709332,
+      "grad_norm": 0.3030833899974823,
+      "learning_rate": 0.00021019736842105264,
+      "loss": 0.4567,
+      "step": 999
+    },
+    {
+      "epoch": 1.109416169740674,
+      "grad_norm": 0.22396822273731232,
+      "learning_rate": 0.00021007401315789473,
+      "loss": 0.4252,
+      "step": 1000
+    },
+    {
+      "epoch": 1.1105255859104146,
+      "grad_norm": 0.285520076751709,
+      "learning_rate": 0.00020995065789473682,
+      "loss": 0.3036,
+      "step": 1001
+    },
+    {
+      "epoch": 1.1116350020801553,
+      "grad_norm": 0.2318202704191208,
+      "learning_rate": 0.00020982730263157892,
+      "loss": 0.44,
+      "step": 1002
+    },
+    {
+      "epoch": 1.112744418249896,
+      "grad_norm": 0.24701035022735596,
+      "learning_rate": 0.00020970394736842104,
+      "loss": 0.5542,
+      "step": 1003
+    },
+    {
+      "epoch": 1.1138538344196367,
+      "grad_norm": 0.25920355319976807,
+      "learning_rate": 0.00020958059210526313,
+      "loss": 0.5066,
+      "step": 1004
+    },
+    {
+      "epoch": 1.1149632505893774,
+      "grad_norm": 0.6409327387809753,
+      "learning_rate": 0.00020945723684210522,
+      "loss": 0.6025,
+      "step": 1005
+    },
+    {
+      "epoch": 1.116072666759118,
+      "grad_norm": 0.469032883644104,
+      "learning_rate": 0.00020933388157894734,
+      "loss": 0.4099,
+      "step": 1006
+    },
+    {
+      "epoch": 1.1171820829288588,
+      "grad_norm": 0.181956484913826,
+      "learning_rate": 0.00020921052631578944,
+      "loss": 0.353,
+      "step": 1007
+    },
+    {
+      "epoch": 1.1182914990985993,
+      "grad_norm": 0.2923588752746582,
+      "learning_rate": 0.00020908717105263158,
+      "loss": 0.5331,
+      "step": 1008
+    },
+    {
+      "epoch": 1.11940091526834,
+      "grad_norm": 0.29170429706573486,
+      "learning_rate": 0.00020896381578947368,
+      "loss": 0.6123,
+      "step": 1009
+    },
+    {
+      "epoch": 1.1205103314380807,
+      "grad_norm": 0.23297952115535736,
+      "learning_rate": 0.00020884046052631577,
+      "loss": 0.3093,
+      "step": 1010
+    },
+    {
+      "epoch": 1.1216197476078213,
+      "grad_norm": 0.3802937865257263,
+      "learning_rate": 0.0002087171052631579,
+      "loss": 0.5309,
+      "step": 1011
+    },
+    {
+      "epoch": 1.122729163777562,
+      "grad_norm": 0.21260103583335876,
+      "learning_rate": 0.00020859374999999998,
+      "loss": 0.5288,
+      "step": 1012
+    },
+    {
+      "epoch": 1.1238385799473027,
+      "grad_norm": 0.22543269395828247,
+      "learning_rate": 0.00020847039473684208,
+      "loss": 0.4474,
+      "step": 1013
+    },
+    {
+      "epoch": 1.1249479961170434,
+      "grad_norm": 0.23437976837158203,
+      "learning_rate": 0.0002083470394736842,
+      "loss": 0.622,
+      "step": 1014
+    },
+    {
+      "epoch": 1.1260574122867841,
+      "grad_norm": 0.22682055830955505,
+      "learning_rate": 0.0002082236842105263,
+      "loss": 0.4028,
+      "step": 1015
+    },
+    {
+      "epoch": 1.1271668284565248,
+      "grad_norm": 0.27149879932403564,
+      "learning_rate": 0.00020810032894736838,
+      "loss": 0.3926,
+      "step": 1016
+    },
+    {
+      "epoch": 1.1282762446262655,
+      "grad_norm": 0.21635794639587402,
+      "learning_rate": 0.00020797697368421053,
+      "loss": 0.5204,
+      "step": 1017
+    },
+    {
+      "epoch": 1.129385660796006,
+      "grad_norm": 0.29137322306632996,
+      "learning_rate": 0.00020785361842105263,
+      "loss": 0.5625,
+      "step": 1018
+    },
+    {
+      "epoch": 1.1304950769657467,
+      "grad_norm": 0.2547966241836548,
+      "learning_rate": 0.00020773026315789472,
+      "loss": 0.7025,
+      "step": 1019
+    },
+    {
+      "epoch": 1.1316044931354874,
+      "grad_norm": 0.34155234694480896,
+      "learning_rate": 0.00020760690789473684,
+      "loss": 0.5504,
+      "step": 1020
+    },
+    {
+      "epoch": 1.132713909305228,
+      "grad_norm": 0.30655375123023987,
+      "learning_rate": 0.00020748355263157893,
+      "loss": 0.8426,
+      "step": 1021
+    },
+    {
+      "epoch": 1.1338233254749688,
+      "grad_norm": 0.24423359334468842,
+      "learning_rate": 0.00020736019736842102,
+      "loss": 0.4466,
+      "step": 1022
+    },
+    {
+      "epoch": 1.1349327416447095,
+      "grad_norm": 0.24093151092529297,
+      "learning_rate": 0.00020723684210526315,
+      "loss": 0.6558,
+      "step": 1023
+    },
+    {
+      "epoch": 1.1360421578144502,
+      "grad_norm": 0.2106793373823166,
+      "learning_rate": 0.00020711348684210524,
+      "loss": 0.4964,
+      "step": 1024
+    },
+    {
+      "epoch": 1.1371515739841909,
+      "grad_norm": 0.27556413412094116,
+      "learning_rate": 0.00020699013157894733,
+      "loss": 0.4517,
+      "step": 1025
+    },
+    {
+      "epoch": 1.1382609901539316,
+      "grad_norm": 0.28604790568351746,
+      "learning_rate": 0.00020686677631578945,
+      "loss": 0.606,
+      "step": 1026
+    },
+    {
+      "epoch": 1.1393704063236723,
+      "grad_norm": 0.2911072075366974,
+      "learning_rate": 0.00020674342105263157,
+      "loss": 0.4299,
+      "step": 1027
+    },
+    {
+      "epoch": 1.140479822493413,
+      "grad_norm": 0.19676262140274048,
+      "learning_rate": 0.00020662006578947367,
+      "loss": 0.3269,
+      "step": 1028
+    },
+    {
+      "epoch": 1.1415892386631534,
+      "grad_norm": 0.2523801624774933,
+      "learning_rate": 0.00020649671052631579,
+      "loss": 0.3888,
+      "step": 1029
+    },
+    {
+      "epoch": 1.1426986548328941,
+      "grad_norm": 0.2896265387535095,
+      "learning_rate": 0.00020637335526315788,
+      "loss": 0.4757,
+      "step": 1030
+    },
+    {
+      "epoch": 1.1438080710026348,
+      "grad_norm": 0.24346299469470978,
+      "learning_rate": 0.00020624999999999997,
+      "loss": 0.5065,
+      "step": 1031
+    },
+    {
+      "epoch": 1.1449174871723755,
+      "grad_norm": 0.2244485467672348,
+      "learning_rate": 0.0002061266447368421,
+      "loss": 0.4212,
+      "step": 1032
+    },
+    {
+      "epoch": 1.1460269033421162,
+      "grad_norm": 0.2615904211997986,
+      "learning_rate": 0.00020600328947368419,
+      "loss": 0.3755,
+      "step": 1033
+    },
+    {
+      "epoch": 1.147136319511857,
+      "grad_norm": 0.2203037440776825,
+      "learning_rate": 0.00020587993421052628,
+      "loss": 0.7468,
+      "step": 1034
+    },
+    {
+      "epoch": 1.1482457356815976,
+      "grad_norm": 0.2502787709236145,
+      "learning_rate": 0.0002057565789473684,
+      "loss": 0.4717,
+      "step": 1035
+    },
+    {
+      "epoch": 1.1493551518513383,
+      "grad_norm": 0.23755724728107452,
+      "learning_rate": 0.00020563322368421052,
+      "loss": 0.4367,
+      "step": 1036
+    },
+    {
+      "epoch": 1.150464568021079,
+      "grad_norm": 0.2541312575340271,
+      "learning_rate": 0.00020550986842105264,
+      "loss": 0.3824,
+      "step": 1037
+    },
+    {
+      "epoch": 1.1515739841908195,
+      "grad_norm": 0.2297431081533432,
+      "learning_rate": 0.00020538651315789473,
+      "loss": 0.5744,
+      "step": 1038
+    },
+    {
+      "epoch": 1.1526834003605602,
+      "grad_norm": 0.25546327233314514,
+      "learning_rate": 0.00020526315789473683,
+      "loss": 0.5484,
+      "step": 1039
+    },
+    {
+      "epoch": 1.1537928165303009,
+      "grad_norm": 0.27455171942710876,
+      "learning_rate": 0.00020513980263157892,
+      "loss": 0.4961,
+      "step": 1040
+    },
+    {
+      "epoch": 1.1549022327000416,
+      "grad_norm": 0.221107617020607,
+      "learning_rate": 0.00020501644736842104,
+      "loss": 0.6225,
+      "step": 1041
+    },
+    {
+      "epoch": 1.1560116488697822,
+      "grad_norm": 0.3334260582923889,
+      "learning_rate": 0.00020489309210526313,
+      "loss": 0.5827,
+      "step": 1042
+    },
+    {
+      "epoch": 1.157121065039523,
+      "grad_norm": 0.2542990446090698,
+      "learning_rate": 0.00020476973684210523,
+      "loss": 0.4444,
+      "step": 1043
+    },
+    {
+      "epoch": 1.1582304812092636,
+      "grad_norm": 0.24342231452465057,
+      "learning_rate": 0.00020464638157894735,
+      "loss": 0.4379,
+      "step": 1044
+    },
+    {
+      "epoch": 1.1593398973790043,
+      "grad_norm": 0.27247416973114014,
+      "learning_rate": 0.00020452302631578944,
+      "loss": 0.5547,
+      "step": 1045
+    },
+    {
+      "epoch": 1.160449313548745,
+      "grad_norm": 0.21573393046855927,
+      "learning_rate": 0.00020439967105263159,
+      "loss": 0.4349,
+      "step": 1046
+    },
+    {
+      "epoch": 1.1615587297184857,
+      "grad_norm": 0.2523028552532196,
+      "learning_rate": 0.00020427631578947368,
+      "loss": 0.4237,
+      "step": 1047
+    },
+    {
+      "epoch": 1.1626681458882264,
+      "grad_norm": 0.30524906516075134,
+      "learning_rate": 0.00020415296052631577,
+      "loss": 0.3916,
+      "step": 1048
+    },
+    {
+      "epoch": 1.163777562057967,
+      "grad_norm": 0.3323182463645935,
+      "learning_rate": 0.0002040296052631579,
+      "loss": 0.6462,
+      "step": 1049
+    },
+    {
+      "epoch": 1.1648869782277076,
+      "grad_norm": 0.31420886516571045,
+      "learning_rate": 0.00020390624999999999,
+      "loss": 0.4049,
+      "step": 1050
+    },
+    {
+      "epoch": 1.1659963943974483,
+      "grad_norm": 0.32108211517333984,
+      "learning_rate": 0.00020378289473684208,
+      "loss": 0.4562,
+      "step": 1051
+    },
+    {
+      "epoch": 1.167105810567189,
+      "grad_norm": 0.1855727732181549,
+      "learning_rate": 0.00020365953947368417,
+      "loss": 0.4673,
+      "step": 1052
+    },
+    {
+      "epoch": 1.1682152267369297,
+      "grad_norm": 0.3211022615432739,
+      "learning_rate": 0.0002035361842105263,
+      "loss": 0.3942,
+      "step": 1053
+    },
+    {
+      "epoch": 1.1693246429066704,
+      "grad_norm": 0.25351840257644653,
+      "learning_rate": 0.00020341282894736839,
+      "loss": 0.5996,
+      "step": 1054
+    },
+    {
+      "epoch": 1.170434059076411,
+      "grad_norm": 0.3350581228733063,
+      "learning_rate": 0.00020328947368421053,
+      "loss": 0.9422,
+      "step": 1055
+    },
+    {
+      "epoch": 1.1715434752461518,
+      "grad_norm": 0.28251969814300537,
+      "learning_rate": 0.00020316611842105263,
+      "loss": 0.5418,
+      "step": 1056
+    },
+    {
+      "epoch": 1.1726528914158925,
+      "grad_norm": 0.25546523928642273,
+      "learning_rate": 0.00020304276315789472,
+      "loss": 0.5415,
+      "step": 1057
+    },
+    {
+      "epoch": 1.173762307585633,
+      "grad_norm": 0.2648818790912628,
+      "learning_rate": 0.00020291940789473684,
+      "loss": 0.5584,
+      "step": 1058
+    },
+    {
+      "epoch": 1.1748717237553739,
+      "grad_norm": 0.3260975480079651,
+      "learning_rate": 0.00020279605263157893,
+      "loss": 0.5427,
+      "step": 1059
+    },
+    {
+      "epoch": 1.1759811399251143,
+      "grad_norm": 0.2364053875207901,
+      "learning_rate": 0.00020267269736842103,
+      "loss": 0.5468,
+      "step": 1060
+    },
+    {
+      "epoch": 1.177090556094855,
+      "grad_norm": 0.3164118230342865,
+      "learning_rate": 0.00020254934210526315,
+      "loss": 0.5242,
+      "step": 1061
+    },
+    {
+      "epoch": 1.1781999722645957,
+      "grad_norm": 0.22556494176387787,
+      "learning_rate": 0.00020242598684210524,
+      "loss": 0.6513,
+      "step": 1062
+    },
+    {
+      "epoch": 1.1793093884343364,
+      "grad_norm": 0.27951955795288086,
+      "learning_rate": 0.00020230263157894733,
+      "loss": 0.6117,
+      "step": 1063
+    },
+    {
+      "epoch": 1.180418804604077,
+      "grad_norm": 0.271635502576828,
+      "learning_rate": 0.00020217927631578945,
+      "loss": 0.3996,
+      "step": 1064
+    },
+    {
+      "epoch": 1.1815282207738178,
+      "grad_norm": 0.23386971652507782,
+      "learning_rate": 0.00020205592105263157,
+      "loss": 0.4866,
+      "step": 1065
+    },
+    {
+      "epoch": 1.1826376369435585,
+      "grad_norm": 0.2693704664707184,
+      "learning_rate": 0.00020193256578947367,
+      "loss": 0.6313,
+      "step": 1066
+    },
+    {
+      "epoch": 1.1837470531132992,
+      "grad_norm": 0.2502618432044983,
+      "learning_rate": 0.0002018092105263158,
+      "loss": 0.5445,
+      "step": 1067
+    },
+    {
+      "epoch": 1.18485646928304,
+      "grad_norm": 0.22942642867565155,
+      "learning_rate": 0.00020168585526315788,
+      "loss": 0.4443,
+      "step": 1068
+    },
+    {
+      "epoch": 1.1859658854527804,
+      "grad_norm": 0.23630082607269287,
+      "learning_rate": 0.00020156249999999997,
+      "loss": 0.5264,
+      "step": 1069
+    },
+    {
+      "epoch": 1.187075301622521,
+      "grad_norm": 0.29252779483795166,
+      "learning_rate": 0.0002014391447368421,
+      "loss": 0.5528,
+      "step": 1070
+    },
+    {
+      "epoch": 1.1881847177922618,
+      "grad_norm": 0.2584315538406372,
+      "learning_rate": 0.0002013157894736842,
+      "loss": 0.5017,
+      "step": 1071
+    },
+    {
+      "epoch": 1.1892941339620025,
+      "grad_norm": 0.28517627716064453,
+      "learning_rate": 0.00020119243421052628,
+      "loss": 0.4536,
+      "step": 1072
+    },
+    {
+      "epoch": 1.1904035501317431,
+      "grad_norm": 0.432355672121048,
+      "learning_rate": 0.0002010690789473684,
+      "loss": 0.592,
+      "step": 1073
+    },
+    {
+      "epoch": 1.1915129663014838,
+      "grad_norm": 0.3467410206794739,
+      "learning_rate": 0.00020094572368421052,
+      "loss": 0.7285,
+      "step": 1074
+    },
+    {
+      "epoch": 1.1926223824712245,
+      "grad_norm": 0.29943886399269104,
+      "learning_rate": 0.0002008223684210526,
+      "loss": 0.7591,
+      "step": 1075
+    },
+    {
+      "epoch": 1.1937317986409652,
+      "grad_norm": 0.17552253603935242,
+      "learning_rate": 0.00020069901315789473,
+      "loss": 0.2609,
+      "step": 1076
+    },
+    {
+      "epoch": 1.194841214810706,
+      "grad_norm": 0.20625296235084534,
+      "learning_rate": 0.00020057565789473683,
+      "loss": 0.3158,
+      "step": 1077
+    },
+    {
+      "epoch": 1.1959506309804466,
+      "grad_norm": 0.27414992451667786,
+      "learning_rate": 0.00020045230263157892,
+      "loss": 0.3384,
+      "step": 1078
+    },
+    {
+      "epoch": 1.1970600471501873,
+      "grad_norm": 0.2584741711616516,
+      "learning_rate": 0.00020032894736842104,
+      "loss": 0.3626,
+      "step": 1079
+    },
+    {
+      "epoch": 1.1981694633199278,
+      "grad_norm": 0.2000085860490799,
+      "learning_rate": 0.00020020559210526313,
+      "loss": 0.3534,
+      "step": 1080
+    },
+    {
+      "epoch": 1.1992788794896685,
+      "grad_norm": 0.4106660783290863,
+      "learning_rate": 0.00020008223684210523,
+      "loss": 0.4994,
+      "step": 1081
+    },
+    {
+      "epoch": 1.2003882956594092,
+      "grad_norm": 0.3149929940700531,
+      "learning_rate": 0.00019995888157894735,
+      "loss": 0.5106,
+      "step": 1082
+    },
+    {
+      "epoch": 1.2014977118291499,
+      "grad_norm": 0.27537012100219727,
+      "learning_rate": 0.00019983552631578944,
+      "loss": 0.403,
+      "step": 1083
+    },
+    {
+      "epoch": 1.2026071279988906,
+      "grad_norm": 0.27250340580940247,
+      "learning_rate": 0.0001997121710526316,
+      "loss": 0.4476,
+      "step": 1084
+    },
+    {
+      "epoch": 1.2037165441686313,
+      "grad_norm": 0.3000098168849945,
+      "learning_rate": 0.00019958881578947368,
+      "loss": 0.5365,
+      "step": 1085
+    },
+    {
+      "epoch": 1.204825960338372,
+      "grad_norm": 0.2616446912288666,
+      "learning_rate": 0.00019946546052631577,
+      "loss": 0.4966,
+      "step": 1086
+    },
+    {
+      "epoch": 1.2059353765081127,
+      "grad_norm": 0.3749598264694214,
+      "learning_rate": 0.0001993421052631579,
+      "loss": 0.4181,
+      "step": 1087
+    },
+    {
+      "epoch": 1.2070447926778534,
+      "grad_norm": 0.22599004209041595,
+      "learning_rate": 0.00019921875,
+      "loss": 0.4952,
+      "step": 1088
+    },
+    {
+      "epoch": 1.2081542088475938,
+      "grad_norm": 0.3596106469631195,
+      "learning_rate": 0.00019909539473684208,
+      "loss": 0.4713,
+      "step": 1089
+    },
+    {
+      "epoch": 1.2092636250173345,
+      "grad_norm": 0.32484951615333557,
+      "learning_rate": 0.00019897203947368417,
+      "loss": 0.5465,
+      "step": 1090
+    },
+    {
+      "epoch": 1.2103730411870752,
+      "grad_norm": 0.3571338951587677,
+      "learning_rate": 0.0001988486842105263,
+      "loss": 0.6788,
+      "step": 1091
+    },
+    {
+      "epoch": 1.211482457356816,
+      "grad_norm": 0.34311506152153015,
+      "learning_rate": 0.0001987253289473684,
+      "loss": 0.5347,
+      "step": 1092
+    },
+    {
+      "epoch": 1.2125918735265566,
+      "grad_norm": 0.32343587279319763,
+      "learning_rate": 0.00019860197368421053,
+      "loss": 0.3366,
+      "step": 1093
+    },
+    {
+      "epoch": 1.2137012896962973,
+      "grad_norm": 0.22507165372371674,
+      "learning_rate": 0.00019847861842105263,
+      "loss": 0.5374,
+      "step": 1094
+    },
+    {
+      "epoch": 1.214810705866038,
+      "grad_norm": 0.25680041313171387,
+      "learning_rate": 0.00019835526315789472,
+      "loss": 0.3549,
+      "step": 1095
+    },
+    {
+      "epoch": 1.2159201220357787,
+      "grad_norm": 0.2538761496543884,
+      "learning_rate": 0.00019823190789473684,
+      "loss": 0.4335,
+      "step": 1096
+    },
+    {
+      "epoch": 1.2170295382055194,
+      "grad_norm": 0.2690007984638214,
+      "learning_rate": 0.00019810855263157893,
+      "loss": 0.7695,
+      "step": 1097
+    },
+    {
+      "epoch": 1.21813895437526,
+      "grad_norm": 0.19778668880462646,
+      "learning_rate": 0.00019798519736842103,
+      "loss": 0.3248,
+      "step": 1098
+    },
+    {
+      "epoch": 1.2192483705450008,
+      "grad_norm": 0.23934458196163177,
+      "learning_rate": 0.00019786184210526315,
+      "loss": 0.3444,
+      "step": 1099
+    },
+    {
+      "epoch": 1.2203577867147413,
+      "grad_norm": 0.2791898846626282,
+      "learning_rate": 0.00019773848684210524,
+      "loss": 0.4964,
+      "step": 1100
+    },
+    {
+      "epoch": 1.221467202884482,
+      "grad_norm": 0.24607348442077637,
+      "learning_rate": 0.00019761513157894733,
+      "loss": 0.59,
+      "step": 1101
+    },
+    {
+      "epoch": 1.2225766190542227,
+      "grad_norm": 0.27295032143592834,
+      "learning_rate": 0.00019749177631578943,
+      "loss": 0.4905,
+      "step": 1102
+    },
+    {
+      "epoch": 1.2236860352239634,
+      "grad_norm": 0.2771230936050415,
+      "learning_rate": 0.00019736842105263157,
+      "loss": 0.66,
+      "step": 1103
+    },
+    {
+      "epoch": 1.224795451393704,
+      "grad_norm": 0.30468347668647766,
+      "learning_rate": 0.00019724506578947367,
+      "loss": 0.4982,
+      "step": 1104
+    },
+    {
+      "epoch": 1.2259048675634447,
+      "grad_norm": 0.2552284300327301,
+      "learning_rate": 0.0001971217105263158,
+      "loss": 0.3414,
+      "step": 1105
+    },
+    {
+      "epoch": 1.2270142837331854,
+      "grad_norm": 0.3029981255531311,
+      "learning_rate": 0.00019699835526315788,
+      "loss": 0.7422,
+      "step": 1106
+    },
+    {
+      "epoch": 1.2281236999029261,
+      "grad_norm": 0.27999430894851685,
+      "learning_rate": 0.00019687499999999997,
+      "loss": 0.5393,
+      "step": 1107
+    },
+    {
+      "epoch": 1.2292331160726668,
+      "grad_norm": 0.25855761766433716,
+      "learning_rate": 0.0001967516447368421,
+      "loss": 0.4018,
+      "step": 1108
+    },
+    {
+      "epoch": 1.2303425322424075,
+      "grad_norm": 0.4557599425315857,
+      "learning_rate": 0.0001966282894736842,
+      "loss": 0.5082,
+      "step": 1109
+    },
+    {
+      "epoch": 1.2314519484121482,
+      "grad_norm": 0.2822202742099762,
+      "learning_rate": 0.00019650493421052628,
+      "loss": 0.6512,
+      "step": 1110
+    },
+    {
+      "epoch": 1.2325613645818887,
+      "grad_norm": 0.31011104583740234,
+      "learning_rate": 0.0001963815789473684,
+      "loss": 0.3538,
+      "step": 1111
+    },
+    {
+      "epoch": 1.2336707807516294,
+      "grad_norm": 0.23474127054214478,
+      "learning_rate": 0.00019625822368421052,
+      "loss": 0.5368,
+      "step": 1112
+    },
+    {
+      "epoch": 1.23478019692137,
+      "grad_norm": 0.2853519320487976,
+      "learning_rate": 0.00019613486842105261,
+      "loss": 0.5207,
+      "step": 1113
+    },
+    {
+      "epoch": 1.2358896130911108,
+      "grad_norm": 0.32621482014656067,
+      "learning_rate": 0.00019601151315789473,
+      "loss": 0.555,
+      "step": 1114
+    },
+    {
+      "epoch": 1.2369990292608515,
+      "grad_norm": 0.2757001519203186,
+      "learning_rate": 0.00019588815789473683,
+      "loss": 0.6936,
+      "step": 1115
+    },
+    {
+      "epoch": 1.2381084454305922,
+      "grad_norm": 0.2345856875181198,
+      "learning_rate": 0.00019576480263157892,
+      "loss": 0.6683,
+      "step": 1116
+    },
+    {
+      "epoch": 1.2392178616003329,
+      "grad_norm": 0.2641811668872833,
+      "learning_rate": 0.00019564144736842104,
+      "loss": 0.4647,
+      "step": 1117
+    },
+    {
+      "epoch": 1.2403272777700736,
+      "grad_norm": 0.2421552538871765,
+      "learning_rate": 0.00019551809210526313,
+      "loss": 0.8658,
+      "step": 1118
+    },
+    {
+      "epoch": 1.2414366939398143,
+      "grad_norm": 0.3254922032356262,
+      "learning_rate": 0.00019539473684210523,
+      "loss": 0.4028,
+      "step": 1119
+    },
+    {
+      "epoch": 1.2425461101095547,
+      "grad_norm": 0.27477791905403137,
+      "learning_rate": 0.00019527138157894735,
+      "loss": 0.4876,
+      "step": 1120
+    },
+    {
+      "epoch": 1.2436555262792954,
+      "grad_norm": 0.37873339653015137,
+      "learning_rate": 0.00019514802631578944,
+      "loss": 0.5121,
+      "step": 1121
+    },
+    {
+      "epoch": 1.2447649424490361,
+      "grad_norm": 0.20323941111564636,
+      "learning_rate": 0.0001950246710526316,
+      "loss": 0.4327,
+      "step": 1122
+    },
+    {
+      "epoch": 1.2458743586187768,
+      "grad_norm": 0.2840193510055542,
+      "learning_rate": 0.00019490131578947368,
+      "loss": 0.3922,
+      "step": 1123
+    },
+    {
+      "epoch": 1.2469837747885175,
+      "grad_norm": 0.24474340677261353,
+      "learning_rate": 0.00019477796052631578,
+      "loss": 0.5334,
+      "step": 1124
+    },
+    {
+      "epoch": 1.2480931909582582,
+      "grad_norm": 0.25993865728378296,
+      "learning_rate": 0.00019465460526315787,
+      "loss": 0.3783,
+      "step": 1125
+    },
+    {
+      "epoch": 1.249202607127999,
+      "grad_norm": 0.3326314687728882,
+      "learning_rate": 0.00019453125,
+      "loss": 0.4944,
+      "step": 1126
+    },
+    {
+      "epoch": 1.2503120232977396,
+      "grad_norm": 0.26182398200035095,
+      "learning_rate": 0.00019440789473684208,
+      "loss": 0.8275,
+      "step": 1127
+    },
+    {
+      "epoch": 1.2514214394674803,
+      "grad_norm": 0.26601549983024597,
+      "learning_rate": 0.00019428453947368417,
+      "loss": 0.4694,
+      "step": 1128
+    },
+    {
+      "epoch": 1.2525308556372208,
+      "grad_norm": 0.27001309394836426,
+      "learning_rate": 0.0001941611842105263,
+      "loss": 0.4887,
+      "step": 1129
+    },
+    {
+      "epoch": 1.2536402718069617,
+      "grad_norm": 0.21264316141605377,
+      "learning_rate": 0.0001940378289473684,
+      "loss": 0.5031,
+      "step": 1130
+    },
+    {
+      "epoch": 1.2547496879767022,
+      "grad_norm": 0.1987910121679306,
+      "learning_rate": 0.00019391447368421054,
+      "loss": 0.3877,
+      "step": 1131
+    },
+    {
+      "epoch": 1.2558591041464429,
+      "grad_norm": 0.33749696612358093,
+      "learning_rate": 0.00019379111842105263,
+      "loss": 0.4784,
+      "step": 1132
+    },
+    {
+      "epoch": 1.2569685203161836,
+      "grad_norm": 0.26460304856300354,
+      "learning_rate": 0.00019366776315789472,
+      "loss": 0.5015,
+      "step": 1133
+    },
+    {
+      "epoch": 1.2580779364859243,
+      "grad_norm": 0.2513628602027893,
+      "learning_rate": 0.00019354440789473684,
+      "loss": 0.3158,
+      "step": 1134
+    },
+    {
+      "epoch": 1.259187352655665,
+      "grad_norm": 0.23956595361232758,
+      "learning_rate": 0.00019342105263157894,
+      "loss": 0.4248,
+      "step": 1135
+    },
+    {
+      "epoch": 1.2602967688254056,
+      "grad_norm": 0.33675435185432434,
+      "learning_rate": 0.00019329769736842103,
+      "loss": 0.5479,
+      "step": 1136
+    },
+    {
+      "epoch": 1.2614061849951463,
+      "grad_norm": 0.2596627473831177,
+      "learning_rate": 0.00019317434210526315,
+      "loss": 0.4415,
+      "step": 1137
+    },
+    {
+      "epoch": 1.262515601164887,
+      "grad_norm": 0.3592386543750763,
+      "learning_rate": 0.00019305098684210524,
+      "loss": 0.5607,
+      "step": 1138
+    },
+    {
+      "epoch": 1.2636250173346277,
+      "grad_norm": 0.3122079074382782,
+      "learning_rate": 0.00019292763157894734,
+      "loss": 0.891,
+      "step": 1139
+    },
+    {
+      "epoch": 1.2647344335043682,
+      "grad_norm": 0.31016677618026733,
+      "learning_rate": 0.00019280427631578943,
+      "loss": 0.5598,
+      "step": 1140
+    },
+    {
+      "epoch": 1.2658438496741091,
+      "grad_norm": 0.28511670231819153,
+      "learning_rate": 0.00019268092105263158,
+      "loss": 0.5666,
+      "step": 1141
+    },
+    {
+      "epoch": 1.2669532658438496,
+      "grad_norm": 0.3194934129714966,
+      "learning_rate": 0.00019255756578947367,
+      "loss": 0.6659,
+      "step": 1142
+    },
+    {
+      "epoch": 1.2680626820135903,
+      "grad_norm": 0.2699849307537079,
+      "learning_rate": 0.0001924342105263158,
+      "loss": 0.3064,
+      "step": 1143
+    },
+    {
+      "epoch": 1.269172098183331,
+      "grad_norm": 0.28152981400489807,
+      "learning_rate": 0.00019231085526315788,
+      "loss": 0.5019,
+      "step": 1144
+    },
+    {
+      "epoch": 1.2702815143530717,
+      "grad_norm": 0.34371238946914673,
+      "learning_rate": 0.00019218749999999998,
+      "loss": 0.4843,
+      "step": 1145
+    },
+    {
+      "epoch": 1.2713909305228124,
+      "grad_norm": 0.2585979402065277,
+      "learning_rate": 0.0001920641447368421,
+      "loss": 0.4984,
+      "step": 1146
+    },
+    {
+      "epoch": 1.272500346692553,
+      "grad_norm": 0.32947319746017456,
+      "learning_rate": 0.0001919407894736842,
+      "loss": 0.5213,
+      "step": 1147
+    },
+    {
+      "epoch": 1.2736097628622938,
+      "grad_norm": 0.22969017922878265,
+      "learning_rate": 0.00019181743421052628,
+      "loss": 0.429,
+      "step": 1148
+    },
+    {
+      "epoch": 1.2747191790320345,
+      "grad_norm": 0.22548457980155945,
+      "learning_rate": 0.0001916940789473684,
+      "loss": 0.3659,
+      "step": 1149
+    },
+    {
+      "epoch": 1.2758285952017752,
+      "grad_norm": 0.2924930155277252,
+      "learning_rate": 0.00019157072368421052,
+      "loss": 0.4644,
+      "step": 1150
+    },
+    {
+      "epoch": 1.2769380113715156,
+      "grad_norm": 0.31746160984039307,
+      "learning_rate": 0.00019144736842105262,
+      "loss": 0.5486,
+      "step": 1151
+    },
+    {
+      "epoch": 1.2780474275412563,
+      "grad_norm": 0.3976684808731079,
+      "learning_rate": 0.00019132401315789474,
+      "loss": 0.6632,
+      "step": 1152
+    },
+    {
+      "epoch": 1.279156843710997,
+      "grad_norm": 0.31601187586784363,
+      "learning_rate": 0.00019120065789473683,
+      "loss": 0.5185,
+      "step": 1153
+    },
+    {
+      "epoch": 1.2802662598807377,
+      "grad_norm": 0.2981434762477875,
+      "learning_rate": 0.00019107730263157892,
+      "loss": 0.4975,
+      "step": 1154
+    },
+    {
+      "epoch": 1.2813756760504784,
+      "grad_norm": 0.2698228657245636,
+      "learning_rate": 0.00019095394736842104,
+      "loss": 0.6271,
+      "step": 1155
+    },
+    {
+      "epoch": 1.2824850922202191,
+      "grad_norm": 0.3439400792121887,
+      "learning_rate": 0.00019083059210526314,
+      "loss": 0.6612,
+      "step": 1156
+    },
+    {
+      "epoch": 1.2835945083899598,
+      "grad_norm": 0.27007582783699036,
+      "learning_rate": 0.00019070723684210523,
+      "loss": 0.5057,
+      "step": 1157
+    },
+    {
+      "epoch": 1.2847039245597005,
+      "grad_norm": 0.29291701316833496,
+      "learning_rate": 0.00019058388157894735,
+      "loss": 0.5788,
+      "step": 1158
+    },
+    {
+      "epoch": 1.2858133407294412,
+      "grad_norm": 0.41224008798599243,
+      "learning_rate": 0.00019046052631578944,
+      "loss": 0.6633,
+      "step": 1159
+    },
+    {
+      "epoch": 1.2869227568991817,
+      "grad_norm": 0.3880039155483246,
+      "learning_rate": 0.00019033717105263156,
+      "loss": 0.5065,
+      "step": 1160
+    },
+    {
+      "epoch": 1.2880321730689226,
+      "grad_norm": 0.3320450186729431,
+      "learning_rate": 0.00019021381578947368,
+      "loss": 0.6414,
+      "step": 1161
+    },
+    {
+      "epoch": 1.289141589238663,
+      "grad_norm": 0.2895980477333069,
+      "learning_rate": 0.00019009046052631578,
+      "loss": 0.842,
+      "step": 1162
+    },
+    {
+      "epoch": 1.2902510054084038,
+      "grad_norm": 0.2721669673919678,
+      "learning_rate": 0.00018996710526315787,
+      "loss": 0.4492,
+      "step": 1163
+    },
+    {
+      "epoch": 1.2913604215781445,
+      "grad_norm": 0.28013283014297485,
+      "learning_rate": 0.00018984375,
+      "loss": 0.3661,
+      "step": 1164
+    },
+    {
+      "epoch": 1.2924698377478852,
+      "grad_norm": 0.22734491527080536,
+      "learning_rate": 0.00018972039473684208,
+      "loss": 0.5751,
+      "step": 1165
+    },
+    {
+      "epoch": 1.2935792539176258,
+      "grad_norm": 0.2372591197490692,
+      "learning_rate": 0.00018959703947368418,
+      "loss": 0.4572,
+      "step": 1166
+    },
+    {
+      "epoch": 1.2946886700873665,
+      "grad_norm": 0.3541167080402374,
+      "learning_rate": 0.0001894736842105263,
+      "loss": 0.6481,
+      "step": 1167
+    },
+    {
+      "epoch": 1.2957980862571072,
+      "grad_norm": 0.23032468557357788,
+      "learning_rate": 0.0001893503289473684,
+      "loss": 0.3583,
+      "step": 1168
+    },
+    {
+      "epoch": 1.296907502426848,
+      "grad_norm": 0.20932866632938385,
+      "learning_rate": 0.00018922697368421054,
+      "loss": 0.3898,
+      "step": 1169
+    },
+    {
+      "epoch": 1.2980169185965886,
+      "grad_norm": 0.2942456603050232,
+      "learning_rate": 0.00018910361842105263,
+      "loss": 0.4076,
+      "step": 1170
+    },
+    {
+      "epoch": 1.299126334766329,
+      "grad_norm": 0.28456342220306396,
+      "learning_rate": 0.00018898026315789472,
+      "loss": 0.2953,
+      "step": 1171
+    },
+    {
+      "epoch": 1.30023575093607,
+      "grad_norm": 0.39009782671928406,
+      "learning_rate": 0.00018885690789473684,
+      "loss": 0.511,
+      "step": 1172
+    },
+    {
+      "epoch": 1.3013451671058105,
+      "grad_norm": 0.22601839900016785,
+      "learning_rate": 0.00018873355263157894,
+      "loss": 0.5164,
+      "step": 1173
+    },
+    {
+      "epoch": 1.3024545832755512,
+      "grad_norm": 0.23257453739643097,
+      "learning_rate": 0.00018861019736842103,
+      "loss": 0.4081,
+      "step": 1174
+    },
+    {
+      "epoch": 1.3035639994452919,
+      "grad_norm": 0.2568961977958679,
+      "learning_rate": 0.00018848684210526312,
+      "loss": 0.4208,
+      "step": 1175
+    },
+    {
+      "epoch": 1.3046734156150326,
+      "grad_norm": 0.33718129992485046,
+      "learning_rate": 0.00018836348684210524,
+      "loss": 0.3738,
+      "step": 1176
+    },
+    {
+      "epoch": 1.3057828317847733,
+      "grad_norm": 0.24113743007183075,
+      "learning_rate": 0.00018824013157894734,
+      "loss": 0.4668,
+      "step": 1177
+    },
+    {
+      "epoch": 1.306892247954514,
+      "grad_norm": 0.2619990110397339,
+      "learning_rate": 0.00018811677631578943,
+      "loss": 0.6427,
+      "step": 1178
+    },
+    {
+      "epoch": 1.3080016641242547,
+      "grad_norm": 0.24512606859207153,
+      "learning_rate": 0.00018799342105263158,
+      "loss": 0.4494,
+      "step": 1179
+    },
+    {
+      "epoch": 1.3091110802939954,
+      "grad_norm": 0.23559564352035522,
+      "learning_rate": 0.00018787006578947367,
+      "loss": 0.4123,
+      "step": 1180
+    },
+    {
+      "epoch": 1.310220496463736,
+      "grad_norm": 0.24036440253257751,
+      "learning_rate": 0.0001877467105263158,
+      "loss": 0.4689,
+      "step": 1181
+    },
+    {
+      "epoch": 1.3113299126334765,
+      "grad_norm": 0.29898273944854736,
+      "learning_rate": 0.00018762335526315788,
+      "loss": 0.2823,
+      "step": 1182
+    },
+    {
+      "epoch": 1.3124393288032172,
+      "grad_norm": 0.2730746567249298,
+      "learning_rate": 0.00018749999999999998,
+      "loss": 0.7834,
+      "step": 1183
+    },
+    {
+      "epoch": 1.313548744972958,
+      "grad_norm": 0.30577996373176575,
+      "learning_rate": 0.0001873766447368421,
+      "loss": 0.6518,
+      "step": 1184
+    },
+    {
+      "epoch": 1.3146581611426986,
+      "grad_norm": 0.1852562427520752,
+      "learning_rate": 0.0001872532894736842,
+      "loss": 0.3995,
+      "step": 1185
+    },
+    {
+      "epoch": 1.3157675773124393,
+      "grad_norm": 0.3734980821609497,
+      "learning_rate": 0.00018712993421052628,
+      "loss": 0.8202,
+      "step": 1186
+    },
+    {
+      "epoch": 1.31687699348218,
+      "grad_norm": 0.1997169852256775,
+      "learning_rate": 0.0001870065789473684,
+      "loss": 0.5287,
+      "step": 1187
+    },
+    {
+      "epoch": 1.3179864096519207,
+      "grad_norm": 0.29843178391456604,
+      "learning_rate": 0.00018688322368421052,
+      "loss": 0.4611,
+      "step": 1188
+    },
+    {
+      "epoch": 1.3190958258216614,
+      "grad_norm": 0.24156174063682556,
+      "learning_rate": 0.00018675986842105262,
+      "loss": 0.7034,
+      "step": 1189
+    },
+    {
+      "epoch": 1.320205241991402,
+      "grad_norm": 0.3248981237411499,
+      "learning_rate": 0.00018663651315789474,
+      "loss": 0.562,
+      "step": 1190
+    },
+    {
+      "epoch": 1.3213146581611426,
+      "grad_norm": 0.22303232550621033,
+      "learning_rate": 0.00018651315789473683,
+      "loss": 0.428,
+      "step": 1191
+    },
+    {
+      "epoch": 1.3224240743308835,
+      "grad_norm": 0.32691171765327454,
+      "learning_rate": 0.00018638980263157892,
+      "loss": 0.5038,
+      "step": 1192
+    },
+    {
+      "epoch": 1.323533490500624,
+      "grad_norm": 0.23463557660579681,
+      "learning_rate": 0.00018626644736842104,
+      "loss": 0.5723,
+      "step": 1193
+    },
+    {
+      "epoch": 1.3246429066703647,
+      "grad_norm": 0.23432673513889313,
+      "learning_rate": 0.00018614309210526314,
+      "loss": 0.2767,
+      "step": 1194
+    },
+    {
+      "epoch": 1.3257523228401054,
+      "grad_norm": 0.30433669686317444,
+      "learning_rate": 0.00018601973684210523,
+      "loss": 0.5931,
+      "step": 1195
+    },
+    {
+      "epoch": 1.326861739009846,
+      "grad_norm": 0.2979690432548523,
+      "learning_rate": 0.00018589638157894735,
+      "loss": 0.6678,
+      "step": 1196
+    },
+    {
+      "epoch": 1.3279711551795867,
+      "grad_norm": 0.279066801071167,
+      "learning_rate": 0.00018577302631578944,
+      "loss": 0.6302,
+      "step": 1197
+    },
+    {
+      "epoch": 1.3290805713493274,
+      "grad_norm": 0.28714966773986816,
+      "learning_rate": 0.00018564967105263156,
+      "loss": 0.494,
+      "step": 1198
+    },
+    {
+      "epoch": 1.3301899875190681,
+      "grad_norm": 0.21045270562171936,
+      "learning_rate": 0.00018552631578947368,
+      "loss": 0.5388,
+      "step": 1199
+    },
+    {
+      "epoch": 1.3312994036888088,
+      "grad_norm": 0.27514612674713135,
+      "learning_rate": 0.00018540296052631578,
+      "loss": 0.3849,
+      "step": 1200
+    },
+    {
+      "epoch": 1.3324088198585495,
+      "grad_norm": 0.3071988523006439,
+      "learning_rate": 0.00018527960526315787,
+      "loss": 0.4609,
+      "step": 1201
+    },
+    {
+      "epoch": 1.33351823602829,
+      "grad_norm": 0.251871794462204,
+      "learning_rate": 0.00018515625,
+      "loss": 0.7988,
+      "step": 1202
+    },
+    {
+      "epoch": 1.334627652198031,
+      "grad_norm": 0.273266077041626,
+      "learning_rate": 0.00018503289473684208,
+      "loss": 0.3886,
+      "step": 1203
+    },
+    {
+      "epoch": 1.3357370683677714,
+      "grad_norm": 0.28145653009414673,
+      "learning_rate": 0.00018490953947368418,
+      "loss": 0.556,
+      "step": 1204
+    },
+    {
+      "epoch": 1.336846484537512,
+      "grad_norm": 0.23886194825172424,
+      "learning_rate": 0.0001847861842105263,
+      "loss": 0.4633,
+      "step": 1205
+    },
+    {
+      "epoch": 1.3379559007072528,
+      "grad_norm": 0.2713840901851654,
+      "learning_rate": 0.0001846628289473684,
+      "loss": 0.6669,
+      "step": 1206
+    },
+    {
+      "epoch": 1.3390653168769935,
+      "grad_norm": 0.19628193974494934,
+      "learning_rate": 0.00018453947368421054,
+      "loss": 0.5268,
+      "step": 1207
+    },
+    {
+      "epoch": 1.3401747330467342,
+      "grad_norm": 0.32730063796043396,
+      "learning_rate": 0.00018441611842105263,
+      "loss": 0.4295,
+      "step": 1208
+    },
+    {
+      "epoch": 1.3412841492164749,
+      "grad_norm": 0.244260773062706,
+      "learning_rate": 0.00018429276315789472,
+      "loss": 0.4108,
+      "step": 1209
+    },
+    {
+      "epoch": 1.3423935653862156,
+      "grad_norm": 0.30566221475601196,
+      "learning_rate": 0.00018416940789473684,
+      "loss": 0.8457,
+      "step": 1210
+    },
+    {
+      "epoch": 1.343502981555956,
+      "grad_norm": 0.2607499659061432,
+      "learning_rate": 0.00018404605263157894,
+      "loss": 0.4235,
+      "step": 1211
+    },
+    {
+      "epoch": 1.344612397725697,
+      "grad_norm": 0.359625905752182,
+      "learning_rate": 0.00018392269736842103,
+      "loss": 0.5669,
+      "step": 1212
+    },
+    {
+      "epoch": 1.3457218138954374,
+      "grad_norm": 0.2476588487625122,
+      "learning_rate": 0.00018379934210526312,
+      "loss": 0.3539,
+      "step": 1213
+    },
+    {
+      "epoch": 1.3468312300651781,
+      "grad_norm": 0.20041054487228394,
+      "learning_rate": 0.00018367598684210524,
+      "loss": 0.3843,
+      "step": 1214
+    },
+    {
+      "epoch": 1.3479406462349188,
+      "grad_norm": 0.4561375677585602,
+      "learning_rate": 0.00018355263157894734,
+      "loss": 0.6868,
+      "step": 1215
+    },
+    {
+      "epoch": 1.3490500624046595,
+      "grad_norm": 0.24837137758731842,
+      "learning_rate": 0.00018342927631578943,
+      "loss": 0.4685,
+      "step": 1216
+    },
+    {
+      "epoch": 1.3501594785744002,
+      "grad_norm": 0.3139680027961731,
+      "learning_rate": 0.00018330592105263158,
+      "loss": 0.5152,
+      "step": 1217
+    },
+    {
+      "epoch": 1.351268894744141,
+      "grad_norm": 0.2819393277168274,
+      "learning_rate": 0.00018318256578947367,
+      "loss": 0.4204,
+      "step": 1218
+    },
+    {
+      "epoch": 1.3523783109138816,
+      "grad_norm": 0.2980377972126007,
+      "learning_rate": 0.0001830592105263158,
+      "loss": 0.4413,
+      "step": 1219
+    },
+    {
+      "epoch": 1.3534877270836223,
+      "grad_norm": 0.29253172874450684,
+      "learning_rate": 0.00018293585526315788,
+      "loss": 0.5169,
+      "step": 1220
+    },
+    {
+      "epoch": 1.354597143253363,
+      "grad_norm": 0.3686385154724121,
+      "learning_rate": 0.00018281249999999998,
+      "loss": 0.4958,
+      "step": 1221
+    },
+    {
+      "epoch": 1.3557065594231035,
+      "grad_norm": 0.29562124609947205,
+      "learning_rate": 0.0001826891447368421,
+      "loss": 0.5406,
+      "step": 1222
+    },
+    {
+      "epoch": 1.3568159755928444,
+      "grad_norm": 0.21728160977363586,
+      "learning_rate": 0.0001825657894736842,
+      "loss": 0.412,
+      "step": 1223
+    },
+    {
+      "epoch": 1.3579253917625849,
+      "grad_norm": 0.23864157497882843,
+      "learning_rate": 0.00018244243421052628,
+      "loss": 0.4752,
+      "step": 1224
+    },
+    {
+      "epoch": 1.3590348079323256,
+      "grad_norm": 0.24741685390472412,
+      "learning_rate": 0.00018231907894736838,
+      "loss": 0.5245,
+      "step": 1225
+    },
+    {
+      "epoch": 1.3601442241020663,
+      "grad_norm": 0.2409430295228958,
+      "learning_rate": 0.00018219572368421053,
+      "loss": 0.4216,
+      "step": 1226
+    },
+    {
+      "epoch": 1.361253640271807,
+      "grad_norm": 0.25885629653930664,
+      "learning_rate": 0.00018207236842105262,
+      "loss": 0.4052,
+      "step": 1227
+    },
+    {
+      "epoch": 1.3623630564415476,
+      "grad_norm": 0.28579777479171753,
+      "learning_rate": 0.00018194901315789474,
+      "loss": 0.4137,
+      "step": 1228
+    },
+    {
+      "epoch": 1.3634724726112883,
+      "grad_norm": 0.23780061304569244,
+      "learning_rate": 0.00018182565789473683,
+      "loss": 0.4971,
+      "step": 1229
+    },
+    {
+      "epoch": 1.364581888781029,
+      "grad_norm": 0.32252663373947144,
+      "learning_rate": 0.00018170230263157893,
+      "loss": 0.5721,
+      "step": 1230
+    },
+    {
+      "epoch": 1.3656913049507697,
+      "grad_norm": 0.3731588125228882,
+      "learning_rate": 0.00018157894736842105,
+      "loss": 0.4581,
+      "step": 1231
+    },
+    {
+      "epoch": 1.3668007211205104,
+      "grad_norm": 0.15041133761405945,
+      "learning_rate": 0.00018145559210526314,
+      "loss": 0.3275,
+      "step": 1232
+    },
+    {
+      "epoch": 1.367910137290251,
+      "grad_norm": 0.4989398419857025,
+      "learning_rate": 0.00018133223684210523,
+      "loss": 0.3791,
+      "step": 1233
+    },
+    {
+      "epoch": 1.3690195534599916,
+      "grad_norm": 0.4500264525413513,
+      "learning_rate": 0.00018120888157894735,
+      "loss": 0.4119,
+      "step": 1234
+    },
+    {
+      "epoch": 1.3701289696297323,
+      "grad_norm": 0.2674502432346344,
+      "learning_rate": 0.00018108552631578945,
+      "loss": 0.6938,
+      "step": 1235
+    },
+    {
+      "epoch": 1.371238385799473,
+      "grad_norm": 0.363046258687973,
+      "learning_rate": 0.00018096217105263157,
+      "loss": 0.4881,
+      "step": 1236
+    },
+    {
+      "epoch": 1.3723478019692137,
+      "grad_norm": 0.38243699073791504,
+      "learning_rate": 0.00018083881578947369,
+      "loss": 0.6211,
+      "step": 1237
+    },
+    {
+      "epoch": 1.3734572181389544,
+      "grad_norm": 0.25102144479751587,
+      "learning_rate": 0.00018071546052631578,
+      "loss": 0.4659,
+      "step": 1238
+    },
+    {
+      "epoch": 1.374566634308695,
+      "grad_norm": 0.546259343624115,
+      "learning_rate": 0.00018059210526315787,
+      "loss": 0.4187,
+      "step": 1239
+    },
+    {
+      "epoch": 1.3756760504784358,
+      "grad_norm": 0.3232324719429016,
+      "learning_rate": 0.00018046875,
+      "loss": 0.457,
+      "step": 1240
+    },
+    {
+      "epoch": 1.3767854666481765,
+      "grad_norm": 0.2647280693054199,
+      "learning_rate": 0.00018034539473684209,
+      "loss": 0.5801,
+      "step": 1241
+    },
+    {
+      "epoch": 1.377894882817917,
+      "grad_norm": 0.43696561455726624,
+      "learning_rate": 0.00018022203947368418,
+      "loss": 0.5091,
+      "step": 1242
+    },
+    {
+      "epoch": 1.3790042989876579,
+      "grad_norm": 0.25493887066841125,
+      "learning_rate": 0.0001800986842105263,
+      "loss": 0.4163,
+      "step": 1243
+    },
+    {
+      "epoch": 1.3801137151573983,
+      "grad_norm": 0.32590964436531067,
+      "learning_rate": 0.0001799753289473684,
+      "loss": 0.4611,
+      "step": 1244
+    },
+    {
+      "epoch": 1.381223131327139,
+      "grad_norm": 0.28104108572006226,
+      "learning_rate": 0.00017985197368421054,
+      "loss": 0.3906,
+      "step": 1245
+    },
+    {
+      "epoch": 1.3823325474968797,
+      "grad_norm": 0.3454284369945526,
+      "learning_rate": 0.00017972861842105263,
+      "loss": 0.5055,
+      "step": 1246
+    },
+    {
+      "epoch": 1.3834419636666204,
+      "grad_norm": 0.3004135489463806,
+      "learning_rate": 0.00017960526315789473,
+      "loss": 0.5906,
+      "step": 1247
+    },
+    {
+      "epoch": 1.3845513798363611,
+      "grad_norm": 0.3798172175884247,
+      "learning_rate": 0.00017948190789473682,
+      "loss": 0.3812,
+      "step": 1248
+    },
+    {
+      "epoch": 1.3856607960061018,
+      "grad_norm": 0.23392631113529205,
+      "learning_rate": 0.00017935855263157894,
+      "loss": 0.3541,
+      "step": 1249
+    },
+    {
+      "epoch": 1.3867702121758425,
+      "grad_norm": 0.28611382842063904,
+      "learning_rate": 0.00017923519736842103,
+      "loss": 0.3994,
+      "step": 1250
+    },
+    {
+      "epoch": 1.3878796283455832,
+      "grad_norm": 0.23118913173675537,
+      "learning_rate": 0.00017911184210526313,
+      "loss": 0.5034,
+      "step": 1251
+    },
+    {
+      "epoch": 1.388989044515324,
+      "grad_norm": 0.3240826725959778,
+      "learning_rate": 0.00017898848684210525,
+      "loss": 0.7616,
+      "step": 1252
+    },
+    {
+      "epoch": 1.3900984606850644,
+      "grad_norm": 0.35286498069763184,
+      "learning_rate": 0.00017886513157894734,
+      "loss": 0.7485,
+      "step": 1253
+    },
+    {
+      "epoch": 1.3912078768548053,
+      "grad_norm": 0.23060156404972076,
+      "learning_rate": 0.00017874177631578943,
+      "loss": 0.4549,
+      "step": 1254
+    },
+    {
+      "epoch": 1.3923172930245458,
+      "grad_norm": 0.3174906075000763,
+      "learning_rate": 0.00017861842105263158,
+      "loss": 0.4251,
+      "step": 1255
+    },
+    {
+      "epoch": 1.3934267091942865,
+      "grad_norm": 0.24730369448661804,
+      "learning_rate": 0.00017849506578947367,
+      "loss": 0.4357,
+      "step": 1256
+    },
+    {
+      "epoch": 1.3945361253640272,
+      "grad_norm": 0.2686748206615448,
+      "learning_rate": 0.0001783717105263158,
+      "loss": 0.3928,
+      "step": 1257
+    },
+    {
+      "epoch": 1.3956455415337679,
+      "grad_norm": 0.2948950231075287,
+      "learning_rate": 0.00017824835526315789,
+      "loss": 0.6208,
+      "step": 1258
+    },
+    {
+      "epoch": 1.3967549577035085,
+      "grad_norm": 0.30887314677238464,
+      "learning_rate": 0.00017812499999999998,
+      "loss": 0.4876,
+      "step": 1259
+    },
+    {
+      "epoch": 1.3978643738732492,
+      "grad_norm": 0.3009176552295685,
+      "learning_rate": 0.0001780016447368421,
+      "loss": 0.3689,
+      "step": 1260
+    },
+    {
+      "epoch": 1.39897379004299,
+      "grad_norm": 0.24447450041770935,
+      "learning_rate": 0.0001778782894736842,
+      "loss": 0.3824,
+      "step": 1261
+    },
+    {
+      "epoch": 1.4000832062127304,
+      "grad_norm": 0.2673259973526001,
+      "learning_rate": 0.00017775493421052629,
+      "loss": 0.5425,
+      "step": 1262
+    },
+    {
+      "epoch": 1.4011926223824713,
+      "grad_norm": 0.273612916469574,
+      "learning_rate": 0.00017763157894736838,
+      "loss": 0.3522,
+      "step": 1263
+    },
+    {
+      "epoch": 1.4023020385522118,
+      "grad_norm": 0.3372279405593872,
+      "learning_rate": 0.00017750822368421053,
+      "loss": 0.4351,
+      "step": 1264
+    },
+    {
+      "epoch": 1.4034114547219525,
+      "grad_norm": 0.23103949427604675,
+      "learning_rate": 0.00017738486842105262,
+      "loss": 0.4426,
+      "step": 1265
+    },
+    {
+      "epoch": 1.4045208708916932,
+      "grad_norm": 0.21246463060379028,
+      "learning_rate": 0.00017726151315789474,
+      "loss": 0.5233,
+      "step": 1266
+    },
+    {
+      "epoch": 1.405630287061434,
+      "grad_norm": 0.2304743230342865,
+      "learning_rate": 0.00017713815789473683,
+      "loss": 0.3163,
+      "step": 1267
+    },
+    {
+      "epoch": 1.4067397032311746,
+      "grad_norm": 0.35149502754211426,
+      "learning_rate": 0.00017701480263157893,
+      "loss": 0.6481,
+      "step": 1268
+    },
+    {
+      "epoch": 1.4078491194009153,
+      "grad_norm": 0.2509346604347229,
+      "learning_rate": 0.00017689144736842105,
+      "loss": 0.6448,
+      "step": 1269
+    },
+    {
+      "epoch": 1.408958535570656,
+      "grad_norm": 0.25470322370529175,
+      "learning_rate": 0.00017676809210526314,
+      "loss": 0.5486,
+      "step": 1270
+    },
+    {
+      "epoch": 1.4100679517403967,
+      "grad_norm": 0.24675153195858002,
+      "learning_rate": 0.00017664473684210523,
+      "loss": 0.4237,
+      "step": 1271
+    },
+    {
+      "epoch": 1.4111773679101374,
+      "grad_norm": 0.38935887813568115,
+      "learning_rate": 0.00017652138157894735,
+      "loss": 0.5769,
+      "step": 1272
+    },
+    {
+      "epoch": 1.4122867840798778,
+      "grad_norm": 0.22933362424373627,
+      "learning_rate": 0.00017639802631578945,
+      "loss": 0.4627,
+      "step": 1273
+    },
+    {
+      "epoch": 1.4133962002496188,
+      "grad_norm": 0.4441911280155182,
+      "learning_rate": 0.00017627467105263157,
+      "loss": 0.5397,
+      "step": 1274
+    },
+    {
+      "epoch": 1.4145056164193592,
+      "grad_norm": 0.24430730938911438,
+      "learning_rate": 0.0001761513157894737,
+      "loss": 0.4419,
+      "step": 1275
+    },
+    {
+      "epoch": 1.4156150325891,
+      "grad_norm": 0.280831515789032,
+      "learning_rate": 0.00017602796052631578,
+      "loss": 0.3993,
+      "step": 1276
+    },
+    {
+      "epoch": 1.4167244487588406,
+      "grad_norm": 0.32151225209236145,
+      "learning_rate": 0.00017590460526315787,
+      "loss": 0.5654,
+      "step": 1277
+    },
+    {
+      "epoch": 1.4178338649285813,
+      "grad_norm": 0.3048468828201294,
+      "learning_rate": 0.00017578125,
+      "loss": 0.4373,
+      "step": 1278
+    },
+    {
+      "epoch": 1.418943281098322,
+      "grad_norm": 0.44277939200401306,
+      "learning_rate": 0.0001756578947368421,
+      "loss": 0.4243,
+      "step": 1279
+    },
+    {
+      "epoch": 1.4200526972680627,
+      "grad_norm": 0.3722161054611206,
+      "learning_rate": 0.00017553453947368418,
+      "loss": 0.5524,
+      "step": 1280
+    },
+    {
+      "epoch": 1.4211621134378034,
+      "grad_norm": 0.24039240181446075,
+      "learning_rate": 0.0001754111842105263,
+      "loss": 0.4882,
+      "step": 1281
+    },
+    {
+      "epoch": 1.422271529607544,
+      "grad_norm": 0.34307026863098145,
+      "learning_rate": 0.0001752878289473684,
+      "loss": 0.5102,
+      "step": 1282
+    },
+    {
+      "epoch": 1.4233809457772848,
+      "grad_norm": 0.2590845227241516,
+      "learning_rate": 0.00017516447368421051,
+      "loss": 0.4354,
+      "step": 1283
+    },
+    {
+      "epoch": 1.4244903619470253,
+      "grad_norm": 0.25644242763519287,
+      "learning_rate": 0.00017504111842105263,
+      "loss": 0.4385,
+      "step": 1284
+    },
+    {
+      "epoch": 1.425599778116766,
+      "grad_norm": 0.25579833984375,
+      "learning_rate": 0.00017491776315789473,
+      "loss": 0.5522,
+      "step": 1285
+    },
+    {
+      "epoch": 1.4267091942865067,
+      "grad_norm": 0.26802870631217957,
+      "learning_rate": 0.00017479440789473682,
+      "loss": 0.369,
+      "step": 1286
+    },
+    {
+      "epoch": 1.4278186104562474,
+      "grad_norm": 0.28027182817459106,
+      "learning_rate": 0.00017467105263157894,
+      "loss": 0.485,
+      "step": 1287
+    },
+    {
+      "epoch": 1.428928026625988,
+      "grad_norm": 0.5787771344184875,
+      "learning_rate": 0.00017454769736842103,
+      "loss": 0.5496,
+      "step": 1288
+    },
+    {
+      "epoch": 1.4300374427957288,
+      "grad_norm": 0.24002310633659363,
+      "learning_rate": 0.00017442434210526313,
+      "loss": 0.4172,
+      "step": 1289
+    },
+    {
+      "epoch": 1.4311468589654694,
+      "grad_norm": 0.34913745522499084,
+      "learning_rate": 0.00017430098684210525,
+      "loss": 0.6889,
+      "step": 1290
+    },
+    {
+      "epoch": 1.4322562751352101,
+      "grad_norm": 0.2884618639945984,
+      "learning_rate": 0.00017417763157894734,
+      "loss": 0.4579,
+      "step": 1291
+    },
+    {
+      "epoch": 1.4333656913049508,
+      "grad_norm": 0.2817135751247406,
+      "learning_rate": 0.00017405427631578943,
+      "loss": 0.4721,
+      "step": 1292
+    },
+    {
+      "epoch": 1.4344751074746913,
+      "grad_norm": 0.2580573558807373,
+      "learning_rate": 0.00017393092105263158,
+      "loss": 0.4966,
+      "step": 1293
+    },
+    {
+      "epoch": 1.4355845236444322,
+      "grad_norm": 0.3415181338787079,
+      "learning_rate": 0.00017380756578947367,
+      "loss": 0.4475,
+      "step": 1294
+    },
+    {
+      "epoch": 1.4366939398141727,
+      "grad_norm": 0.2799202799797058,
+      "learning_rate": 0.0001736842105263158,
+      "loss": 0.4047,
+      "step": 1295
+    },
+    {
+      "epoch": 1.4378033559839134,
+      "grad_norm": 0.3765754699707031,
+      "learning_rate": 0.0001735608552631579,
+      "loss": 0.4161,
+      "step": 1296
+    },
+    {
+      "epoch": 1.438912772153654,
+      "grad_norm": 0.2828143835067749,
+      "learning_rate": 0.00017343749999999998,
+      "loss": 0.544,
+      "step": 1297
+    },
+    {
+      "epoch": 1.4400221883233948,
+      "grad_norm": 0.4222780466079712,
+      "learning_rate": 0.00017331414473684207,
+      "loss": 0.4053,
+      "step": 1298
+    },
+    {
+      "epoch": 1.4411316044931355,
+      "grad_norm": 0.2224740982055664,
+      "learning_rate": 0.0001731907894736842,
+      "loss": 0.343,
+      "step": 1299
+    },
+    {
+      "epoch": 1.4422410206628762,
+      "grad_norm": 0.301512748003006,
+      "learning_rate": 0.0001730674342105263,
+      "loss": 0.5385,
+      "step": 1300
+    },
+    {
+      "epoch": 1.4433504368326169,
+      "grad_norm": 0.24199537932872772,
+      "learning_rate": 0.00017294407894736838,
+      "loss": 0.4618,
+      "step": 1301
+    },
+    {
+      "epoch": 1.4444598530023576,
+      "grad_norm": 0.2662793695926666,
+      "learning_rate": 0.00017282072368421053,
+      "loss": 0.3893,
+      "step": 1302
+    },
+    {
+      "epoch": 1.4455692691720983,
+      "grad_norm": 0.6027406454086304,
+      "learning_rate": 0.00017269736842105262,
+      "loss": 0.7918,
+      "step": 1303
+    },
+    {
+      "epoch": 1.4466786853418387,
+      "grad_norm": 0.33173203468322754,
+      "learning_rate": 0.00017257401315789474,
+      "loss": 0.3696,
+      "step": 1304
+    },
+    {
+      "epoch": 1.4477881015115797,
+      "grad_norm": 0.42119306325912476,
+      "learning_rate": 0.00017245065789473683,
+      "loss": 0.3504,
+      "step": 1305
+    },
+    {
+      "epoch": 1.4488975176813201,
+      "grad_norm": 0.33368048071861267,
+      "learning_rate": 0.00017232730263157893,
+      "loss": 0.543,
+      "step": 1306
+    },
+    {
+      "epoch": 1.4500069338510608,
+      "grad_norm": 0.3724362850189209,
+      "learning_rate": 0.00017220394736842105,
+      "loss": 0.483,
+      "step": 1307
+    },
+    {
+      "epoch": 1.4511163500208015,
+      "grad_norm": 0.28803780674934387,
+      "learning_rate": 0.00017208059210526314,
+      "loss": 0.6044,
+      "step": 1308
+    },
+    {
+      "epoch": 1.4522257661905422,
+      "grad_norm": 0.29041630029678345,
+      "learning_rate": 0.00017195723684210523,
+      "loss": 0.4351,
+      "step": 1309
+    },
+    {
+      "epoch": 1.453335182360283,
+      "grad_norm": 0.27980196475982666,
+      "learning_rate": 0.00017183388157894735,
+      "loss": 0.2806,
+      "step": 1310
+    },
+    {
+      "epoch": 1.4544445985300236,
+      "grad_norm": 0.5144875049591064,
+      "learning_rate": 0.00017171052631578945,
+      "loss": 0.6004,
+      "step": 1311
+    },
+    {
+      "epoch": 1.4555540146997643,
+      "grad_norm": 0.41391658782958984,
+      "learning_rate": 0.00017158717105263157,
+      "loss": 0.3426,
+      "step": 1312
+    },
+    {
+      "epoch": 1.456663430869505,
+      "grad_norm": 0.38778960704803467,
+      "learning_rate": 0.0001714638157894737,
+      "loss": 0.5716,
+      "step": 1313
+    },
+    {
+      "epoch": 1.4577728470392457,
+      "grad_norm": 0.29660889506340027,
+      "learning_rate": 0.00017134046052631578,
+      "loss": 0.6084,
+      "step": 1314
+    },
+    {
+      "epoch": 1.4588822632089862,
+      "grad_norm": 0.24347136914730072,
+      "learning_rate": 0.00017121710526315787,
+      "loss": 0.6538,
+      "step": 1315
+    },
+    {
+      "epoch": 1.4599916793787269,
+      "grad_norm": 0.2733883857727051,
+      "learning_rate": 0.00017109375,
+      "loss": 0.5797,
+      "step": 1316
+    },
+    {
+      "epoch": 1.4611010955484676,
+      "grad_norm": 0.33641162514686584,
+      "learning_rate": 0.0001709703947368421,
+      "loss": 0.4837,
+      "step": 1317
+    },
+    {
+      "epoch": 1.4622105117182083,
+      "grad_norm": 0.24747195839881897,
+      "learning_rate": 0.00017084703947368418,
+      "loss": 0.4617,
+      "step": 1318
+    },
+    {
+      "epoch": 1.463319927887949,
+      "grad_norm": 0.2453019767999649,
+      "learning_rate": 0.0001707236842105263,
+      "loss": 0.5242,
+      "step": 1319
+    },
+    {
+      "epoch": 1.4644293440576897,
+      "grad_norm": 0.2680438160896301,
+      "learning_rate": 0.0001706003289473684,
+      "loss": 0.3985,
+      "step": 1320
+    },
+    {
+      "epoch": 1.4655387602274303,
+      "grad_norm": 0.22616985440254211,
+      "learning_rate": 0.00017047697368421051,
+      "loss": 0.5428,
+      "step": 1321
+    },
+    {
+      "epoch": 1.466648176397171,
+      "grad_norm": 0.25485649704933167,
+      "learning_rate": 0.00017035361842105264,
+      "loss": 0.4411,
+      "step": 1322
+    },
+    {
+      "epoch": 1.4677575925669117,
+      "grad_norm": 0.19676417112350464,
+      "learning_rate": 0.00017023026315789473,
+      "loss": 0.4987,
+      "step": 1323
+    },
+    {
+      "epoch": 1.4688670087366522,
+      "grad_norm": 0.23980014026165009,
+      "learning_rate": 0.00017010690789473682,
+      "loss": 0.2853,
+      "step": 1324
+    },
+    {
+      "epoch": 1.4699764249063931,
+      "grad_norm": 0.31723615527153015,
+      "learning_rate": 0.00016998355263157894,
+      "loss": 0.3271,
+      "step": 1325
+    },
+    {
+      "epoch": 1.4710858410761336,
+      "grad_norm": 0.2783150374889374,
+      "learning_rate": 0.00016986019736842103,
+      "loss": 0.4648,
+      "step": 1326
+    },
+    {
+      "epoch": 1.4721952572458743,
+      "grad_norm": 0.3233512341976166,
+      "learning_rate": 0.00016973684210526313,
+      "loss": 0.4557,
+      "step": 1327
+    },
+    {
+      "epoch": 1.473304673415615,
+      "grad_norm": 0.2481250911951065,
+      "learning_rate": 0.00016961348684210525,
+      "loss": 0.5224,
+      "step": 1328
+    },
+    {
+      "epoch": 1.4744140895853557,
+      "grad_norm": 0.317909300327301,
+      "learning_rate": 0.00016949013157894734,
+      "loss": 0.642,
+      "step": 1329
+    },
+    {
+      "epoch": 1.4755235057550964,
+      "grad_norm": 0.25988101959228516,
+      "learning_rate": 0.00016936677631578943,
+      "loss": 0.3956,
+      "step": 1330
+    },
+    {
+      "epoch": 1.476632921924837,
+      "grad_norm": 0.2583842873573303,
+      "learning_rate": 0.00016924342105263158,
+      "loss": 0.4031,
+      "step": 1331
+    },
+    {
+      "epoch": 1.4777423380945778,
+      "grad_norm": 0.2636638283729553,
+      "learning_rate": 0.00016912006578947368,
+      "loss": 0.4351,
+      "step": 1332
+    },
+    {
+      "epoch": 1.4788517542643185,
+      "grad_norm": 0.2658294439315796,
+      "learning_rate": 0.00016899671052631577,
+      "loss": 0.4895,
+      "step": 1333
+    },
+    {
+      "epoch": 1.4799611704340592,
+      "grad_norm": 0.192026287317276,
+      "learning_rate": 0.0001688733552631579,
+      "loss": 0.5236,
+      "step": 1334
+    },
+    {
+      "epoch": 1.4810705866037996,
+      "grad_norm": 0.2988239824771881,
+      "learning_rate": 0.00016874999999999998,
+      "loss": 0.4434,
+      "step": 1335
+    },
+    {
+      "epoch": 1.4821800027735406,
+      "grad_norm": 0.32316145300865173,
+      "learning_rate": 0.00016862664473684207,
+      "loss": 0.5385,
+      "step": 1336
+    },
+    {
+      "epoch": 1.483289418943281,
+      "grad_norm": 0.3037776052951813,
+      "learning_rate": 0.0001685032894736842,
+      "loss": 0.3864,
+      "step": 1337
+    },
+    {
+      "epoch": 1.4843988351130217,
+      "grad_norm": 0.3819033205509186,
+      "learning_rate": 0.0001683799342105263,
+      "loss": 0.4825,
+      "step": 1338
+    },
+    {
+      "epoch": 1.4855082512827624,
+      "grad_norm": 0.346760630607605,
+      "learning_rate": 0.00016825657894736838,
+      "loss": 0.5131,
+      "step": 1339
+    },
+    {
+      "epoch": 1.4866176674525031,
+      "grad_norm": 0.27673614025115967,
+      "learning_rate": 0.00016813322368421053,
+      "loss": 0.4316,
+      "step": 1340
+    },
+    {
+      "epoch": 1.4877270836222438,
+      "grad_norm": 0.21999450027942657,
+      "learning_rate": 0.00016800986842105262,
+      "loss": 0.4106,
+      "step": 1341
+    },
+    {
+      "epoch": 1.4888364997919845,
+      "grad_norm": 0.30723118782043457,
+      "learning_rate": 0.00016788651315789474,
+      "loss": 0.5362,
+      "step": 1342
+    },
+    {
+      "epoch": 1.4899459159617252,
+      "grad_norm": 0.4165399372577667,
+      "learning_rate": 0.00016776315789473684,
+      "loss": 0.6185,
+      "step": 1343
+    },
+    {
+      "epoch": 1.4910553321314657,
+      "grad_norm": 0.2928377091884613,
+      "learning_rate": 0.00016763980263157893,
+      "loss": 0.4006,
+      "step": 1344
+    },
+    {
+      "epoch": 1.4921647483012066,
+      "grad_norm": 0.405435711145401,
+      "learning_rate": 0.00016751644736842105,
+      "loss": 0.4493,
+      "step": 1345
+    },
+    {
+      "epoch": 1.493274164470947,
+      "grad_norm": 0.2776191830635071,
+      "learning_rate": 0.00016739309210526314,
+      "loss": 0.4531,
+      "step": 1346
+    },
+    {
+      "epoch": 1.4943835806406878,
+      "grad_norm": 0.19967693090438843,
+      "learning_rate": 0.00016726973684210524,
+      "loss": 0.5397,
+      "step": 1347
+    },
+    {
+      "epoch": 1.4954929968104285,
+      "grad_norm": 0.22307896614074707,
+      "learning_rate": 0.00016714638157894733,
+      "loss": 0.6485,
+      "step": 1348
+    },
+    {
+      "epoch": 1.4966024129801692,
+      "grad_norm": 0.34726205468177795,
+      "learning_rate": 0.00016702302631578945,
+      "loss": 0.7418,
+      "step": 1349
+    },
+    {
+      "epoch": 1.4977118291499099,
+      "grad_norm": 0.2557240128517151,
+      "learning_rate": 0.00016689967105263157,
+      "loss": 0.2897,
+      "step": 1350
+    },
+    {
+      "epoch": 1.4988212453196506,
+      "grad_norm": 0.2983255386352539,
+      "learning_rate": 0.0001667763157894737,
+      "loss": 0.4885,
+      "step": 1351
+    },
+    {
+      "epoch": 1.4999306614893912,
+      "grad_norm": 0.2455969750881195,
+      "learning_rate": 0.00016665296052631578,
+      "loss": 0.5027,
+      "step": 1352
+    },
+    {
+      "epoch": 1.5010400776591317,
+      "grad_norm": 0.2705221474170685,
+      "learning_rate": 0.00016652960526315788,
+      "loss": 0.5434,
+      "step": 1353
+    },
+    {
+      "epoch": 1.5021494938288726,
+      "grad_norm": 0.27649638056755066,
+      "learning_rate": 0.00016640625,
+      "loss": 0.5844,
+      "step": 1354
+    },
+    {
+      "epoch": 1.5032589099986131,
+      "grad_norm": 0.28695183992385864,
+      "learning_rate": 0.0001662828947368421,
+      "loss": 0.4698,
+      "step": 1355
+    },
+    {
+      "epoch": 1.504368326168354,
+      "grad_norm": 0.3684578537940979,
+      "learning_rate": 0.00016615953947368418,
+      "loss": 0.5555,
+      "step": 1356
+    },
+    {
+      "epoch": 1.5054777423380945,
+      "grad_norm": 0.1808445006608963,
+      "learning_rate": 0.0001660361842105263,
+      "loss": 0.4829,
+      "step": 1357
+    },
+    {
+      "epoch": 1.5065871585078352,
+      "grad_norm": 0.37919580936431885,
+      "learning_rate": 0.0001659128289473684,
+      "loss": 0.4,
+      "step": 1358
+    },
+    {
+      "epoch": 1.507696574677576,
+      "grad_norm": 0.3226320743560791,
+      "learning_rate": 0.00016578947368421052,
+      "loss": 0.6081,
+      "step": 1359
+    },
+    {
+      "epoch": 1.5088059908473166,
+      "grad_norm": 0.378694623708725,
+      "learning_rate": 0.00016566611842105264,
+      "loss": 0.5211,
+      "step": 1360
+    },
+    {
+      "epoch": 1.5099154070170573,
+      "grad_norm": 0.418722003698349,
+      "learning_rate": 0.00016554276315789473,
+      "loss": 0.67,
+      "step": 1361
+    },
+    {
+      "epoch": 1.511024823186798,
+      "grad_norm": 0.3537873327732086,
+      "learning_rate": 0.00016541940789473682,
+      "loss": 0.6805,
+      "step": 1362
+    },
+    {
+      "epoch": 1.5121342393565387,
+      "grad_norm": 0.2202320247888565,
+      "learning_rate": 0.00016529605263157894,
+      "loss": 0.3873,
+      "step": 1363
+    },
+    {
+      "epoch": 1.5132436555262792,
+      "grad_norm": 0.36664825677871704,
+      "learning_rate": 0.00016517269736842104,
+      "loss": 0.4605,
+      "step": 1364
+    },
+    {
+      "epoch": 1.51435307169602,
+      "grad_norm": 0.25324442982673645,
+      "learning_rate": 0.00016504934210526313,
+      "loss": 0.6109,
+      "step": 1365
+    },
+    {
+      "epoch": 1.5154624878657605,
+      "grad_norm": 0.19155505299568176,
+      "learning_rate": 0.00016492598684210525,
+      "loss": 0.7548,
+      "step": 1366
+    },
+    {
+      "epoch": 1.5165719040355015,
+      "grad_norm": 0.3349843919277191,
+      "learning_rate": 0.00016480263157894734,
+      "loss": 0.3317,
+      "step": 1367
+    },
+    {
+      "epoch": 1.517681320205242,
+      "grad_norm": 0.26652729511260986,
+      "learning_rate": 0.00016467927631578944,
+      "loss": 0.6128,
+      "step": 1368
+    },
+    {
+      "epoch": 1.5187907363749826,
+      "grad_norm": 0.30250903964042664,
+      "learning_rate": 0.00016455592105263158,
+      "loss": 0.6961,
+      "step": 1369
+    },
+    {
+      "epoch": 1.5199001525447233,
+      "grad_norm": 0.4225883185863495,
+      "learning_rate": 0.00016443256578947368,
+      "loss": 0.5264,
+      "step": 1370
+    },
+    {
+      "epoch": 1.521009568714464,
+      "grad_norm": 0.331551730632782,
+      "learning_rate": 0.00016430921052631577,
+      "loss": 0.4245,
+      "step": 1371
+    },
+    {
+      "epoch": 1.5221189848842047,
+      "grad_norm": 0.2630516588687897,
+      "learning_rate": 0.0001641858552631579,
+      "loss": 0.4204,
+      "step": 1372
+    },
+    {
+      "epoch": 1.5232284010539454,
+      "grad_norm": 0.2230089157819748,
+      "learning_rate": 0.00016406249999999998,
+      "loss": 0.4686,
+      "step": 1373
+    },
+    {
+      "epoch": 1.524337817223686,
+      "grad_norm": 0.33370235562324524,
+      "learning_rate": 0.00016393914473684208,
+      "loss": 0.69,
+      "step": 1374
+    },
+    {
+      "epoch": 1.5254472333934266,
+      "grad_norm": 0.3383263349533081,
+      "learning_rate": 0.0001638157894736842,
+      "loss": 0.5046,
+      "step": 1375
+    },
+    {
+      "epoch": 1.5265566495631675,
+      "grad_norm": 0.3292944133281708,
+      "learning_rate": 0.0001636924342105263,
+      "loss": 0.5178,
+      "step": 1376
+    },
+    {
+      "epoch": 1.527666065732908,
+      "grad_norm": 0.2474631369113922,
+      "learning_rate": 0.00016356907894736838,
+      "loss": 0.3783,
+      "step": 1377
+    },
+    {
+      "epoch": 1.528775481902649,
+      "grad_norm": 0.306476354598999,
+      "learning_rate": 0.00016344572368421053,
+      "loss": 0.3773,
+      "step": 1378
+    },
+    {
+      "epoch": 1.5298848980723894,
+      "grad_norm": 0.2079583704471588,
+      "learning_rate": 0.00016332236842105262,
+      "loss": 0.4554,
+      "step": 1379
+    },
+    {
+      "epoch": 1.53099431424213,
+      "grad_norm": 0.4730626940727234,
+      "learning_rate": 0.00016319901315789474,
+      "loss": 0.3571,
+      "step": 1380
+    },
+    {
+      "epoch": 1.5321037304118708,
+      "grad_norm": 0.22660337388515472,
+      "learning_rate": 0.00016307565789473684,
+      "loss": 0.4086,
+      "step": 1381
+    },
+    {
+      "epoch": 1.5332131465816115,
+      "grad_norm": 0.444742888212204,
+      "learning_rate": 0.00016295230263157893,
+      "loss": 0.4406,
+      "step": 1382
+    },
+    {
+      "epoch": 1.5343225627513521,
+      "grad_norm": 0.34402474761009216,
+      "learning_rate": 0.00016282894736842102,
+      "loss": 0.6871,
+      "step": 1383
+    },
+    {
+      "epoch": 1.5354319789210926,
+      "grad_norm": 0.31494930386543274,
+      "learning_rate": 0.00016270559210526314,
+      "loss": 0.5023,
+      "step": 1384
+    },
+    {
+      "epoch": 1.5365413950908335,
+      "grad_norm": 0.24129949510097504,
+      "learning_rate": 0.00016258223684210524,
+      "loss": 0.3006,
+      "step": 1385
+    },
+    {
+      "epoch": 1.537650811260574,
+      "grad_norm": 0.37455177307128906,
+      "learning_rate": 0.00016245888157894733,
+      "loss": 0.3357,
+      "step": 1386
+    },
+    {
+      "epoch": 1.538760227430315,
+      "grad_norm": 0.22518782317638397,
+      "learning_rate": 0.00016233552631578945,
+      "loss": 0.4212,
+      "step": 1387
+    },
+    {
+      "epoch": 1.5398696436000554,
+      "grad_norm": 0.3566707372665405,
+      "learning_rate": 0.00016221217105263157,
+      "loss": 0.432,
+      "step": 1388
+    },
+    {
+      "epoch": 1.540979059769796,
+      "grad_norm": 0.3053068518638611,
+      "learning_rate": 0.0001620888157894737,
+      "loss": 0.3547,
+      "step": 1389
+    },
+    {
+      "epoch": 1.5420884759395368,
+      "grad_norm": 0.26762375235557556,
+      "learning_rate": 0.00016196546052631578,
+      "loss": 0.4258,
+      "step": 1390
+    },
+    {
+      "epoch": 1.5431978921092775,
+      "grad_norm": 0.44275879859924316,
+      "learning_rate": 0.00016184210526315788,
+      "loss": 0.5698,
+      "step": 1391
+    },
+    {
+      "epoch": 1.5443073082790182,
+      "grad_norm": 0.3341034948825836,
+      "learning_rate": 0.00016171875,
+      "loss": 0.6197,
+      "step": 1392
+    },
+    {
+      "epoch": 1.5454167244487589,
+      "grad_norm": 0.21536946296691895,
+      "learning_rate": 0.0001615953947368421,
+      "loss": 0.586,
+      "step": 1393
+    },
+    {
+      "epoch": 1.5465261406184996,
+      "grad_norm": 0.518312394618988,
+      "learning_rate": 0.00016147203947368418,
+      "loss": 0.663,
+      "step": 1394
+    },
+    {
+      "epoch": 1.54763555678824,
+      "grad_norm": 0.31936115026474,
+      "learning_rate": 0.0001613486842105263,
+      "loss": 0.5038,
+      "step": 1395
+    },
+    {
+      "epoch": 1.548744972957981,
+      "grad_norm": 0.4910357892513275,
+      "learning_rate": 0.0001612253289473684,
+      "loss": 0.5321,
+      "step": 1396
+    },
+    {
+      "epoch": 1.5498543891277214,
+      "grad_norm": 0.35702967643737793,
+      "learning_rate": 0.00016110197368421052,
+      "loss": 0.5254,
+      "step": 1397
+    },
+    {
+      "epoch": 1.5509638052974624,
+      "grad_norm": 0.46834543347358704,
+      "learning_rate": 0.00016097861842105264,
+      "loss": 0.5789,
+      "step": 1398
+    },
+    {
+      "epoch": 1.5520732214672028,
+      "grad_norm": 0.48472335934638977,
+      "learning_rate": 0.00016085526315789473,
+      "loss": 0.4901,
+      "step": 1399
+    },
+    {
+      "epoch": 1.5531826376369435,
+      "grad_norm": 0.28571802377700806,
+      "learning_rate": 0.00016073190789473682,
+      "loss": 0.458,
+      "step": 1400
+    },
+    {
+      "epoch": 1.5542920538066842,
+      "grad_norm": 0.30422210693359375,
+      "learning_rate": 0.00016060855263157894,
+      "loss": 0.4369,
+      "step": 1401
+    },
+    {
+      "epoch": 1.555401469976425,
+      "grad_norm": 0.34759703278541565,
+      "learning_rate": 0.00016048519736842104,
+      "loss": 0.5257,
+      "step": 1402
+    },
+    {
+      "epoch": 1.5565108861461656,
+      "grad_norm": 0.305867999792099,
+      "learning_rate": 0.00016036184210526313,
+      "loss": 0.5571,
+      "step": 1403
+    },
+    {
+      "epoch": 1.557620302315906,
+      "grad_norm": 0.2919771075248718,
+      "learning_rate": 0.00016023848684210525,
+      "loss": 0.6567,
+      "step": 1404
+    },
+    {
+      "epoch": 1.558729718485647,
+      "grad_norm": 0.24828073382377625,
+      "learning_rate": 0.00016011513157894734,
+      "loss": 0.409,
+      "step": 1405
+    },
+    {
+      "epoch": 1.5598391346553875,
+      "grad_norm": 0.48059597611427307,
+      "learning_rate": 0.00015999177631578944,
+      "loss": 0.5028,
+      "step": 1406
+    },
+    {
+      "epoch": 1.5609485508251284,
+      "grad_norm": 0.28389695286750793,
+      "learning_rate": 0.00015986842105263158,
+      "loss": 0.369,
+      "step": 1407
+    },
+    {
+      "epoch": 1.5620579669948689,
+      "grad_norm": 0.505401074886322,
+      "learning_rate": 0.00015974506578947368,
+      "loss": 0.6917,
+      "step": 1408
+    },
+    {
+      "epoch": 1.5631673831646096,
+      "grad_norm": 0.24662017822265625,
+      "learning_rate": 0.00015962171052631577,
+      "loss": 0.3777,
+      "step": 1409
+    },
+    {
+      "epoch": 1.5642767993343503,
+      "grad_norm": 0.38750240206718445,
+      "learning_rate": 0.0001594983552631579,
+      "loss": 0.6647,
+      "step": 1410
+    },
+    {
+      "epoch": 1.565386215504091,
+      "grad_norm": 0.41798150539398193,
+      "learning_rate": 0.00015937499999999998,
+      "loss": 0.5661,
+      "step": 1411
+    },
+    {
+      "epoch": 1.5664956316738317,
+      "grad_norm": 0.24084101617336273,
+      "learning_rate": 0.00015925164473684208,
+      "loss": 0.5066,
+      "step": 1412
+    },
+    {
+      "epoch": 1.5676050478435724,
+      "grad_norm": 0.3744387924671173,
+      "learning_rate": 0.0001591282894736842,
+      "loss": 0.4722,
+      "step": 1413
+    },
+    {
+      "epoch": 1.568714464013313,
+      "grad_norm": 0.2724044919013977,
+      "learning_rate": 0.0001590049342105263,
+      "loss": 0.5259,
+      "step": 1414
+    },
+    {
+      "epoch": 1.5698238801830535,
+      "grad_norm": 0.2745903432369232,
+      "learning_rate": 0.00015888157894736838,
+      "loss": 0.4011,
+      "step": 1415
+    },
+    {
+      "epoch": 1.5709332963527944,
+      "grad_norm": 0.20273329317569733,
+      "learning_rate": 0.00015875822368421053,
+      "loss": 0.5408,
+      "step": 1416
+    },
+    {
+      "epoch": 1.572042712522535,
+      "grad_norm": 0.38695916533470154,
+      "learning_rate": 0.00015863486842105262,
+      "loss": 0.6258,
+      "step": 1417
+    },
+    {
+      "epoch": 1.5731521286922758,
+      "grad_norm": 0.26426389813423157,
+      "learning_rate": 0.00015851151315789474,
+      "loss": 0.4185,
+      "step": 1418
+    },
+    {
+      "epoch": 1.5742615448620163,
+      "grad_norm": 0.2685663104057312,
+      "learning_rate": 0.00015838815789473684,
+      "loss": 1.0182,
+      "step": 1419
+    },
+    {
+      "epoch": 1.575370961031757,
+      "grad_norm": 0.3128146231174469,
+      "learning_rate": 0.00015826480263157893,
+      "loss": 0.4687,
+      "step": 1420
+    },
+    {
+      "epoch": 1.5764803772014977,
+      "grad_norm": 0.24564802646636963,
+      "learning_rate": 0.00015814144736842102,
+      "loss": 0.4918,
+      "step": 1421
+    },
+    {
+      "epoch": 1.5775897933712384,
+      "grad_norm": 0.3619256913661957,
+      "learning_rate": 0.00015801809210526314,
+      "loss": 0.4946,
+      "step": 1422
+    },
+    {
+      "epoch": 1.578699209540979,
+      "grad_norm": 0.4339245557785034,
+      "learning_rate": 0.00015789473684210524,
+      "loss": 0.5112,
+      "step": 1423
+    },
+    {
+      "epoch": 1.5798086257107198,
+      "grad_norm": 0.18522101640701294,
+      "learning_rate": 0.00015777138157894733,
+      "loss": 0.3891,
+      "step": 1424
+    },
+    {
+      "epoch": 1.5809180418804605,
+      "grad_norm": 0.23903268575668335,
+      "learning_rate": 0.00015764802631578945,
+      "loss": 0.3706,
+      "step": 1425
+    },
+    {
+      "epoch": 1.582027458050201,
+      "grad_norm": 0.6427960991859436,
+      "learning_rate": 0.00015752467105263157,
+      "loss": 0.6339,
+      "step": 1426
+    },
+    {
+      "epoch": 1.5831368742199419,
+      "grad_norm": 0.2820015847682953,
+      "learning_rate": 0.0001574013157894737,
+      "loss": 0.4439,
+      "step": 1427
+    },
+    {
+      "epoch": 1.5842462903896823,
+      "grad_norm": 0.26673081517219543,
+      "learning_rate": 0.00015727796052631579,
+      "loss": 0.5116,
+      "step": 1428
+    },
+    {
+      "epoch": 1.5853557065594233,
+      "grad_norm": 0.36776989698410034,
+      "learning_rate": 0.00015715460526315788,
+      "loss": 0.4931,
+      "step": 1429
+    },
+    {
+      "epoch": 1.5864651227291637,
+      "grad_norm": 0.3894679844379425,
+      "learning_rate": 0.00015703125,
+      "loss": 0.478,
+      "step": 1430
+    },
+    {
+      "epoch": 1.5875745388989044,
+      "grad_norm": 0.27240705490112305,
+      "learning_rate": 0.0001569078947368421,
+      "loss": 0.3737,
+      "step": 1431
+    },
+    {
+      "epoch": 1.5886839550686451,
+      "grad_norm": 0.32550275325775146,
+      "learning_rate": 0.00015678453947368418,
+      "loss": 0.5774,
+      "step": 1432
+    },
+    {
+      "epoch": 1.5897933712383858,
+      "grad_norm": 0.32884177565574646,
+      "learning_rate": 0.00015666118421052628,
+      "loss": 0.5408,
+      "step": 1433
+    },
+    {
+      "epoch": 1.5909027874081265,
+      "grad_norm": 0.45859280228614807,
+      "learning_rate": 0.0001565378289473684,
+      "loss": 0.7546,
+      "step": 1434
+    },
+    {
+      "epoch": 1.592012203577867,
+      "grad_norm": 0.29456886649131775,
+      "learning_rate": 0.00015641447368421052,
+      "loss": 0.3665,
+      "step": 1435
+    },
+    {
+      "epoch": 1.593121619747608,
+      "grad_norm": 0.22899344563484192,
+      "learning_rate": 0.00015629111842105264,
+      "loss": 0.4733,
+      "step": 1436
+    },
+    {
+      "epoch": 1.5942310359173484,
+      "grad_norm": 0.3687454164028168,
+      "learning_rate": 0.00015616776315789473,
+      "loss": 0.4508,
+      "step": 1437
+    },
+    {
+      "epoch": 1.5953404520870893,
+      "grad_norm": 0.2763974964618683,
+      "learning_rate": 0.00015604440789473683,
+      "loss": 0.5613,
+      "step": 1438
+    },
+    {
+      "epoch": 1.5964498682568298,
+      "grad_norm": 0.4642561376094818,
+      "learning_rate": 0.00015592105263157895,
+      "loss": 0.5261,
+      "step": 1439
+    },
+    {
+      "epoch": 1.5975592844265705,
+      "grad_norm": 0.46116307377815247,
+      "learning_rate": 0.00015579769736842104,
+      "loss": 0.5183,
+      "step": 1440
+    },
+    {
+      "epoch": 1.5986687005963112,
+      "grad_norm": 0.42349570989608765,
+      "learning_rate": 0.00015567434210526313,
+      "loss": 0.4358,
+      "step": 1441
+    },
+    {
+      "epoch": 1.5997781167660519,
+      "grad_norm": 0.5927262902259827,
+      "learning_rate": 0.00015555098684210525,
+      "loss": 0.4545,
+      "step": 1442
+    },
+    {
+      "epoch": 1.6008875329357926,
+      "grad_norm": 0.2763030230998993,
+      "learning_rate": 0.00015542763157894735,
+      "loss": 0.598,
+      "step": 1443
+    },
+    {
+      "epoch": 1.6019969491055333,
+      "grad_norm": 0.2903679311275482,
+      "learning_rate": 0.00015530427631578944,
+      "loss": 0.4782,
+      "step": 1444
+    },
+    {
+      "epoch": 1.603106365275274,
+      "grad_norm": 0.31723347306251526,
+      "learning_rate": 0.00015518092105263159,
+      "loss": 0.4793,
+      "step": 1445
+    },
+    {
+      "epoch": 1.6042157814450144,
+      "grad_norm": 0.3059498965740204,
+      "learning_rate": 0.00015505756578947368,
+      "loss": 0.5086,
+      "step": 1446
+    },
+    {
+      "epoch": 1.6053251976147553,
+      "grad_norm": 0.5435683131217957,
+      "learning_rate": 0.00015493421052631577,
+      "loss": 0.5502,
+      "step": 1447
+    },
+    {
+      "epoch": 1.6064346137844958,
+      "grad_norm": 0.28451740741729736,
+      "learning_rate": 0.0001548108552631579,
+      "loss": 0.3415,
+      "step": 1448
+    },
+    {
+      "epoch": 1.6075440299542367,
+      "grad_norm": 0.27565130591392517,
+      "learning_rate": 0.00015468749999999999,
+      "loss": 0.4961,
+      "step": 1449
+    },
+    {
+      "epoch": 1.6086534461239772,
+      "grad_norm": 0.38885173201560974,
+      "learning_rate": 0.00015456414473684208,
+      "loss": 0.6244,
+      "step": 1450
+    },
+    {
+      "epoch": 1.609762862293718,
+      "grad_norm": 0.3009326457977295,
+      "learning_rate": 0.0001544407894736842,
+      "loss": 0.3191,
+      "step": 1451
+    },
+    {
+      "epoch": 1.6108722784634586,
+      "grad_norm": 0.2374914437532425,
+      "learning_rate": 0.0001543174342105263,
+      "loss": 0.4029,
+      "step": 1452
+    },
+    {
+      "epoch": 1.6119816946331993,
+      "grad_norm": 0.24939504265785217,
+      "learning_rate": 0.00015419407894736839,
+      "loss": 0.6444,
+      "step": 1453
+    },
+    {
+      "epoch": 1.61309111080294,
+      "grad_norm": 0.2710380554199219,
+      "learning_rate": 0.00015407072368421053,
+      "loss": 0.5412,
+      "step": 1454
+    },
+    {
+      "epoch": 1.6142005269726805,
+      "grad_norm": 0.32970866560935974,
+      "learning_rate": 0.00015394736842105263,
+      "loss": 0.4512,
+      "step": 1455
+    },
+    {
+      "epoch": 1.6153099431424214,
+      "grad_norm": 0.4523448944091797,
+      "learning_rate": 0.00015382401315789472,
+      "loss": 0.5883,
+      "step": 1456
+    },
+    {
+      "epoch": 1.6164193593121619,
+      "grad_norm": 0.25891023874282837,
+      "learning_rate": 0.00015370065789473684,
+      "loss": 0.5516,
+      "step": 1457
+    },
+    {
+      "epoch": 1.6175287754819028,
+      "grad_norm": 0.37495937943458557,
+      "learning_rate": 0.00015357730263157893,
+      "loss": 0.5841,
+      "step": 1458
+    },
+    {
+      "epoch": 1.6186381916516432,
+      "grad_norm": 0.33875572681427,
+      "learning_rate": 0.00015345394736842103,
+      "loss": 0.6034,
+      "step": 1459
+    },
+    {
+      "epoch": 1.6197476078213842,
+      "grad_norm": 0.2283666431903839,
+      "learning_rate": 0.00015333059210526315,
+      "loss": 0.3142,
+      "step": 1460
+    },
+    {
+      "epoch": 1.6208570239911246,
+      "grad_norm": 0.35166046023368835,
+      "learning_rate": 0.00015320723684210524,
+      "loss": 0.5688,
+      "step": 1461
+    },
+    {
+      "epoch": 1.6219664401608653,
+      "grad_norm": 0.3472052812576294,
+      "learning_rate": 0.00015308388157894733,
+      "loss": 0.2508,
+      "step": 1462
+    },
+    {
+      "epoch": 1.623075856330606,
+      "grad_norm": 0.30597633123397827,
+      "learning_rate": 0.00015296052631578945,
+      "loss": 0.5829,
+      "step": 1463
+    },
+    {
+      "epoch": 1.6241852725003467,
+      "grad_norm": 0.7042750716209412,
+      "learning_rate": 0.00015283717105263157,
+      "loss": 0.5834,
+      "step": 1464
+    },
+    {
+      "epoch": 1.6252946886700874,
+      "grad_norm": 0.3710060119628906,
+      "learning_rate": 0.0001527138157894737,
+      "loss": 0.4427,
+      "step": 1465
+    },
+    {
+      "epoch": 1.626404104839828,
+      "grad_norm": 0.3646557927131653,
+      "learning_rate": 0.00015259046052631579,
+      "loss": 0.5418,
+      "step": 1466
+    },
+    {
+      "epoch": 1.6275135210095688,
+      "grad_norm": 0.357334703207016,
+      "learning_rate": 0.00015246710526315788,
+      "loss": 0.5551,
+      "step": 1467
+    },
+    {
+      "epoch": 1.6286229371793093,
+      "grad_norm": 0.2936727702617645,
+      "learning_rate": 0.00015234375,
+      "loss": 0.5441,
+      "step": 1468
+    },
+    {
+      "epoch": 1.6297323533490502,
+      "grad_norm": 0.24167072772979736,
+      "learning_rate": 0.0001522203947368421,
+      "loss": 0.6756,
+      "step": 1469
+    },
+    {
+      "epoch": 1.6308417695187907,
+      "grad_norm": 0.3860384523868561,
+      "learning_rate": 0.00015209703947368419,
+      "loss": 0.6466,
+      "step": 1470
+    },
+    {
+      "epoch": 1.6319511856885314,
+      "grad_norm": 0.20689241588115692,
+      "learning_rate": 0.00015197368421052628,
+      "loss": 0.28,
+      "step": 1471
+    },
+    {
+      "epoch": 1.633060601858272,
+      "grad_norm": 0.2413812130689621,
+      "learning_rate": 0.0001518503289473684,
+      "loss": 0.5198,
+      "step": 1472
+    },
+    {
+      "epoch": 1.6341700180280128,
+      "grad_norm": 0.2979881167411804,
+      "learning_rate": 0.00015172697368421052,
+      "loss": 0.5016,
+      "step": 1473
+    },
+    {
+      "epoch": 1.6352794341977535,
+      "grad_norm": 0.22364138066768646,
+      "learning_rate": 0.00015160361842105264,
+      "loss": 0.3996,
+      "step": 1474
+    },
+    {
+      "epoch": 1.6363888503674942,
+      "grad_norm": 0.27691081166267395,
+      "learning_rate": 0.00015148026315789473,
+      "loss": 0.4038,
+      "step": 1475
+    },
+    {
+      "epoch": 1.6374982665372348,
+      "grad_norm": 0.300599604845047,
+      "learning_rate": 0.00015135690789473683,
+      "loss": 0.4522,
+      "step": 1476
+    },
+    {
+      "epoch": 1.6386076827069753,
+      "grad_norm": 0.2906956672668457,
+      "learning_rate": 0.00015123355263157895,
+      "loss": 0.5978,
+      "step": 1477
+    },
+    {
+      "epoch": 1.6397170988767162,
+      "grad_norm": 0.3278878629207611,
+      "learning_rate": 0.00015111019736842104,
+      "loss": 0.445,
+      "step": 1478
+    },
+    {
+      "epoch": 1.6408265150464567,
+      "grad_norm": 0.25733959674835205,
+      "learning_rate": 0.00015098684210526313,
+      "loss": 0.4896,
+      "step": 1479
+    },
+    {
+      "epoch": 1.6419359312161976,
+      "grad_norm": 0.33139604330062866,
+      "learning_rate": 0.00015086348684210525,
+      "loss": 0.54,
+      "step": 1480
+    },
+    {
+      "epoch": 1.643045347385938,
+      "grad_norm": 0.6023613810539246,
+      "learning_rate": 0.00015074013157894735,
+      "loss": 0.6114,
+      "step": 1481
+    },
+    {
+      "epoch": 1.6441547635556788,
+      "grad_norm": 0.22838066518306732,
+      "learning_rate": 0.00015061677631578944,
+      "loss": 0.426,
+      "step": 1482
+    },
+    {
+      "epoch": 1.6452641797254195,
+      "grad_norm": 0.26457536220550537,
+      "learning_rate": 0.0001504934210526316,
+      "loss": 0.4963,
+      "step": 1483
+    },
+    {
+      "epoch": 1.6463735958951602,
+      "grad_norm": 0.30254966020584106,
+      "learning_rate": 0.00015037006578947368,
+      "loss": 0.5112,
+      "step": 1484
+    },
+    {
+      "epoch": 1.6474830120649009,
+      "grad_norm": 0.23207102715969086,
+      "learning_rate": 0.00015024671052631577,
+      "loss": 0.4926,
+      "step": 1485
+    },
+    {
+      "epoch": 1.6485924282346414,
+      "grad_norm": 0.25486233830451965,
+      "learning_rate": 0.0001501233552631579,
+      "loss": 0.4387,
+      "step": 1486
+    },
+    {
+      "epoch": 1.6497018444043823,
+      "grad_norm": 0.3489553928375244,
+      "learning_rate": 0.00015,
+      "loss": 0.6869,
+      "step": 1487
+    },
+    {
+      "epoch": 1.6508112605741228,
+      "grad_norm": 0.28267863392829895,
+      "learning_rate": 0.00014987664473684208,
+      "loss": 0.4191,
+      "step": 1488
+    },
+    {
+      "epoch": 1.6519206767438637,
+      "grad_norm": 0.2585887014865875,
+      "learning_rate": 0.0001497532894736842,
+      "loss": 0.5599,
+      "step": 1489
+    },
+    {
+      "epoch": 1.6530300929136041,
+      "grad_norm": 0.27951672673225403,
+      "learning_rate": 0.00014962993421052632,
+      "loss": 0.6091,
+      "step": 1490
+    },
+    {
+      "epoch": 1.6541395090833448,
+      "grad_norm": 0.33680734038352966,
+      "learning_rate": 0.00014950657894736841,
+      "loss": 0.5986,
+      "step": 1491
+    },
+    {
+      "epoch": 1.6552489252530855,
+      "grad_norm": 0.33777835965156555,
+      "learning_rate": 0.0001493832236842105,
+      "loss": 0.6051,
+      "step": 1492
+    },
+    {
+      "epoch": 1.6563583414228262,
+      "grad_norm": 0.29178673028945923,
+      "learning_rate": 0.00014925986842105263,
+      "loss": 0.3462,
+      "step": 1493
+    },
+    {
+      "epoch": 1.657467757592567,
+      "grad_norm": 0.23857684433460236,
+      "learning_rate": 0.00014913651315789472,
+      "loss": 0.3521,
+      "step": 1494
+    },
+    {
+      "epoch": 1.6585771737623076,
+      "grad_norm": 0.33080053329467773,
+      "learning_rate": 0.00014901315789473684,
+      "loss": 0.4185,
+      "step": 1495
+    },
+    {
+      "epoch": 1.6596865899320483,
+      "grad_norm": 0.35032498836517334,
+      "learning_rate": 0.00014888980263157893,
+      "loss": 0.6305,
+      "step": 1496
+    },
+    {
+      "epoch": 1.6607960061017888,
+      "grad_norm": 0.2609749138355255,
+      "learning_rate": 0.00014876644736842103,
+      "loss": 0.4313,
+      "step": 1497
+    },
+    {
+      "epoch": 1.6619054222715297,
+      "grad_norm": 0.3178192675113678,
+      "learning_rate": 0.00014864309210526315,
+      "loss": 0.6381,
+      "step": 1498
+    },
+    {
+      "epoch": 1.6630148384412702,
+      "grad_norm": 0.3444008231163025,
+      "learning_rate": 0.00014851973684210527,
+      "loss": 0.6142,
+      "step": 1499
+    },
+    {
+      "epoch": 1.664124254611011,
+      "grad_norm": 0.24173513054847717,
+      "learning_rate": 0.00014839638157894736,
+      "loss": 0.5568,
+      "step": 1500
+    },
+    {
+      "epoch": 1.6652336707807516,
+      "grad_norm": 0.23622675240039825,
+      "learning_rate": 0.00014827302631578945,
+      "loss": 0.4888,
+      "step": 1501
+    },
+    {
+      "epoch": 1.6663430869504923,
+      "grad_norm": 0.24951039254665375,
+      "learning_rate": 0.00014814967105263157,
+      "loss": 0.5554,
+      "step": 1502
+    },
+    {
+      "epoch": 1.667452503120233,
+      "grad_norm": 0.20648309588432312,
+      "learning_rate": 0.00014802631578947367,
+      "loss": 0.3302,
+      "step": 1503
+    },
+    {
+      "epoch": 1.6685619192899737,
+      "grad_norm": 0.2786915898323059,
+      "learning_rate": 0.0001479029605263158,
+      "loss": 0.3773,
+      "step": 1504
+    },
+    {
+      "epoch": 1.6696713354597144,
+      "grad_norm": 0.35018453001976013,
+      "learning_rate": 0.00014777960526315788,
+      "loss": 0.5903,
+      "step": 1505
+    },
+    {
+      "epoch": 1.670780751629455,
+      "grad_norm": 0.34121614694595337,
+      "learning_rate": 0.00014765624999999997,
+      "loss": 0.4813,
+      "step": 1506
+    },
+    {
+      "epoch": 1.6718901677991957,
+      "grad_norm": 0.41687941551208496,
+      "learning_rate": 0.0001475328947368421,
+      "loss": 0.7389,
+      "step": 1507
+    },
+    {
+      "epoch": 1.6729995839689362,
+      "grad_norm": 0.27919813990592957,
+      "learning_rate": 0.0001474095394736842,
+      "loss": 0.3984,
+      "step": 1508
+    },
+    {
+      "epoch": 1.6741090001386771,
+      "grad_norm": 0.2847552001476288,
+      "learning_rate": 0.0001472861842105263,
+      "loss": 0.3091,
+      "step": 1509
+    },
+    {
+      "epoch": 1.6752184163084176,
+      "grad_norm": 0.3107469379901886,
+      "learning_rate": 0.0001471628289473684,
+      "loss": 0.4399,
+      "step": 1510
+    },
+    {
+      "epoch": 1.6763278324781585,
+      "grad_norm": 0.21151646971702576,
+      "learning_rate": 0.00014703947368421052,
+      "loss": 0.3691,
+      "step": 1511
+    },
+    {
+      "epoch": 1.677437248647899,
+      "grad_norm": 0.27195483446121216,
+      "learning_rate": 0.00014691611842105261,
+      "loss": 0.5531,
+      "step": 1512
+    },
+    {
+      "epoch": 1.6785466648176397,
+      "grad_norm": 0.2964016795158386,
+      "learning_rate": 0.0001467927631578947,
+      "loss": 0.4978,
+      "step": 1513
+    },
+    {
+      "epoch": 1.6796560809873804,
+      "grad_norm": 0.39583608508110046,
+      "learning_rate": 0.00014666940789473683,
+      "loss": 0.6652,
+      "step": 1514
+    },
+    {
+      "epoch": 1.680765497157121,
+      "grad_norm": 0.23227983713150024,
+      "learning_rate": 0.00014654605263157895,
+      "loss": 0.5401,
+      "step": 1515
+    },
+    {
+      "epoch": 1.6818749133268618,
+      "grad_norm": 0.2583910822868347,
+      "learning_rate": 0.00014642269736842104,
+      "loss": 0.5246,
+      "step": 1516
+    },
+    {
+      "epoch": 1.6829843294966023,
+      "grad_norm": 0.32901066541671753,
+      "learning_rate": 0.00014629934210526313,
+      "loss": 0.4348,
+      "step": 1517
+    },
+    {
+      "epoch": 1.6840937456663432,
+      "grad_norm": 0.2796122133731842,
+      "learning_rate": 0.00014617598684210525,
+      "loss": 0.2749,
+      "step": 1518
+    },
+    {
+      "epoch": 1.6852031618360837,
+      "grad_norm": 0.2737559378147125,
+      "learning_rate": 0.00014605263157894735,
+      "loss": 0.3413,
+      "step": 1519
+    },
+    {
+      "epoch": 1.6863125780058246,
+      "grad_norm": 0.5847452282905579,
+      "learning_rate": 0.00014592927631578947,
+      "loss": 0.5046,
+      "step": 1520
+    },
+    {
+      "epoch": 1.687421994175565,
+      "grad_norm": 0.42608511447906494,
+      "learning_rate": 0.00014580592105263156,
+      "loss": 0.511,
+      "step": 1521
+    },
+    {
+      "epoch": 1.6885314103453057,
+      "grad_norm": 0.3839153051376343,
+      "learning_rate": 0.00014568256578947365,
+      "loss": 0.4812,
+      "step": 1522
+    },
+    {
+      "epoch": 1.6896408265150464,
+      "grad_norm": 0.2144833356142044,
+      "learning_rate": 0.00014555921052631577,
+      "loss": 0.479,
+      "step": 1523
+    },
+    {
+      "epoch": 1.6907502426847871,
+      "grad_norm": 0.34472957253456116,
+      "learning_rate": 0.0001454358552631579,
+      "loss": 0.356,
+      "step": 1524
+    },
+    {
+      "epoch": 1.6918596588545278,
+      "grad_norm": 0.29683181643486023,
+      "learning_rate": 0.0001453125,
+      "loss": 0.385,
+      "step": 1525
+    },
+    {
+      "epoch": 1.6929690750242685,
+      "grad_norm": 0.3433467447757721,
+      "learning_rate": 0.00014518914473684208,
+      "loss": 0.4606,
+      "step": 1526
+    },
+    {
+      "epoch": 1.6940784911940092,
+      "grad_norm": 0.3034195601940155,
+      "learning_rate": 0.0001450657894736842,
+      "loss": 0.5594,
+      "step": 1527
+    },
+    {
+      "epoch": 1.6951879073637497,
+      "grad_norm": 0.49384185671806335,
+      "learning_rate": 0.00014494243421052632,
+      "loss": 0.6995,
+      "step": 1528
+    },
+    {
+      "epoch": 1.6962973235334906,
+      "grad_norm": 0.5246427655220032,
+      "learning_rate": 0.00014481907894736842,
+      "loss": 0.6432,
+      "step": 1529
+    },
+    {
+      "epoch": 1.697406739703231,
+      "grad_norm": 0.26392480731010437,
+      "learning_rate": 0.0001446957236842105,
+      "loss": 0.5626,
+      "step": 1530
+    },
+    {
+      "epoch": 1.698516155872972,
+      "grad_norm": 0.27015626430511475,
+      "learning_rate": 0.0001445723684210526,
+      "loss": 0.4244,
+      "step": 1531
+    },
+    {
+      "epoch": 1.6996255720427125,
+      "grad_norm": 0.3427369296550751,
+      "learning_rate": 0.00014444901315789472,
+      "loss": 0.4015,
+      "step": 1532
+    },
+    {
+      "epoch": 1.7007349882124532,
+      "grad_norm": 0.4389760196208954,
+      "learning_rate": 0.00014432565789473684,
+      "loss": 0.654,
+      "step": 1533
+    },
+    {
+      "epoch": 1.7018444043821939,
+      "grad_norm": 0.36939921975135803,
+      "learning_rate": 0.00014420230263157894,
+      "loss": 0.6009,
+      "step": 1534
+    },
+    {
+      "epoch": 1.7029538205519346,
+      "grad_norm": 0.2916509211063385,
+      "learning_rate": 0.00014407894736842103,
+      "loss": 0.4624,
+      "step": 1535
+    },
+    {
+      "epoch": 1.7040632367216753,
+      "grad_norm": 0.5189476013183594,
+      "learning_rate": 0.00014395559210526315,
+      "loss": 0.5838,
+      "step": 1536
+    },
+    {
+      "epoch": 1.7051726528914157,
+      "grad_norm": 0.2686052620410919,
+      "learning_rate": 0.00014383223684210527,
+      "loss": 0.9601,
+      "step": 1537
+    },
+    {
+      "epoch": 1.7062820690611566,
+      "grad_norm": 0.28845494985580444,
+      "learning_rate": 0.00014370888157894736,
+      "loss": 0.3672,
+      "step": 1538
+    },
+    {
+      "epoch": 1.7073914852308971,
+      "grad_norm": 0.21178792417049408,
+      "learning_rate": 0.00014358552631578946,
+      "loss": 0.3064,
+      "step": 1539
+    },
+    {
+      "epoch": 1.708500901400638,
+      "grad_norm": 0.2538648843765259,
+      "learning_rate": 0.00014346217105263158,
+      "loss": 0.4975,
+      "step": 1540
+    },
+    {
+      "epoch": 1.7096103175703785,
+      "grad_norm": 0.4006761908531189,
+      "learning_rate": 0.00014333881578947367,
+      "loss": 0.5619,
+      "step": 1541
+    },
+    {
+      "epoch": 1.7107197337401192,
+      "grad_norm": 0.3077350854873657,
+      "learning_rate": 0.0001432154605263158,
+      "loss": 0.6509,
+      "step": 1542
+    },
+    {
+      "epoch": 1.71182914990986,
+      "grad_norm": 0.29142218828201294,
+      "learning_rate": 0.00014309210526315788,
+      "loss": 0.4765,
+      "step": 1543
+    },
+    {
+      "epoch": 1.7129385660796006,
+      "grad_norm": 0.3905639350414276,
+      "learning_rate": 0.00014296874999999998,
+      "loss": 0.8906,
+      "step": 1544
+    },
+    {
+      "epoch": 1.7140479822493413,
+      "grad_norm": 0.32861366868019104,
+      "learning_rate": 0.0001428453947368421,
+      "loss": 0.6136,
+      "step": 1545
+    },
+    {
+      "epoch": 1.715157398419082,
+      "grad_norm": 0.20155826210975647,
+      "learning_rate": 0.0001427220394736842,
+      "loss": 0.4392,
+      "step": 1546
+    },
+    {
+      "epoch": 1.7162668145888227,
+      "grad_norm": 0.35804450511932373,
+      "learning_rate": 0.0001425986842105263,
+      "loss": 0.6582,
+      "step": 1547
+    },
+    {
+      "epoch": 1.7173762307585632,
+      "grad_norm": 0.26054689288139343,
+      "learning_rate": 0.0001424753289473684,
+      "loss": 0.4493,
+      "step": 1548
+    },
+    {
+      "epoch": 1.718485646928304,
+      "grad_norm": 0.25189530849456787,
+      "learning_rate": 0.00014235197368421052,
+      "loss": 0.4806,
+      "step": 1549
+    },
+    {
+      "epoch": 1.7195950630980446,
+      "grad_norm": 0.3394787609577179,
+      "learning_rate": 0.00014222861842105262,
+      "loss": 0.5681,
+      "step": 1550
+    },
+    {
+      "epoch": 1.7207044792677855,
+      "grad_norm": 0.3084029257297516,
+      "learning_rate": 0.0001421052631578947,
+      "loss": 0.6309,
+      "step": 1551
+    },
+    {
+      "epoch": 1.721813895437526,
+      "grad_norm": 0.3268156945705414,
+      "learning_rate": 0.00014198190789473683,
+      "loss": 0.4937,
+      "step": 1552
+    },
+    {
+      "epoch": 1.7229233116072666,
+      "grad_norm": 0.2543306350708008,
+      "learning_rate": 0.00014185855263157895,
+      "loss": 0.3895,
+      "step": 1553
+    },
+    {
+      "epoch": 1.7240327277770073,
+      "grad_norm": 0.2428501844406128,
+      "learning_rate": 0.00014173519736842104,
+      "loss": 0.4724,
+      "step": 1554
+    },
+    {
+      "epoch": 1.725142143946748,
+      "grad_norm": 0.34834590554237366,
+      "learning_rate": 0.00014161184210526314,
+      "loss": 0.4686,
+      "step": 1555
+    },
+    {
+      "epoch": 1.7262515601164887,
+      "grad_norm": 0.2403583824634552,
+      "learning_rate": 0.00014148848684210526,
+      "loss": 0.3847,
+      "step": 1556
+    },
+    {
+      "epoch": 1.7273609762862294,
+      "grad_norm": 0.38176214694976807,
+      "learning_rate": 0.00014136513157894735,
+      "loss": 0.4869,
+      "step": 1557
+    },
+    {
+      "epoch": 1.7284703924559701,
+      "grad_norm": 0.2659490406513214,
+      "learning_rate": 0.00014124177631578947,
+      "loss": 0.4449,
+      "step": 1558
+    },
+    {
+      "epoch": 1.7295798086257106,
+      "grad_norm": 0.2423594892024994,
+      "learning_rate": 0.00014111842105263156,
+      "loss": 0.5206,
+      "step": 1559
+    },
+    {
+      "epoch": 1.7306892247954515,
+      "grad_norm": 0.3293440639972687,
+      "learning_rate": 0.00014099506578947366,
+      "loss": 0.621,
+      "step": 1560
+    },
+    {
+      "epoch": 1.731798640965192,
+      "grad_norm": 0.27292686700820923,
+      "learning_rate": 0.00014087171052631578,
+      "loss": 0.4955,
+      "step": 1561
+    },
+    {
+      "epoch": 1.732908057134933,
+      "grad_norm": 0.3719004690647125,
+      "learning_rate": 0.0001407483552631579,
+      "loss": 0.4081,
+      "step": 1562
+    },
+    {
+      "epoch": 1.7340174733046734,
+      "grad_norm": 0.3784489631652832,
+      "learning_rate": 0.000140625,
+      "loss": 0.4212,
+      "step": 1563
+    },
+    {
+      "epoch": 1.735126889474414,
+      "grad_norm": 0.24494099617004395,
+      "learning_rate": 0.00014050164473684208,
+      "loss": 0.4698,
+      "step": 1564
+    },
+    {
+      "epoch": 1.7362363056441548,
+      "grad_norm": 0.2339191883802414,
+      "learning_rate": 0.0001403782894736842,
+      "loss": 0.3486,
+      "step": 1565
+    },
+    {
+      "epoch": 1.7373457218138955,
+      "grad_norm": 0.321445107460022,
+      "learning_rate": 0.00014025493421052632,
+      "loss": 0.8323,
+      "step": 1566
+    },
+    {
+      "epoch": 1.7384551379836362,
+      "grad_norm": 0.3625154495239258,
+      "learning_rate": 0.00014013157894736842,
+      "loss": 0.3676,
+      "step": 1567
+    },
+    {
+      "epoch": 1.7395645541533766,
+      "grad_norm": 0.30214935541152954,
+      "learning_rate": 0.0001400082236842105,
+      "loss": 0.611,
+      "step": 1568
+    },
+    {
+      "epoch": 1.7406739703231175,
+      "grad_norm": 0.3197210133075714,
+      "learning_rate": 0.0001398848684210526,
+      "loss": 0.5491,
+      "step": 1569
+    },
+    {
+      "epoch": 1.741783386492858,
+      "grad_norm": 0.31939029693603516,
+      "learning_rate": 0.00013976151315789472,
+      "loss": 0.5426,
+      "step": 1570
+    },
+    {
+      "epoch": 1.742892802662599,
+      "grad_norm": 0.470907986164093,
+      "learning_rate": 0.00013963815789473684,
+      "loss": 0.6622,
+      "step": 1571
+    },
+    {
+      "epoch": 1.7440022188323394,
+      "grad_norm": 0.2162821739912033,
+      "learning_rate": 0.00013951480263157894,
+      "loss": 0.5728,
+      "step": 1572
+    },
+    {
+      "epoch": 1.74511163500208,
+      "grad_norm": 0.24964164197444916,
+      "learning_rate": 0.00013939144736842103,
+      "loss": 0.6929,
+      "step": 1573
+    },
+    {
+      "epoch": 1.7462210511718208,
+      "grad_norm": 0.34951767325401306,
+      "learning_rate": 0.00013926809210526315,
+      "loss": 0.3177,
+      "step": 1574
+    },
+    {
+      "epoch": 1.7473304673415615,
+      "grad_norm": 0.2987998425960541,
+      "learning_rate": 0.00013914473684210527,
+      "loss": 0.4462,
+      "step": 1575
+    },
+    {
+      "epoch": 1.7484398835113022,
+      "grad_norm": 0.362047016620636,
+      "learning_rate": 0.00013902138157894736,
+      "loss": 0.4171,
+      "step": 1576
+    },
+    {
+      "epoch": 1.749549299681043,
+      "grad_norm": 0.2592370808124542,
+      "learning_rate": 0.00013889802631578946,
+      "loss": 0.2222,
+      "step": 1577
+    },
+    {
+      "epoch": 1.7506587158507836,
+      "grad_norm": 0.26247555017471313,
+      "learning_rate": 0.00013877467105263158,
+      "loss": 0.5498,
+      "step": 1578
+    },
+    {
+      "epoch": 1.751768132020524,
+      "grad_norm": 0.28997063636779785,
+      "learning_rate": 0.00013865131578947367,
+      "loss": 0.537,
+      "step": 1579
+    },
+    {
+      "epoch": 1.752877548190265,
+      "grad_norm": 0.31275662779808044,
+      "learning_rate": 0.0001385279605263158,
+      "loss": 0.5331,
+      "step": 1580
+    },
+    {
+      "epoch": 1.7539869643600055,
+      "grad_norm": 0.3327484130859375,
+      "learning_rate": 0.00013840460526315788,
+      "loss": 0.4344,
+      "step": 1581
+    },
+    {
+      "epoch": 1.7550963805297464,
+      "grad_norm": 0.31380587816238403,
+      "learning_rate": 0.00013828124999999998,
+      "loss": 0.641,
+      "step": 1582
+    },
+    {
+      "epoch": 1.7562057966994868,
+      "grad_norm": 0.2786813974380493,
+      "learning_rate": 0.0001381578947368421,
+      "loss": 0.3636,
+      "step": 1583
+    },
+    {
+      "epoch": 1.7573152128692275,
+      "grad_norm": 0.33598342537879944,
+      "learning_rate": 0.0001380345394736842,
+      "loss": 0.5949,
+      "step": 1584
+    },
+    {
+      "epoch": 1.7584246290389682,
+      "grad_norm": 0.28291746973991394,
+      "learning_rate": 0.0001379111842105263,
+      "loss": 0.3358,
+      "step": 1585
+    },
+    {
+      "epoch": 1.759534045208709,
+      "grad_norm": 0.22912530601024628,
+      "learning_rate": 0.0001377878289473684,
+      "loss": 0.4407,
+      "step": 1586
+    },
+    {
+      "epoch": 1.7606434613784496,
+      "grad_norm": 0.3488161861896515,
+      "learning_rate": 0.00013766447368421052,
+      "loss": 0.4591,
+      "step": 1587
+    },
+    {
+      "epoch": 1.76175287754819,
+      "grad_norm": 0.31319087743759155,
+      "learning_rate": 0.00013754111842105262,
+      "loss": 0.496,
+      "step": 1588
+    },
+    {
+      "epoch": 1.762862293717931,
+      "grad_norm": 0.31536537408828735,
+      "learning_rate": 0.0001374177631578947,
+      "loss": 0.4566,
+      "step": 1589
+    },
+    {
+      "epoch": 1.7639717098876715,
+      "grad_norm": 0.38071408867836,
+      "learning_rate": 0.00013729440789473683,
+      "loss": 0.4728,
+      "step": 1590
+    },
+    {
+      "epoch": 1.7650811260574124,
+      "grad_norm": 0.25146248936653137,
+      "learning_rate": 0.00013717105263157895,
+      "loss": 0.3651,
+      "step": 1591
+    },
+    {
+      "epoch": 1.7661905422271529,
+      "grad_norm": 0.262510746717453,
+      "learning_rate": 0.00013704769736842104,
+      "loss": 0.3982,
+      "step": 1592
+    },
+    {
+      "epoch": 1.7672999583968938,
+      "grad_norm": 0.2857152819633484,
+      "learning_rate": 0.00013692434210526314,
+      "loss": 0.3916,
+      "step": 1593
+    },
+    {
+      "epoch": 1.7684093745666343,
+      "grad_norm": 0.5477368831634521,
+      "learning_rate": 0.00013680098684210526,
+      "loss": 0.5775,
+      "step": 1594
+    },
+    {
+      "epoch": 1.769518790736375,
+      "grad_norm": 0.30496469140052795,
+      "learning_rate": 0.00013667763157894735,
+      "loss": 0.6863,
+      "step": 1595
+    },
+    {
+      "epoch": 1.7706282069061157,
+      "grad_norm": 0.32225456833839417,
+      "learning_rate": 0.00013655427631578947,
+      "loss": 0.4028,
+      "step": 1596
+    },
+    {
+      "epoch": 1.7717376230758564,
+      "grad_norm": 0.25836458802223206,
+      "learning_rate": 0.00013643092105263156,
+      "loss": 0.4638,
+      "step": 1597
+    },
+    {
+      "epoch": 1.772847039245597,
+      "grad_norm": 0.24989454448223114,
+      "learning_rate": 0.00013630756578947366,
+      "loss": 0.5425,
+      "step": 1598
+    },
+    {
+      "epoch": 1.7739564554153375,
+      "grad_norm": 0.46502557396888733,
+      "learning_rate": 0.00013618421052631578,
+      "loss": 0.5317,
+      "step": 1599
+    },
+    {
+      "epoch": 1.7750658715850784,
+      "grad_norm": 0.32870060205459595,
+      "learning_rate": 0.0001360608552631579,
+      "loss": 0.5497,
+      "step": 1600
+    },
+    {
+      "epoch": 1.776175287754819,
+      "grad_norm": 0.24722667038440704,
+      "learning_rate": 0.0001359375,
+      "loss": 0.5929,
+      "step": 1601
+    },
+    {
+      "epoch": 1.7772847039245598,
+      "grad_norm": 0.3317899703979492,
+      "learning_rate": 0.00013581414473684208,
+      "loss": 0.4435,
+      "step": 1602
+    },
+    {
+      "epoch": 1.7783941200943003,
+      "grad_norm": 0.36830225586891174,
+      "learning_rate": 0.0001356907894736842,
+      "loss": 0.5876,
+      "step": 1603
+    },
+    {
+      "epoch": 1.779503536264041,
+      "grad_norm": 0.23982636630535126,
+      "learning_rate": 0.0001355674342105263,
+      "loss": 0.4906,
+      "step": 1604
+    },
+    {
+      "epoch": 1.7806129524337817,
+      "grad_norm": 0.3808034062385559,
+      "learning_rate": 0.00013544407894736842,
+      "loss": 0.6202,
+      "step": 1605
+    },
+    {
+      "epoch": 1.7817223686035224,
+      "grad_norm": 0.3208853006362915,
+      "learning_rate": 0.0001353207236842105,
+      "loss": 0.4002,
+      "step": 1606
+    },
+    {
+      "epoch": 1.782831784773263,
+      "grad_norm": 0.42497456073760986,
+      "learning_rate": 0.0001351973684210526,
+      "loss": 0.6453,
+      "step": 1607
+    },
+    {
+      "epoch": 1.7839412009430038,
+      "grad_norm": 0.5558460354804993,
+      "learning_rate": 0.00013507401315789472,
+      "loss": 0.424,
+      "step": 1608
+    },
+    {
+      "epoch": 1.7850506171127445,
+      "grad_norm": 0.38764357566833496,
+      "learning_rate": 0.00013495065789473684,
+      "loss": 0.4115,
+      "step": 1609
+    },
+    {
+      "epoch": 1.786160033282485,
+      "grad_norm": 0.3829018771648407,
+      "learning_rate": 0.00013482730263157894,
+      "loss": 0.4814,
+      "step": 1610
+    },
+    {
+      "epoch": 1.7872694494522259,
+      "grad_norm": 0.4661031663417816,
+      "learning_rate": 0.00013470394736842103,
+      "loss": 0.6416,
+      "step": 1611
+    },
+    {
+      "epoch": 1.7883788656219664,
+      "grad_norm": 0.2798513174057007,
+      "learning_rate": 0.00013458059210526315,
+      "loss": 0.4603,
+      "step": 1612
+    },
+    {
+      "epoch": 1.7894882817917073,
+      "grad_norm": 0.3700726330280304,
+      "learning_rate": 0.00013445723684210527,
+      "loss": 0.4821,
+      "step": 1613
+    },
+    {
+      "epoch": 1.7905976979614477,
+      "grad_norm": 0.35398468375205994,
+      "learning_rate": 0.00013433388157894736,
+      "loss": 0.5852,
+      "step": 1614
+    },
+    {
+      "epoch": 1.7917071141311884,
+      "grad_norm": 0.37176424264907837,
+      "learning_rate": 0.00013421052631578946,
+      "loss": 0.605,
+      "step": 1615
+    },
+    {
+      "epoch": 1.7928165303009291,
+      "grad_norm": 0.2966163754463196,
+      "learning_rate": 0.00013408717105263158,
+      "loss": 0.4163,
+      "step": 1616
+    },
+    {
+      "epoch": 1.7939259464706698,
+      "grad_norm": 0.3742397129535675,
+      "learning_rate": 0.00013396381578947367,
+      "loss": 0.4507,
+      "step": 1617
+    },
+    {
+      "epoch": 1.7950353626404105,
+      "grad_norm": 0.36498209834098816,
+      "learning_rate": 0.0001338404605263158,
+      "loss": 0.8315,
+      "step": 1618
+    },
+    {
+      "epoch": 1.796144778810151,
+      "grad_norm": 0.4069786071777344,
+      "learning_rate": 0.00013371710526315788,
+      "loss": 0.5758,
+      "step": 1619
+    },
+    {
+      "epoch": 1.797254194979892,
+      "grad_norm": 0.30974453687667847,
+      "learning_rate": 0.00013359374999999998,
+      "loss": 0.7786,
+      "step": 1620
+    },
+    {
+      "epoch": 1.7983636111496324,
+      "grad_norm": 0.5354030728340149,
+      "learning_rate": 0.0001334703947368421,
+      "loss": 0.5513,
+      "step": 1621
+    },
+    {
+      "epoch": 1.7994730273193733,
+      "grad_norm": 0.24419055879116058,
+      "learning_rate": 0.0001333470394736842,
+      "loss": 0.4483,
+      "step": 1622
+    },
+    {
+      "epoch": 1.8005824434891138,
+      "grad_norm": 0.33314335346221924,
+      "learning_rate": 0.0001332236842105263,
+      "loss": 0.5992,
+      "step": 1623
+    },
+    {
+      "epoch": 1.8016918596588545,
+      "grad_norm": 0.22179794311523438,
+      "learning_rate": 0.0001331003289473684,
+      "loss": 0.382,
+      "step": 1624
+    },
+    {
+      "epoch": 1.8028012758285952,
+      "grad_norm": 0.26683637499809265,
+      "learning_rate": 0.00013297697368421052,
+      "loss": 0.5549,
+      "step": 1625
+    },
+    {
+      "epoch": 1.8039106919983359,
+      "grad_norm": 0.2577199339866638,
+      "learning_rate": 0.00013285361842105262,
+      "loss": 0.5963,
+      "step": 1626
+    },
+    {
+      "epoch": 1.8050201081680766,
+      "grad_norm": 0.30272090435028076,
+      "learning_rate": 0.0001327302631578947,
+      "loss": 0.4767,
+      "step": 1627
+    },
+    {
+      "epoch": 1.8061295243378173,
+      "grad_norm": 0.4484618306159973,
+      "learning_rate": 0.00013260690789473683,
+      "loss": 0.4336,
+      "step": 1628
+    },
+    {
+      "epoch": 1.807238940507558,
+      "grad_norm": 0.3869413733482361,
+      "learning_rate": 0.00013248355263157892,
+      "loss": 0.3799,
+      "step": 1629
+    },
+    {
+      "epoch": 1.8083483566772984,
+      "grad_norm": 0.26756906509399414,
+      "learning_rate": 0.00013236019736842104,
+      "loss": 0.5128,
+      "step": 1630
+    },
+    {
+      "epoch": 1.8094577728470393,
+      "grad_norm": 0.21581970155239105,
+      "learning_rate": 0.00013223684210526314,
+      "loss": 0.5776,
+      "step": 1631
+    },
+    {
+      "epoch": 1.8105671890167798,
+      "grad_norm": 0.5925078392028809,
+      "learning_rate": 0.00013211348684210526,
+      "loss": 0.4828,
+      "step": 1632
+    },
+    {
+      "epoch": 1.8116766051865207,
+      "grad_norm": 0.29944899678230286,
+      "learning_rate": 0.00013199013157894735,
+      "loss": 0.5219,
+      "step": 1633
+    },
+    {
+      "epoch": 1.8127860213562612,
+      "grad_norm": 0.3090478479862213,
+      "learning_rate": 0.00013186677631578947,
+      "loss": 0.7766,
+      "step": 1634
+    },
+    {
+      "epoch": 1.813895437526002,
+      "grad_norm": 0.27250027656555176,
+      "learning_rate": 0.00013174342105263156,
+      "loss": 0.5283,
+      "step": 1635
+    },
+    {
+      "epoch": 1.8150048536957426,
+      "grad_norm": 0.30475983023643494,
+      "learning_rate": 0.00013162006578947366,
+      "loss": 0.4438,
+      "step": 1636
+    },
+    {
+      "epoch": 1.8161142698654833,
+      "grad_norm": 0.4503616690635681,
+      "learning_rate": 0.00013149671052631578,
+      "loss": 0.6286,
+      "step": 1637
+    },
+    {
+      "epoch": 1.817223686035224,
+      "grad_norm": 0.3719213306903839,
+      "learning_rate": 0.0001313733552631579,
+      "loss": 0.4938,
+      "step": 1638
+    },
+    {
+      "epoch": 1.8183331022049647,
+      "grad_norm": 0.2590722143650055,
+      "learning_rate": 0.00013125,
+      "loss": 0.3978,
+      "step": 1639
+    },
+    {
+      "epoch": 1.8194425183747054,
+      "grad_norm": 0.4789052903652191,
+      "learning_rate": 0.00013112664473684209,
+      "loss": 0.4542,
+      "step": 1640
+    },
+    {
+      "epoch": 1.8205519345444459,
+      "grad_norm": 0.3678234815597534,
+      "learning_rate": 0.0001310032894736842,
+      "loss": 0.4437,
+      "step": 1641
+    },
+    {
+      "epoch": 1.8216613507141868,
+      "grad_norm": 0.21813832223415375,
+      "learning_rate": 0.0001308799342105263,
+      "loss": 0.3382,
+      "step": 1642
+    },
+    {
+      "epoch": 1.8227707668839273,
+      "grad_norm": 0.2665456235408783,
+      "learning_rate": 0.00013075657894736842,
+      "loss": 0.3803,
+      "step": 1643
+    },
+    {
+      "epoch": 1.8238801830536682,
+      "grad_norm": 0.26693305373191833,
+      "learning_rate": 0.0001306332236842105,
+      "loss": 0.4494,
+      "step": 1644
+    },
+    {
+      "epoch": 1.8249895992234086,
+      "grad_norm": 0.22977307438850403,
+      "learning_rate": 0.0001305098684210526,
+      "loss": 0.3472,
+      "step": 1645
+    },
+    {
+      "epoch": 1.8260990153931493,
+      "grad_norm": 0.38384175300598145,
+      "learning_rate": 0.00013038651315789473,
+      "loss": 0.7401,
+      "step": 1646
+    },
+    {
+      "epoch": 1.82720843156289,
+      "grad_norm": 0.26645827293395996,
+      "learning_rate": 0.00013026315789473685,
+      "loss": 0.5021,
+      "step": 1647
+    },
+    {
+      "epoch": 1.8283178477326307,
+      "grad_norm": 0.28554990887641907,
+      "learning_rate": 0.00013013980263157894,
+      "loss": 0.3429,
+      "step": 1648
+    },
+    {
+      "epoch": 1.8294272639023714,
+      "grad_norm": 0.23377835750579834,
+      "learning_rate": 0.00013001644736842103,
+      "loss": 0.4639,
+      "step": 1649
+    },
+    {
+      "epoch": 1.830536680072112,
+      "grad_norm": 0.323998361825943,
+      "learning_rate": 0.00012989309210526315,
+      "loss": 0.5517,
+      "step": 1650
+    },
+    {
+      "epoch": 1.8316460962418528,
+      "grad_norm": 0.24397112429141998,
+      "learning_rate": 0.00012976973684210527,
+      "loss": 0.4443,
+      "step": 1651
+    },
+    {
+      "epoch": 1.8327555124115933,
+      "grad_norm": 0.23545107245445251,
+      "learning_rate": 0.00012964638157894737,
+      "loss": 0.6529,
+      "step": 1652
+    },
+    {
+      "epoch": 1.8338649285813342,
+      "grad_norm": 0.3781031668186188,
+      "learning_rate": 0.00012952302631578946,
+      "loss": 0.3811,
+      "step": 1653
+    },
+    {
+      "epoch": 1.8349743447510747,
+      "grad_norm": 0.3218782842159271,
+      "learning_rate": 0.00012939967105263155,
+      "loss": 0.5818,
+      "step": 1654
+    },
+    {
+      "epoch": 1.8360837609208154,
+      "grad_norm": 0.31816890835762024,
+      "learning_rate": 0.00012927631578947367,
+      "loss": 0.515,
+      "step": 1655
+    },
+    {
+      "epoch": 1.837193177090556,
+      "grad_norm": 0.3660028278827667,
+      "learning_rate": 0.0001291529605263158,
+      "loss": 0.6154,
+      "step": 1656
+    },
+    {
+      "epoch": 1.8383025932602968,
+      "grad_norm": 0.25920218229293823,
+      "learning_rate": 0.00012902960526315789,
+      "loss": 0.5614,
+      "step": 1657
+    },
+    {
+      "epoch": 1.8394120094300375,
+      "grad_norm": 0.3921451270580292,
+      "learning_rate": 0.00012890624999999998,
+      "loss": 0.5109,
+      "step": 1658
+    },
+    {
+      "epoch": 1.8405214255997782,
+      "grad_norm": 0.30347323417663574,
+      "learning_rate": 0.0001287828947368421,
+      "loss": 0.5139,
+      "step": 1659
+    },
+    {
+      "epoch": 1.8416308417695189,
+      "grad_norm": 0.36900901794433594,
+      "learning_rate": 0.0001286595394736842,
+      "loss": 0.3862,
+      "step": 1660
+    },
+    {
+      "epoch": 1.8427402579392593,
+      "grad_norm": 0.23627950251102448,
+      "learning_rate": 0.0001285361842105263,
+      "loss": 0.3962,
+      "step": 1661
+    },
+    {
+      "epoch": 1.8438496741090002,
+      "grad_norm": 0.3626163601875305,
+      "learning_rate": 0.0001284128289473684,
+      "loss": 0.5779,
+      "step": 1662
+    },
+    {
+      "epoch": 1.8449590902787407,
+      "grad_norm": 0.3031785488128662,
+      "learning_rate": 0.00012828947368421053,
+      "loss": 0.4422,
+      "step": 1663
+    },
+    {
+      "epoch": 1.8460685064484816,
+      "grad_norm": 0.27116191387176514,
+      "learning_rate": 0.00012816611842105262,
+      "loss": 0.4825,
+      "step": 1664
+    },
+    {
+      "epoch": 1.8471779226182221,
+      "grad_norm": 0.23863159120082855,
+      "learning_rate": 0.0001280427631578947,
+      "loss": 0.6105,
+      "step": 1665
+    },
+    {
+      "epoch": 1.8482873387879628,
+      "grad_norm": 0.3026638329029083,
+      "learning_rate": 0.00012791940789473683,
+      "loss": 0.5599,
+      "step": 1666
+    },
+    {
+      "epoch": 1.8493967549577035,
+      "grad_norm": 0.2904566526412964,
+      "learning_rate": 0.00012779605263157893,
+      "loss": 0.4156,
+      "step": 1667
+    },
+    {
+      "epoch": 1.8505061711274442,
+      "grad_norm": 0.2892657518386841,
+      "learning_rate": 0.00012767269736842105,
+      "loss": 0.3779,
+      "step": 1668
+    },
+    {
+      "epoch": 1.851615587297185,
+      "grad_norm": 0.27468252182006836,
+      "learning_rate": 0.00012754934210526314,
+      "loss": 0.4811,
+      "step": 1669
+    },
+    {
+      "epoch": 1.8527250034669254,
+      "grad_norm": 0.33178287744522095,
+      "learning_rate": 0.00012742598684210526,
+      "loss": 0.6331,
+      "step": 1670
+    },
+    {
+      "epoch": 1.8538344196366663,
+      "grad_norm": 0.3048073351383209,
+      "learning_rate": 0.00012730263157894735,
+      "loss": 0.4763,
+      "step": 1671
+    },
+    {
+      "epoch": 1.8549438358064068,
+      "grad_norm": 0.2505081593990326,
+      "learning_rate": 0.00012717927631578947,
+      "loss": 0.6395,
+      "step": 1672
+    },
+    {
+      "epoch": 1.8560532519761477,
+      "grad_norm": 0.3426123261451721,
+      "learning_rate": 0.00012705592105263157,
+      "loss": 0.4118,
+      "step": 1673
+    },
+    {
+      "epoch": 1.8571626681458882,
+      "grad_norm": 0.2770869731903076,
+      "learning_rate": 0.00012693256578947366,
+      "loss": 0.4372,
+      "step": 1674
+    },
+    {
+      "epoch": 1.8582720843156288,
+      "grad_norm": 0.28371554613113403,
+      "learning_rate": 0.00012680921052631578,
+      "loss": 0.5352,
+      "step": 1675
+    },
+    {
+      "epoch": 1.8593815004853695,
+      "grad_norm": 0.424926221370697,
+      "learning_rate": 0.0001266858552631579,
+      "loss": 0.5825,
+      "step": 1676
+    },
+    {
+      "epoch": 1.8604909166551102,
+      "grad_norm": 0.42135924100875854,
+      "learning_rate": 0.0001265625,
+      "loss": 0.54,
+      "step": 1677
+    },
+    {
+      "epoch": 1.861600332824851,
+      "grad_norm": 0.35227394104003906,
+      "learning_rate": 0.00012643914473684209,
+      "loss": 0.4149,
+      "step": 1678
+    },
+    {
+      "epoch": 1.8627097489945916,
+      "grad_norm": 0.368327796459198,
+      "learning_rate": 0.00012631578947368418,
+      "loss": 0.5194,
+      "step": 1679
+    },
+    {
+      "epoch": 1.8638191651643323,
+      "grad_norm": 0.31259453296661377,
+      "learning_rate": 0.0001261924342105263,
+      "loss": 0.5897,
+      "step": 1680
+    },
+    {
+      "epoch": 1.8649285813340728,
+      "grad_norm": 0.42234233021736145,
+      "learning_rate": 0.00012606907894736842,
+      "loss": 0.3769,
+      "step": 1681
+    },
+    {
+      "epoch": 1.8660379975038137,
+      "grad_norm": 0.258651465177536,
+      "learning_rate": 0.0001259457236842105,
+      "loss": 0.443,
+      "step": 1682
+    },
+    {
+      "epoch": 1.8671474136735542,
+      "grad_norm": 0.3242909014225006,
+      "learning_rate": 0.0001258223684210526,
+      "loss": 0.5423,
+      "step": 1683
+    },
+    {
+      "epoch": 1.868256829843295,
+      "grad_norm": 0.3746740221977234,
+      "learning_rate": 0.00012569901315789473,
+      "loss": 0.4365,
+      "step": 1684
+    },
+    {
+      "epoch": 1.8693662460130356,
+      "grad_norm": 0.2767789363861084,
+      "learning_rate": 0.00012557565789473685,
+      "loss": 0.5011,
+      "step": 1685
+    },
+    {
+      "epoch": 1.8704756621827763,
+      "grad_norm": 0.41377684473991394,
+      "learning_rate": 0.00012545230263157894,
+      "loss": 0.4733,
+      "step": 1686
+    },
+    {
+      "epoch": 1.871585078352517,
+      "grad_norm": 0.2723773717880249,
+      "learning_rate": 0.00012532894736842103,
+      "loss": 0.5801,
+      "step": 1687
+    },
+    {
+      "epoch": 1.8726944945222577,
+      "grad_norm": 0.28274834156036377,
+      "learning_rate": 0.00012520559210526315,
+      "loss": 0.6009,
+      "step": 1688
+    },
+    {
+      "epoch": 1.8738039106919984,
+      "grad_norm": 0.3209463655948639,
+      "learning_rate": 0.00012508223684210525,
+      "loss": 0.5851,
+      "step": 1689
+    },
+    {
+      "epoch": 1.874913326861739,
+      "grad_norm": 0.24118223786354065,
+      "learning_rate": 0.00012495888157894737,
+      "loss": 0.6186,
+      "step": 1690
+    },
+    {
+      "epoch": 1.8760227430314798,
+      "grad_norm": 0.3865971565246582,
+      "learning_rate": 0.00012483552631578946,
+      "loss": 0.5411,
+      "step": 1691
+    },
+    {
+      "epoch": 1.8771321592012202,
+      "grad_norm": 0.23888447880744934,
+      "learning_rate": 0.00012471217105263155,
+      "loss": 0.3907,
+      "step": 1692
+    },
+    {
+      "epoch": 1.8782415753709611,
+      "grad_norm": 0.290234237909317,
+      "learning_rate": 0.00012458881578947367,
+      "loss": 0.3822,
+      "step": 1693
+    },
+    {
+      "epoch": 1.8793509915407016,
+      "grad_norm": 0.2845550775527954,
+      "learning_rate": 0.0001244654605263158,
+      "loss": 0.4248,
+      "step": 1694
+    },
+    {
+      "epoch": 1.8804604077104425,
+      "grad_norm": 0.3374759554862976,
+      "learning_rate": 0.0001243421052631579,
+      "loss": 0.7508,
+      "step": 1695
+    },
+    {
+      "epoch": 1.881569823880183,
+      "grad_norm": 0.39034581184387207,
+      "learning_rate": 0.00012421874999999998,
+      "loss": 0.4806,
+      "step": 1696
+    },
+    {
+      "epoch": 1.8826792400499237,
+      "grad_norm": 0.39774978160858154,
+      "learning_rate": 0.0001240953947368421,
+      "loss": 0.2821,
+      "step": 1697
+    },
+    {
+      "epoch": 1.8837886562196644,
+      "grad_norm": 0.31861943006515503,
+      "learning_rate": 0.0001239720394736842,
+      "loss": 0.487,
+      "step": 1698
+    },
+    {
+      "epoch": 1.884898072389405,
+      "grad_norm": 0.3267800807952881,
+      "learning_rate": 0.00012384868421052631,
+      "loss": 0.4185,
+      "step": 1699
+    },
+    {
+      "epoch": 1.8860074885591458,
+      "grad_norm": 0.28482627868652344,
+      "learning_rate": 0.0001237253289473684,
+      "loss": 0.5111,
+      "step": 1700
+    },
+    {
+      "epoch": 1.8871169047288863,
+      "grad_norm": 0.3203260898590088,
+      "learning_rate": 0.00012360197368421053,
+      "loss": 0.465,
+      "step": 1701
+    },
+    {
+      "epoch": 1.8882263208986272,
+      "grad_norm": 0.2945539057254791,
+      "learning_rate": 0.00012347861842105262,
+      "loss": 0.615,
+      "step": 1702
+    },
+    {
+      "epoch": 1.8893357370683677,
+      "grad_norm": 0.29036056995391846,
+      "learning_rate": 0.00012335526315789471,
+      "loss": 0.5022,
+      "step": 1703
+    },
+    {
+      "epoch": 1.8904451532381086,
+      "grad_norm": 0.22323249280452728,
+      "learning_rate": 0.00012323190789473683,
+      "loss": 0.5774,
+      "step": 1704
+    },
+    {
+      "epoch": 1.891554569407849,
+      "grad_norm": 0.3879876434803009,
+      "learning_rate": 0.00012310855263157893,
+      "loss": 0.732,
+      "step": 1705
+    },
+    {
+      "epoch": 1.8926639855775897,
+      "grad_norm": 0.49169594049453735,
+      "learning_rate": 0.00012298519736842105,
+      "loss": 0.6079,
+      "step": 1706
+    },
+    {
+      "epoch": 1.8937734017473304,
+      "grad_norm": 0.32131388783454895,
+      "learning_rate": 0.00012286184210526314,
+      "loss": 0.4357,
+      "step": 1707
+    },
+    {
+      "epoch": 1.8948828179170711,
+      "grad_norm": 0.2757743299007416,
+      "learning_rate": 0.00012273848684210526,
+      "loss": 0.4277,
+      "step": 1708
+    },
+    {
+      "epoch": 1.8959922340868118,
+      "grad_norm": 0.2627353370189667,
+      "learning_rate": 0.00012261513157894735,
+      "loss": 0.4186,
+      "step": 1709
+    },
+    {
+      "epoch": 1.8971016502565525,
+      "grad_norm": 0.24395854771137238,
+      "learning_rate": 0.00012249177631578947,
+      "loss": 0.4422,
+      "step": 1710
+    },
+    {
+      "epoch": 1.8982110664262932,
+      "grad_norm": 0.2920277416706085,
+      "learning_rate": 0.00012236842105263157,
+      "loss": 0.3717,
+      "step": 1711
+    },
+    {
+      "epoch": 1.8993204825960337,
+      "grad_norm": 0.22231972217559814,
+      "learning_rate": 0.00012224506578947366,
+      "loss": 0.4537,
+      "step": 1712
+    },
+    {
+      "epoch": 1.9004298987657746,
+      "grad_norm": 0.2766577899456024,
+      "learning_rate": 0.00012212171052631578,
+      "loss": 0.4264,
+      "step": 1713
+    },
+    {
+      "epoch": 1.901539314935515,
+      "grad_norm": 0.30484381318092346,
+      "learning_rate": 0.00012199835526315789,
+      "loss": 0.4421,
+      "step": 1714
+    },
+    {
+      "epoch": 1.902648731105256,
+      "grad_norm": 0.36658528447151184,
+      "learning_rate": 0.000121875,
+      "loss": 0.4917,
+      "step": 1715
+    },
+    {
+      "epoch": 1.9037581472749965,
+      "grad_norm": 0.38978394865989685,
+      "learning_rate": 0.00012175164473684209,
+      "loss": 0.5901,
+      "step": 1716
+    },
+    {
+      "epoch": 1.9048675634447372,
+      "grad_norm": 0.3046998381614685,
+      "learning_rate": 0.0001216282894736842,
+      "loss": 0.4495,
+      "step": 1717
+    },
+    {
+      "epoch": 1.9059769796144779,
+      "grad_norm": 0.31667011976242065,
+      "learning_rate": 0.00012150493421052631,
+      "loss": 0.4073,
+      "step": 1718
+    },
+    {
+      "epoch": 1.9070863957842186,
+      "grad_norm": 0.3211687207221985,
+      "learning_rate": 0.00012138157894736841,
+      "loss": 0.4975,
+      "step": 1719
+    },
+    {
+      "epoch": 1.9081958119539593,
+      "grad_norm": 0.2827535569667816,
+      "learning_rate": 0.00012125822368421051,
+      "loss": 0.4363,
+      "step": 1720
+    },
+    {
+      "epoch": 1.9093052281236997,
+      "grad_norm": 0.28672489523887634,
+      "learning_rate": 0.00012113486842105262,
+      "loss": 0.5762,
+      "step": 1721
+    },
+    {
+      "epoch": 1.9104146442934407,
+      "grad_norm": 0.3268757462501526,
+      "learning_rate": 0.00012101151315789471,
+      "loss": 0.4126,
+      "step": 1722
+    },
+    {
+      "epoch": 1.9115240604631811,
+      "grad_norm": 0.3554566204547882,
+      "learning_rate": 0.00012088815789473683,
+      "loss": 0.5062,
+      "step": 1723
+    },
+    {
+      "epoch": 1.912633476632922,
+      "grad_norm": 0.3198055326938629,
+      "learning_rate": 0.00012076480263157894,
+      "loss": 0.4958,
+      "step": 1724
+    },
+    {
+      "epoch": 1.9137428928026625,
+      "grad_norm": 0.3657841980457306,
+      "learning_rate": 0.00012064144736842103,
+      "loss": 0.5931,
+      "step": 1725
+    },
+    {
+      "epoch": 1.9148523089724034,
+      "grad_norm": 0.45995911955833435,
+      "learning_rate": 0.00012051809210526314,
+      "loss": 0.5862,
+      "step": 1726
+    },
+    {
+      "epoch": 1.915961725142144,
+      "grad_norm": 0.4919174015522003,
+      "learning_rate": 0.00012039473684210526,
+      "loss": 0.452,
+      "step": 1727
+    },
+    {
+      "epoch": 1.9170711413118846,
+      "grad_norm": 0.3233271539211273,
+      "learning_rate": 0.00012027138157894737,
+      "loss": 0.6559,
+      "step": 1728
+    },
+    {
+      "epoch": 1.9181805574816253,
+      "grad_norm": 0.396419495344162,
+      "learning_rate": 0.00012014802631578946,
+      "loss": 0.7268,
+      "step": 1729
+    },
+    {
+      "epoch": 1.919289973651366,
+      "grad_norm": 0.2332264930009842,
+      "learning_rate": 0.00012002467105263157,
+      "loss": 0.3874,
+      "step": 1730
+    },
+    {
+      "epoch": 1.9203993898211067,
+      "grad_norm": 0.20889733731746674,
+      "learning_rate": 0.00011990131578947366,
+      "loss": 0.5648,
+      "step": 1731
+    },
+    {
+      "epoch": 1.9215088059908472,
+      "grad_norm": 0.25143593549728394,
+      "learning_rate": 0.00011977796052631578,
+      "loss": 0.4837,
+      "step": 1732
+    },
+    {
+      "epoch": 1.922618222160588,
+      "grad_norm": 0.2032875120639801,
+      "learning_rate": 0.00011965460526315789,
+      "loss": 0.4553,
+      "step": 1733
+    },
+    {
+      "epoch": 1.9237276383303286,
+      "grad_norm": 0.2216006964445114,
+      "learning_rate": 0.00011953125,
+      "loss": 0.3256,
+      "step": 1734
+    },
+    {
+      "epoch": 1.9248370545000695,
+      "grad_norm": 0.34091660380363464,
+      "learning_rate": 0.00011940789473684209,
+      "loss": 0.534,
+      "step": 1735
+    },
+    {
+      "epoch": 1.92594647066981,
+      "grad_norm": 0.35081061720848083,
+      "learning_rate": 0.0001192845394736842,
+      "loss": 0.4549,
+      "step": 1736
+    },
+    {
+      "epoch": 1.9270558868395506,
+      "grad_norm": 0.37153178453445435,
+      "learning_rate": 0.00011916118421052632,
+      "loss": 0.507,
+      "step": 1737
+    },
+    {
+      "epoch": 1.9281653030092913,
+      "grad_norm": 0.3207988142967224,
+      "learning_rate": 0.00011903782894736841,
+      "loss": 0.6104,
+      "step": 1738
+    },
+    {
+      "epoch": 1.929274719179032,
+      "grad_norm": 0.378360390663147,
+      "learning_rate": 0.00011891447368421052,
+      "loss": 0.5371,
+      "step": 1739
+    },
+    {
+      "epoch": 1.9303841353487727,
+      "grad_norm": 0.3643793761730194,
+      "learning_rate": 0.00011879111842105262,
+      "loss": 0.506,
+      "step": 1740
+    },
+    {
+      "epoch": 1.9314935515185134,
+      "grad_norm": 0.35685864090919495,
+      "learning_rate": 0.00011866776315789471,
+      "loss": 0.6205,
+      "step": 1741
+    },
+    {
+      "epoch": 1.9326029676882541,
+      "grad_norm": 0.349833220243454,
+      "learning_rate": 0.00011854440789473684,
+      "loss": 0.587,
+      "step": 1742
+    },
+    {
+      "epoch": 1.9337123838579946,
+      "grad_norm": 0.3674916923046112,
+      "learning_rate": 0.00011842105263157894,
+      "loss": 0.5616,
+      "step": 1743
+    },
+    {
+      "epoch": 1.9348218000277355,
+      "grad_norm": 0.4197103679180145,
+      "learning_rate": 0.00011829769736842104,
+      "loss": 0.6715,
+      "step": 1744
+    },
+    {
+      "epoch": 1.935931216197476,
+      "grad_norm": 0.2582911550998688,
+      "learning_rate": 0.00011817434210526314,
+      "loss": 0.3392,
+      "step": 1745
+    },
+    {
+      "epoch": 1.937040632367217,
+      "grad_norm": 0.3199860751628876,
+      "learning_rate": 0.00011805098684210526,
+      "loss": 0.4765,
+      "step": 1746
+    },
+    {
+      "epoch": 1.9381500485369574,
+      "grad_norm": 0.28448477387428284,
+      "learning_rate": 0.00011792763157894737,
+      "loss": 0.5821,
+      "step": 1747
+    },
+    {
+      "epoch": 1.939259464706698,
+      "grad_norm": 0.4114968180656433,
+      "learning_rate": 0.00011780427631578946,
+      "loss": 0.48,
+      "step": 1748
+    },
+    {
+      "epoch": 1.9403688808764388,
+      "grad_norm": 0.3065422773361206,
+      "learning_rate": 0.00011768092105263157,
+      "loss": 0.4095,
+      "step": 1749
+    },
+    {
+      "epoch": 1.9414782970461795,
+      "grad_norm": 0.33260229229927063,
+      "learning_rate": 0.00011755756578947366,
+      "loss": 0.5081,
+      "step": 1750
+    },
+    {
+      "epoch": 1.9425877132159202,
+      "grad_norm": 0.3044232130050659,
+      "learning_rate": 0.00011743421052631578,
+      "loss": 0.6083,
+      "step": 1751
+    },
+    {
+      "epoch": 1.9436971293856606,
+      "grad_norm": 0.3583667278289795,
+      "learning_rate": 0.00011731085526315789,
+      "loss": 0.6418,
+      "step": 1752
+    },
+    {
+      "epoch": 1.9448065455554016,
+      "grad_norm": 0.5549653172492981,
+      "learning_rate": 0.0001171875,
+      "loss": 0.8511,
+      "step": 1753
+    },
+    {
+      "epoch": 1.945915961725142,
+      "grad_norm": 0.24391904473304749,
+      "learning_rate": 0.00011706414473684209,
+      "loss": 0.4204,
+      "step": 1754
+    },
+    {
+      "epoch": 1.947025377894883,
+      "grad_norm": 0.4001742899417877,
+      "learning_rate": 0.0001169407894736842,
+      "loss": 0.4757,
+      "step": 1755
+    },
+    {
+      "epoch": 1.9481347940646234,
+      "grad_norm": 0.49259909987449646,
+      "learning_rate": 0.00011681743421052632,
+      "loss": 0.4522,
+      "step": 1756
+    },
+    {
+      "epoch": 1.9492442102343641,
+      "grad_norm": 0.3102129399776459,
+      "learning_rate": 0.00011669407894736841,
+      "loss": 0.3831,
+      "step": 1757
+    },
+    {
+      "epoch": 1.9503536264041048,
+      "grad_norm": 0.2639727294445038,
+      "learning_rate": 0.00011657072368421052,
+      "loss": 0.6364,
+      "step": 1758
+    },
+    {
+      "epoch": 1.9514630425738455,
+      "grad_norm": 0.31318995356559753,
+      "learning_rate": 0.00011644736842105262,
+      "loss": 0.5422,
+      "step": 1759
+    },
+    {
+      "epoch": 1.9525724587435862,
+      "grad_norm": 0.2877756357192993,
+      "learning_rate": 0.00011632401315789472,
+      "loss": 0.3984,
+      "step": 1760
+    },
+    {
+      "epoch": 1.953681874913327,
+      "grad_norm": 0.36178058385849,
+      "learning_rate": 0.00011620065789473684,
+      "loss": 0.6806,
+      "step": 1761
+    },
+    {
+      "epoch": 1.9547912910830676,
+      "grad_norm": 0.36867088079452515,
+      "learning_rate": 0.00011607730263157894,
+      "loss": 0.529,
+      "step": 1762
+    },
+    {
+      "epoch": 1.955900707252808,
+      "grad_norm": 0.25498855113983154,
+      "learning_rate": 0.00011595394736842104,
+      "loss": 0.6411,
+      "step": 1763
+    },
+    {
+      "epoch": 1.957010123422549,
+      "grad_norm": 0.26043468713760376,
+      "learning_rate": 0.00011583059210526314,
+      "loss": 0.5102,
+      "step": 1764
+    },
+    {
+      "epoch": 1.9581195395922895,
+      "grad_norm": 0.40660566091537476,
+      "learning_rate": 0.00011570723684210526,
+      "loss": 0.5582,
+      "step": 1765
+    },
+    {
+      "epoch": 1.9592289557620304,
+      "grad_norm": 0.4207366406917572,
+      "learning_rate": 0.00011558388157894736,
+      "loss": 0.7263,
+      "step": 1766
+    },
+    {
+      "epoch": 1.9603383719317709,
+      "grad_norm": 0.35944870114326477,
+      "learning_rate": 0.00011546052631578946,
+      "loss": 0.4644,
+      "step": 1767
+    },
+    {
+      "epoch": 1.9614477881015115,
+      "grad_norm": 0.2992507219314575,
+      "learning_rate": 0.00011533717105263157,
+      "loss": 0.7683,
+      "step": 1768
+    },
+    {
+      "epoch": 1.9625572042712522,
+      "grad_norm": 0.3475952744483948,
+      "learning_rate": 0.00011521381578947366,
+      "loss": 0.6288,
+      "step": 1769
+    },
+    {
+      "epoch": 1.963666620440993,
+      "grad_norm": 0.29175207018852234,
+      "learning_rate": 0.00011509046052631578,
+      "loss": 0.355,
+      "step": 1770
+    },
+    {
+      "epoch": 1.9647760366107336,
+      "grad_norm": 0.3024480640888214,
+      "learning_rate": 0.00011496710526315789,
+      "loss": 0.5916,
+      "step": 1771
+    },
+    {
+      "epoch": 1.9658854527804743,
+      "grad_norm": 0.310245543718338,
+      "learning_rate": 0.00011484375,
+      "loss": 0.6658,
+      "step": 1772
+    },
+    {
+      "epoch": 1.966994868950215,
+      "grad_norm": 0.2858862578868866,
+      "learning_rate": 0.00011472039473684209,
+      "loss": 0.5808,
+      "step": 1773
+    },
+    {
+      "epoch": 1.9681042851199555,
+      "grad_norm": 0.19843228161334991,
+      "learning_rate": 0.0001145970394736842,
+      "loss": 0.3837,
+      "step": 1774
+    },
+    {
+      "epoch": 1.9692137012896964,
+      "grad_norm": 0.37114304304122925,
+      "learning_rate": 0.00011447368421052632,
+      "loss": 0.7252,
+      "step": 1775
+    },
+    {
+      "epoch": 1.970323117459437,
+      "grad_norm": 0.3807290196418762,
+      "learning_rate": 0.00011435032894736841,
+      "loss": 0.4429,
+      "step": 1776
+    },
+    {
+      "epoch": 1.9714325336291778,
+      "grad_norm": 0.2850121557712555,
+      "learning_rate": 0.00011422697368421052,
+      "loss": 0.3464,
+      "step": 1777
+    },
+    {
+      "epoch": 1.9725419497989183,
+      "grad_norm": 0.34873002767562866,
+      "learning_rate": 0.00011410361842105262,
+      "loss": 0.4814,
+      "step": 1778
+    },
+    {
+      "epoch": 1.973651365968659,
+      "grad_norm": 0.42871007323265076,
+      "learning_rate": 0.00011398026315789472,
+      "loss": 0.4912,
+      "step": 1779
+    },
+    {
+      "epoch": 1.9747607821383997,
+      "grad_norm": 0.3286532163619995,
+      "learning_rate": 0.00011385690789473684,
+      "loss": 0.4862,
+      "step": 1780
+    },
+    {
+      "epoch": 1.9758701983081404,
+      "grad_norm": 0.3135276436805725,
+      "learning_rate": 0.00011373355263157894,
+      "loss": 0.3872,
+      "step": 1781
+    },
+    {
+      "epoch": 1.976979614477881,
+      "grad_norm": 0.37062501907348633,
+      "learning_rate": 0.00011361019736842104,
+      "loss": 0.5004,
+      "step": 1782
+    },
+    {
+      "epoch": 1.9780890306476215,
+      "grad_norm": 0.28763630986213684,
+      "learning_rate": 0.00011348684210526314,
+      "loss": 0.5367,
+      "step": 1783
+    },
+    {
+      "epoch": 1.9791984468173625,
+      "grad_norm": 0.34978562593460083,
+      "learning_rate": 0.00011336348684210526,
+      "loss": 0.5041,
+      "step": 1784
+    },
+    {
+      "epoch": 1.980307862987103,
+      "grad_norm": 0.2940448820590973,
+      "learning_rate": 0.00011324013157894736,
+      "loss": 0.5121,
+      "step": 1785
+    },
+    {
+      "epoch": 1.9814172791568438,
+      "grad_norm": 0.24150650203227997,
+      "learning_rate": 0.00011311677631578946,
+      "loss": 0.4354,
+      "step": 1786
+    },
+    {
+      "epoch": 1.9825266953265843,
+      "grad_norm": 0.24752016365528107,
+      "learning_rate": 0.00011299342105263157,
+      "loss": 0.3303,
+      "step": 1787
+    },
+    {
+      "epoch": 1.983636111496325,
+      "grad_norm": 0.2988849878311157,
+      "learning_rate": 0.00011287006578947366,
+      "loss": 0.5873,
+      "step": 1788
+    },
+    {
+      "epoch": 1.9847455276660657,
+      "grad_norm": 0.548851490020752,
+      "learning_rate": 0.00011274671052631578,
+      "loss": 0.6516,
+      "step": 1789
+    },
+    {
+      "epoch": 1.9858549438358064,
+      "grad_norm": 0.3005162477493286,
+      "learning_rate": 0.00011262335526315789,
+      "loss": 0.4203,
+      "step": 1790
+    },
+    {
+      "epoch": 1.986964360005547,
+      "grad_norm": 0.3434782922267914,
+      "learning_rate": 0.0001125,
+      "loss": 0.6023,
+      "step": 1791
+    },
+    {
+      "epoch": 1.9880737761752878,
+      "grad_norm": 0.27085399627685547,
+      "learning_rate": 0.00011237664473684209,
+      "loss": 0.4336,
+      "step": 1792
+    },
+    {
+      "epoch": 1.9891831923450285,
+      "grad_norm": 0.24659699201583862,
+      "learning_rate": 0.0001122532894736842,
+      "loss": 0.4351,
+      "step": 1793
+    },
+    {
+      "epoch": 1.990292608514769,
+      "grad_norm": 0.2878054976463318,
+      "learning_rate": 0.00011212993421052632,
+      "loss": 0.3109,
+      "step": 1794
+    },
+    {
+      "epoch": 1.9914020246845099,
+      "grad_norm": 0.2754107117652893,
+      "learning_rate": 0.00011200657894736841,
+      "loss": 0.5065,
+      "step": 1795
+    },
+    {
+      "epoch": 1.9925114408542504,
+      "grad_norm": 0.31422141194343567,
+      "learning_rate": 0.00011188322368421052,
+      "loss": 0.5294,
+      "step": 1796
+    },
+    {
+      "epoch": 1.9936208570239913,
+      "grad_norm": 0.2437220960855484,
+      "learning_rate": 0.00011175986842105262,
+      "loss": 0.4636,
+      "step": 1797
+    },
+    {
+      "epoch": 1.9947302731937318,
+      "grad_norm": 0.3113705515861511,
+      "learning_rate": 0.00011163651315789472,
+      "loss": 0.4084,
+      "step": 1798
+    },
+    {
+      "epoch": 1.9958396893634724,
+      "grad_norm": 0.2959713935852051,
+      "learning_rate": 0.00011151315789473684,
+      "loss": 0.6161,
+      "step": 1799
+    },
+    {
+      "epoch": 1.9969491055332131,
+      "grad_norm": 0.29905256628990173,
+      "learning_rate": 0.00011138980263157894,
+      "loss": 0.3853,
+      "step": 1800
+    },
+    {
+      "epoch": 1.9980585217029538,
+      "grad_norm": 0.3135545551776886,
+      "learning_rate": 0.00011126644736842104,
+      "loss": 0.5653,
+      "step": 1801
+    },
+    {
+      "epoch": 1.9991679378726945,
+      "grad_norm": 0.3632647395133972,
+      "learning_rate": 0.00011114309210526314,
+      "loss": 0.3835,
+      "step": 1802
+    },
+    {
+      "epoch": 2.000277354042435,
+      "grad_norm": 0.3683667480945587,
+      "learning_rate": 0.00011101973684210526,
+      "loss": 0.4497,
+      "step": 1803
+    },
+    {
+      "epoch": 2.001386770212176,
+      "grad_norm": 0.26978781819343567,
+      "learning_rate": 0.00011089638157894736,
+      "loss": 0.584,
+      "step": 1804
+    },
+    {
+      "epoch": 2.0024961863819164,
+      "grad_norm": 0.2260834127664566,
+      "learning_rate": 0.00011077302631578946,
+      "loss": 0.5175,
+      "step": 1805
+    },
+    {
+      "epoch": 2.0036056025516573,
+      "grad_norm": 0.2791745364665985,
+      "learning_rate": 0.00011064967105263157,
+      "loss": 0.5489,
+      "step": 1806
+    },
+    {
+      "epoch": 2.004715018721398,
+      "grad_norm": 0.4569042921066284,
+      "learning_rate": 0.00011052631578947366,
+      "loss": 0.4872,
+      "step": 1807
+    },
+    {
+      "epoch": 2.0058244348911387,
+      "grad_norm": 0.2634184956550598,
+      "learning_rate": 0.00011040296052631578,
+      "loss": 0.4137,
+      "step": 1808
+    },
+    {
+      "epoch": 2.006933851060879,
+      "grad_norm": 0.3725602626800537,
+      "learning_rate": 0.00011027960526315789,
+      "loss": 0.3663,
+      "step": 1809
+    },
+    {
+      "epoch": 2.00804326723062,
+      "grad_norm": 0.19589465856552124,
+      "learning_rate": 0.00011015624999999998,
+      "loss": 0.3241,
+      "step": 1810
+    },
+    {
+      "epoch": 2.0091526834003606,
+      "grad_norm": 0.2446906864643097,
+      "learning_rate": 0.00011003289473684209,
+      "loss": 0.3818,
+      "step": 1811
+    },
+    {
+      "epoch": 2.010262099570101,
+      "grad_norm": 0.2932548224925995,
+      "learning_rate": 0.0001099095394736842,
+      "loss": 0.4009,
+      "step": 1812
+    },
+    {
+      "epoch": 2.011371515739842,
+      "grad_norm": 0.23010744154453278,
+      "learning_rate": 0.00010978618421052632,
+      "loss": 0.5868,
+      "step": 1813
+    },
+    {
+      "epoch": 2.0124809319095824,
+      "grad_norm": 0.24582666158676147,
+      "learning_rate": 0.00010966282894736841,
+      "loss": 0.2866,
+      "step": 1814
+    },
+    {
+      "epoch": 2.0135903480793234,
+      "grad_norm": 0.2688146233558655,
+      "learning_rate": 0.00010953947368421052,
+      "loss": 0.3321,
+      "step": 1815
+    },
+    {
+      "epoch": 2.014699764249064,
+      "grad_norm": 0.35448578000068665,
+      "learning_rate": 0.00010941611842105262,
+      "loss": 0.5611,
+      "step": 1816
+    },
+    {
+      "epoch": 2.0158091804188047,
+      "grad_norm": 0.3180113732814789,
+      "learning_rate": 0.00010929276315789472,
+      "loss": 0.3527,
+      "step": 1817
+    },
+    {
+      "epoch": 2.016918596588545,
+      "grad_norm": 0.27800217270851135,
+      "learning_rate": 0.00010916940789473684,
+      "loss": 0.339,
+      "step": 1818
+    },
+    {
+      "epoch": 2.018028012758286,
+      "grad_norm": 0.34227412939071655,
+      "learning_rate": 0.00010904605263157894,
+      "loss": 0.4474,
+      "step": 1819
+    },
+    {
+      "epoch": 2.0191374289280266,
+      "grad_norm": 0.3180390000343323,
+      "learning_rate": 0.00010892269736842104,
+      "loss": 0.5412,
+      "step": 1820
+    },
+    {
+      "epoch": 2.020246845097767,
+      "grad_norm": 0.4531157314777374,
+      "learning_rate": 0.00010879934210526314,
+      "loss": 0.4999,
+      "step": 1821
+    },
+    {
+      "epoch": 2.021356261267508,
+      "grad_norm": 0.3139798045158386,
+      "learning_rate": 0.00010867598684210526,
+      "loss": 0.3725,
+      "step": 1822
+    },
+    {
+      "epoch": 2.0224656774372485,
+      "grad_norm": 0.2892252206802368,
+      "learning_rate": 0.00010855263157894736,
+      "loss": 0.462,
+      "step": 1823
+    },
+    {
+      "epoch": 2.0235750936069894,
+      "grad_norm": 0.34606751799583435,
+      "learning_rate": 0.00010842927631578946,
+      "loss": 0.2938,
+      "step": 1824
+    },
+    {
+      "epoch": 2.02468450977673,
+      "grad_norm": 0.3713940680027008,
+      "learning_rate": 0.00010830592105263157,
+      "loss": 0.3484,
+      "step": 1825
+    },
+    {
+      "epoch": 2.025793925946471,
+      "grad_norm": 0.2926501929759979,
+      "learning_rate": 0.00010818256578947366,
+      "loss": 0.3333,
+      "step": 1826
+    },
+    {
+      "epoch": 2.0269033421162113,
+      "grad_norm": 0.29994428157806396,
+      "learning_rate": 0.00010805921052631578,
+      "loss": 0.5876,
+      "step": 1827
+    },
+    {
+      "epoch": 2.028012758285952,
+      "grad_norm": 0.26852184534072876,
+      "learning_rate": 0.00010793585526315789,
+      "loss": 0.3795,
+      "step": 1828
+    },
+    {
+      "epoch": 2.0291221744556927,
+      "grad_norm": 0.3186289072036743,
+      "learning_rate": 0.00010781249999999998,
+      "loss": 0.4261,
+      "step": 1829
+    },
+    {
+      "epoch": 2.0302315906254336,
+      "grad_norm": 0.22164680063724518,
+      "learning_rate": 0.00010768914473684209,
+      "loss": 0.4383,
+      "step": 1830
+    },
+    {
+      "epoch": 2.031341006795174,
+      "grad_norm": 0.4684840142726898,
+      "learning_rate": 0.0001075657894736842,
+      "loss": 0.4621,
+      "step": 1831
+    },
+    {
+      "epoch": 2.0324504229649145,
+      "grad_norm": 0.27373453974723816,
+      "learning_rate": 0.00010744243421052632,
+      "loss": 0.4259,
+      "step": 1832
+    },
+    {
+      "epoch": 2.0335598391346554,
+      "grad_norm": 0.3046364188194275,
+      "learning_rate": 0.00010731907894736841,
+      "loss": 0.5675,
+      "step": 1833
+    },
+    {
+      "epoch": 2.034669255304396,
+      "grad_norm": 0.2961323857307434,
+      "learning_rate": 0.00010719572368421052,
+      "loss": 0.3838,
+      "step": 1834
+    },
+    {
+      "epoch": 2.035778671474137,
+      "grad_norm": 0.3641231656074524,
+      "learning_rate": 0.00010707236842105261,
+      "loss": 0.2031,
+      "step": 1835
+    },
+    {
+      "epoch": 2.0368880876438773,
+      "grad_norm": 0.38065147399902344,
+      "learning_rate": 0.00010694901315789472,
+      "loss": 0.3094,
+      "step": 1836
+    },
+    {
+      "epoch": 2.037997503813618,
+      "grad_norm": 0.3846987783908844,
+      "learning_rate": 0.00010682565789473684,
+      "loss": 0.4203,
+      "step": 1837
+    },
+    {
+      "epoch": 2.0391069199833587,
+      "grad_norm": 0.2999848425388336,
+      "learning_rate": 0.00010670230263157895,
+      "loss": 0.3044,
+      "step": 1838
+    },
+    {
+      "epoch": 2.0402163361530996,
+      "grad_norm": 0.4001493453979492,
+      "learning_rate": 0.00010657894736842104,
+      "loss": 0.3758,
+      "step": 1839
+    },
+    {
+      "epoch": 2.04132575232284,
+      "grad_norm": 0.42989227175712585,
+      "learning_rate": 0.00010645559210526315,
+      "loss": 0.4854,
+      "step": 1840
+    },
+    {
+      "epoch": 2.042435168492581,
+      "grad_norm": 0.3566846549510956,
+      "learning_rate": 0.00010633223684210527,
+      "loss": 0.4105,
+      "step": 1841
+    },
+    {
+      "epoch": 2.0435445846623215,
+      "grad_norm": 0.41669943928718567,
+      "learning_rate": 0.00010620888157894736,
+      "loss": 0.4248,
+      "step": 1842
+    },
+    {
+      "epoch": 2.044654000832062,
+      "grad_norm": 0.31254488229751587,
+      "learning_rate": 0.00010608552631578947,
+      "loss": 0.3155,
+      "step": 1843
+    },
+    {
+      "epoch": 2.045763417001803,
+      "grad_norm": 0.2741456925868988,
+      "learning_rate": 0.00010596217105263157,
+      "loss": 0.4663,
+      "step": 1844
+    },
+    {
+      "epoch": 2.0468728331715433,
+      "grad_norm": 0.40784788131713867,
+      "learning_rate": 0.00010583881578947367,
+      "loss": 0.2822,
+      "step": 1845
+    },
+    {
+      "epoch": 2.0479822493412843,
+      "grad_norm": 0.3757185935974121,
+      "learning_rate": 0.00010571546052631579,
+      "loss": 0.4443,
+      "step": 1846
+    },
+    {
+      "epoch": 2.0490916655110247,
+      "grad_norm": 0.38732078671455383,
+      "learning_rate": 0.00010559210526315789,
+      "loss": 0.4025,
+      "step": 1847
+    },
+    {
+      "epoch": 2.0502010816807656,
+      "grad_norm": 0.34661343693733215,
+      "learning_rate": 0.00010546874999999999,
+      "loss": 0.3945,
+      "step": 1848
+    },
+    {
+      "epoch": 2.051310497850506,
+      "grad_norm": 0.41781237721443176,
+      "learning_rate": 0.00010534539473684209,
+      "loss": 0.3046,
+      "step": 1849
+    },
+    {
+      "epoch": 2.052419914020247,
+      "grad_norm": 0.3018251955509186,
+      "learning_rate": 0.0001052220394736842,
+      "loss": 0.3751,
+      "step": 1850
+    },
+    {
+      "epoch": 2.0535293301899875,
+      "grad_norm": 0.2182953655719757,
+      "learning_rate": 0.00010509868421052632,
+      "loss": 0.4354,
+      "step": 1851
+    },
+    {
+      "epoch": 2.054638746359728,
+      "grad_norm": 0.48397496342658997,
+      "learning_rate": 0.00010497532894736841,
+      "loss": 0.362,
+      "step": 1852
+    },
+    {
+      "epoch": 2.055748162529469,
+      "grad_norm": 0.3845345079898834,
+      "learning_rate": 0.00010485197368421052,
+      "loss": 0.3564,
+      "step": 1853
+    },
+    {
+      "epoch": 2.0568575786992094,
+      "grad_norm": 0.2810097932815552,
+      "learning_rate": 0.00010472861842105261,
+      "loss": 0.2759,
+      "step": 1854
+    },
+    {
+      "epoch": 2.0579669948689503,
+      "grad_norm": 0.27831992506980896,
+      "learning_rate": 0.00010460526315789472,
+      "loss": 0.4981,
+      "step": 1855
+    },
+    {
+      "epoch": 2.0590764110386908,
+      "grad_norm": 0.48267292976379395,
+      "learning_rate": 0.00010448190789473684,
+      "loss": 0.5129,
+      "step": 1856
+    },
+    {
+      "epoch": 2.0601858272084317,
+      "grad_norm": 0.3351428508758545,
+      "learning_rate": 0.00010435855263157895,
+      "loss": 0.4921,
+      "step": 1857
+    },
+    {
+      "epoch": 2.061295243378172,
+      "grad_norm": 0.3631199598312378,
+      "learning_rate": 0.00010423519736842104,
+      "loss": 0.3983,
+      "step": 1858
+    },
+    {
+      "epoch": 2.062404659547913,
+      "grad_norm": 0.369219571352005,
+      "learning_rate": 0.00010411184210526315,
+      "loss": 0.4716,
+      "step": 1859
+    },
+    {
+      "epoch": 2.0635140757176536,
+      "grad_norm": 0.43210768699645996,
+      "learning_rate": 0.00010398848684210527,
+      "loss": 0.3501,
+      "step": 1860
+    },
+    {
+      "epoch": 2.0646234918873945,
+      "grad_norm": 0.41098493337631226,
+      "learning_rate": 0.00010386513157894736,
+      "loss": 0.4026,
+      "step": 1861
+    },
+    {
+      "epoch": 2.065732908057135,
+      "grad_norm": 0.36239397525787354,
+      "learning_rate": 0.00010374177631578947,
+      "loss": 0.297,
+      "step": 1862
+    },
+    {
+      "epoch": 2.0668423242268754,
+      "grad_norm": 0.41763097047805786,
+      "learning_rate": 0.00010361842105263157,
+      "loss": 0.3034,
+      "step": 1863
+    },
+    {
+      "epoch": 2.0679517403966163,
+      "grad_norm": 0.37006324529647827,
+      "learning_rate": 0.00010349506578947367,
+      "loss": 0.5594,
+      "step": 1864
+    },
+    {
+      "epoch": 2.069061156566357,
+      "grad_norm": 0.4518885612487793,
+      "learning_rate": 0.00010337171052631579,
+      "loss": 0.5642,
+      "step": 1865
+    },
+    {
+      "epoch": 2.0701705727360977,
+      "grad_norm": 0.3855383098125458,
+      "learning_rate": 0.00010324835526315789,
+      "loss": 0.2811,
+      "step": 1866
+    },
+    {
+      "epoch": 2.071279988905838,
+      "grad_norm": 0.3048069477081299,
+      "learning_rate": 0.00010312499999999999,
+      "loss": 0.3574,
+      "step": 1867
+    },
+    {
+      "epoch": 2.072389405075579,
+      "grad_norm": 0.28566887974739075,
+      "learning_rate": 0.00010300164473684209,
+      "loss": 0.2174,
+      "step": 1868
+    },
+    {
+      "epoch": 2.0734988212453196,
+      "grad_norm": 0.394229918718338,
+      "learning_rate": 0.0001028782894736842,
+      "loss": 0.3636,
+      "step": 1869
+    },
+    {
+      "epoch": 2.0746082374150605,
+      "grad_norm": 0.31521254777908325,
+      "learning_rate": 0.00010275493421052632,
+      "loss": 0.5267,
+      "step": 1870
+    },
+    {
+      "epoch": 2.075717653584801,
+      "grad_norm": 0.3841816782951355,
+      "learning_rate": 0.00010263157894736841,
+      "loss": 0.4475,
+      "step": 1871
+    },
+    {
+      "epoch": 2.0768270697545415,
+      "grad_norm": 0.3173518776893616,
+      "learning_rate": 0.00010250822368421052,
+      "loss": 0.3575,
+      "step": 1872
+    },
+    {
+      "epoch": 2.0779364859242824,
+      "grad_norm": 0.4765770435333252,
+      "learning_rate": 0.00010238486842105261,
+      "loss": 0.3616,
+      "step": 1873
+    },
+    {
+      "epoch": 2.079045902094023,
+      "grad_norm": 0.3841620683670044,
+      "learning_rate": 0.00010226151315789472,
+      "loss": 0.4769,
+      "step": 1874
+    },
+    {
+      "epoch": 2.0801553182637638,
+      "grad_norm": 0.3756863474845886,
+      "learning_rate": 0.00010213815789473684,
+      "loss": 0.3619,
+      "step": 1875
+    },
+    {
+      "epoch": 2.0812647344335042,
+      "grad_norm": 0.29783549904823303,
+      "learning_rate": 0.00010201480263157895,
+      "loss": 0.5892,
+      "step": 1876
+    },
+    {
+      "epoch": 2.082374150603245,
+      "grad_norm": 0.3997184634208679,
+      "learning_rate": 0.00010189144736842104,
+      "loss": 0.34,
+      "step": 1877
+    },
+    {
+      "epoch": 2.0834835667729856,
+      "grad_norm": 0.31789451837539673,
+      "learning_rate": 0.00010176809210526315,
+      "loss": 0.3102,
+      "step": 1878
+    },
+    {
+      "epoch": 2.0845929829427265,
+      "grad_norm": 0.3776637017726898,
+      "learning_rate": 0.00010164473684210527,
+      "loss": 0.3004,
+      "step": 1879
+    },
+    {
+      "epoch": 2.085702399112467,
+      "grad_norm": 0.3050936162471771,
+      "learning_rate": 0.00010152138157894736,
+      "loss": 0.3768,
+      "step": 1880
+    },
+    {
+      "epoch": 2.086811815282208,
+      "grad_norm": 0.33904218673706055,
+      "learning_rate": 0.00010139802631578947,
+      "loss": 0.461,
+      "step": 1881
+    },
+    {
+      "epoch": 2.0879212314519484,
+      "grad_norm": 0.5047959089279175,
+      "learning_rate": 0.00010127467105263157,
+      "loss": 0.4089,
+      "step": 1882
+    },
+    {
+      "epoch": 2.089030647621689,
+      "grad_norm": 0.3899175822734833,
+      "learning_rate": 0.00010115131578947367,
+      "loss": 0.5338,
+      "step": 1883
+    },
+    {
+      "epoch": 2.09014006379143,
+      "grad_norm": 0.5013115406036377,
+      "learning_rate": 0.00010102796052631579,
+      "loss": 0.4203,
+      "step": 1884
+    },
+    {
+      "epoch": 2.0912494799611703,
+      "grad_norm": 0.2986677289009094,
+      "learning_rate": 0.0001009046052631579,
+      "loss": 0.3183,
+      "step": 1885
+    },
+    {
+      "epoch": 2.092358896130911,
+      "grad_norm": 0.26325130462646484,
+      "learning_rate": 0.00010078124999999999,
+      "loss": 0.3204,
+      "step": 1886
+    },
+    {
+      "epoch": 2.0934683123006517,
+      "grad_norm": 0.333397775888443,
+      "learning_rate": 0.0001006578947368421,
+      "loss": 0.292,
+      "step": 1887
+    },
+    {
+      "epoch": 2.0945777284703926,
+      "grad_norm": 0.40086430311203003,
+      "learning_rate": 0.0001005345394736842,
+      "loss": 0.3311,
+      "step": 1888
+    },
+    {
+      "epoch": 2.095687144640133,
+      "grad_norm": 0.3059875965118408,
+      "learning_rate": 0.0001004111842105263,
+      "loss": 0.3572,
+      "step": 1889
+    },
+    {
+      "epoch": 2.096796560809874,
+      "grad_norm": 0.40417563915252686,
+      "learning_rate": 0.00010028782894736841,
+      "loss": 0.3889,
+      "step": 1890
+    },
+    {
+      "epoch": 2.0979059769796145,
+      "grad_norm": 0.45205605030059814,
+      "learning_rate": 0.00010016447368421052,
+      "loss": 0.5045,
+      "step": 1891
+    },
+    {
+      "epoch": 2.0990153931493554,
+      "grad_norm": 0.28313323855400085,
+      "learning_rate": 0.00010004111842105261,
+      "loss": 0.4171,
+      "step": 1892
+    },
+    {
+      "epoch": 2.100124809319096,
+      "grad_norm": 0.41001975536346436,
+      "learning_rate": 9.991776315789472e-05,
+      "loss": 0.2781,
+      "step": 1893
+    },
+    {
+      "epoch": 2.1012342254888363,
+      "grad_norm": 0.2708085775375366,
+      "learning_rate": 9.979440789473684e-05,
+      "loss": 0.3806,
+      "step": 1894
+    },
+    {
+      "epoch": 2.1023436416585772,
+      "grad_norm": 0.3741215467453003,
+      "learning_rate": 9.967105263157895e-05,
+      "loss": 0.2874,
+      "step": 1895
+    },
+    {
+      "epoch": 2.1034530578283177,
+      "grad_norm": 0.2658732831478119,
+      "learning_rate": 9.954769736842104e-05,
+      "loss": 0.3757,
+      "step": 1896
+    },
+    {
+      "epoch": 2.1045624739980586,
+      "grad_norm": 0.3492313623428345,
+      "learning_rate": 9.942434210526315e-05,
+      "loss": 0.4064,
+      "step": 1897
+    },
+    {
+      "epoch": 2.105671890167799,
+      "grad_norm": 0.34106922149658203,
+      "learning_rate": 9.930098684210527e-05,
+      "loss": 0.2377,
+      "step": 1898
+    },
+    {
+      "epoch": 2.10678130633754,
+      "grad_norm": 0.3154791593551636,
+      "learning_rate": 9.917763157894736e-05,
+      "loss": 0.4634,
+      "step": 1899
+    },
+    {
+      "epoch": 2.1078907225072805,
+      "grad_norm": 0.3917715549468994,
+      "learning_rate": 9.905427631578947e-05,
+      "loss": 0.4126,
+      "step": 1900
+    },
+    {
+      "epoch": 2.1090001386770214,
+      "grad_norm": 0.32718661427497864,
+      "learning_rate": 9.893092105263157e-05,
+      "loss": 0.3976,
+      "step": 1901
+    },
+    {
+      "epoch": 2.110109554846762,
+      "grad_norm": 0.2590242028236389,
+      "learning_rate": 9.880756578947367e-05,
+      "loss": 0.3127,
+      "step": 1902
+    },
+    {
+      "epoch": 2.1112189710165024,
+      "grad_norm": 0.44344115257263184,
+      "learning_rate": 9.868421052631579e-05,
+      "loss": 0.3864,
+      "step": 1903
+    },
+    {
+      "epoch": 2.1123283871862433,
+      "grad_norm": 0.425987184047699,
+      "learning_rate": 9.85608552631579e-05,
+      "loss": 0.4428,
+      "step": 1904
+    },
+    {
+      "epoch": 2.1134378033559837,
+      "grad_norm": 0.5364298224449158,
+      "learning_rate": 9.843749999999999e-05,
+      "loss": 0.3975,
+      "step": 1905
+    },
+    {
+      "epoch": 2.1145472195257247,
+      "grad_norm": 0.4158439338207245,
+      "learning_rate": 9.83141447368421e-05,
+      "loss": 0.499,
+      "step": 1906
+    },
+    {
+      "epoch": 2.115656635695465,
+      "grad_norm": 0.30733615159988403,
+      "learning_rate": 9.81907894736842e-05,
+      "loss": 0.3568,
+      "step": 1907
+    },
+    {
+      "epoch": 2.116766051865206,
+      "grad_norm": 0.2557796835899353,
+      "learning_rate": 9.806743421052631e-05,
+      "loss": 0.5255,
+      "step": 1908
+    },
+    {
+      "epoch": 2.1178754680349465,
+      "grad_norm": 0.36079320311546326,
+      "learning_rate": 9.794407894736841e-05,
+      "loss": 0.4101,
+      "step": 1909
+    },
+    {
+      "epoch": 2.1189848842046874,
+      "grad_norm": 0.3143673837184906,
+      "learning_rate": 9.782072368421052e-05,
+      "loss": 0.4313,
+      "step": 1910
+    },
+    {
+      "epoch": 2.120094300374428,
+      "grad_norm": 0.44542989134788513,
+      "learning_rate": 9.769736842105261e-05,
+      "loss": 0.3331,
+      "step": 1911
+    },
+    {
+      "epoch": 2.121203716544169,
+      "grad_norm": 0.28402179479599,
+      "learning_rate": 9.757401315789472e-05,
+      "loss": 0.3088,
+      "step": 1912
+    },
+    {
+      "epoch": 2.1223131327139093,
+      "grad_norm": 0.4471670687198639,
+      "learning_rate": 9.745065789473684e-05,
+      "loss": 0.4129,
+      "step": 1913
+    },
+    {
+      "epoch": 2.12342254888365,
+      "grad_norm": 0.29794514179229736,
+      "learning_rate": 9.732730263157893e-05,
+      "loss": 0.3326,
+      "step": 1914
+    },
+    {
+      "epoch": 2.1245319650533907,
+      "grad_norm": 0.37658748030662537,
+      "learning_rate": 9.720394736842104e-05,
+      "loss": 0.3271,
+      "step": 1915
+    },
+    {
+      "epoch": 2.125641381223131,
+      "grad_norm": 0.31780439615249634,
+      "learning_rate": 9.708059210526315e-05,
+      "loss": 0.3144,
+      "step": 1916
+    },
+    {
+      "epoch": 2.126750797392872,
+      "grad_norm": 0.3516882359981537,
+      "learning_rate": 9.695723684210527e-05,
+      "loss": 0.1996,
+      "step": 1917
+    },
+    {
+      "epoch": 2.1278602135626126,
+      "grad_norm": 0.3291231095790863,
+      "learning_rate": 9.683388157894736e-05,
+      "loss": 0.7135,
+      "step": 1918
+    },
+    {
+      "epoch": 2.1289696297323535,
+      "grad_norm": 0.4481741189956665,
+      "learning_rate": 9.671052631578947e-05,
+      "loss": 0.2855,
+      "step": 1919
+    },
+    {
+      "epoch": 2.130079045902094,
+      "grad_norm": 0.43044859170913696,
+      "learning_rate": 9.658717105263157e-05,
+      "loss": 0.4145,
+      "step": 1920
+    },
+    {
+      "epoch": 2.131188462071835,
+      "grad_norm": 0.3299560546875,
+      "learning_rate": 9.646381578947367e-05,
+      "loss": 0.5771,
+      "step": 1921
+    },
+    {
+      "epoch": 2.1322978782415754,
+      "grad_norm": 0.4290536940097809,
+      "learning_rate": 9.634046052631579e-05,
+      "loss": 0.492,
+      "step": 1922
+    },
+    {
+      "epoch": 2.133407294411316,
+      "grad_norm": 0.344137966632843,
+      "learning_rate": 9.62171052631579e-05,
+      "loss": 0.3803,
+      "step": 1923
+    },
+    {
+      "epoch": 2.1345167105810567,
+      "grad_norm": 0.511370062828064,
+      "learning_rate": 9.609374999999999e-05,
+      "loss": 0.3662,
+      "step": 1924
+    },
+    {
+      "epoch": 2.135626126750797,
+      "grad_norm": 0.4554339349269867,
+      "learning_rate": 9.59703947368421e-05,
+      "loss": 0.3805,
+      "step": 1925
+    },
+    {
+      "epoch": 2.136735542920538,
+      "grad_norm": 0.36867454648017883,
+      "learning_rate": 9.58470394736842e-05,
+      "loss": 0.4913,
+      "step": 1926
+    },
+    {
+      "epoch": 2.1378449590902786,
+      "grad_norm": 0.36078619956970215,
+      "learning_rate": 9.572368421052631e-05,
+      "loss": 0.3083,
+      "step": 1927
+    },
+    {
+      "epoch": 2.1389543752600195,
+      "grad_norm": 0.369831919670105,
+      "learning_rate": 9.560032894736841e-05,
+      "loss": 0.2677,
+      "step": 1928
+    },
+    {
+      "epoch": 2.14006379142976,
+      "grad_norm": 0.4087238013744354,
+      "learning_rate": 9.547697368421052e-05,
+      "loss": 0.4678,
+      "step": 1929
+    },
+    {
+      "epoch": 2.141173207599501,
+      "grad_norm": 0.5325888991355896,
+      "learning_rate": 9.535361842105261e-05,
+      "loss": 0.3679,
+      "step": 1930
+    },
+    {
+      "epoch": 2.1422826237692414,
+      "grad_norm": 0.281781405210495,
+      "learning_rate": 9.523026315789472e-05,
+      "loss": 0.4423,
+      "step": 1931
+    },
+    {
+      "epoch": 2.1433920399389823,
+      "grad_norm": 0.29980483651161194,
+      "learning_rate": 9.510690789473684e-05,
+      "loss": 0.3535,
+      "step": 1932
+    },
+    {
+      "epoch": 2.144501456108723,
+      "grad_norm": 0.39666828513145447,
+      "learning_rate": 9.498355263157893e-05,
+      "loss": 0.5298,
+      "step": 1933
+    },
+    {
+      "epoch": 2.1456108722784633,
+      "grad_norm": 0.4402129352092743,
+      "learning_rate": 9.486019736842104e-05,
+      "loss": 0.4866,
+      "step": 1934
+    },
+    {
+      "epoch": 2.146720288448204,
+      "grad_norm": 0.4045298397541046,
+      "learning_rate": 9.473684210526315e-05,
+      "loss": 0.5098,
+      "step": 1935
+    },
+    {
+      "epoch": 2.1478297046179446,
+      "grad_norm": 0.3659813404083252,
+      "learning_rate": 9.461348684210527e-05,
+      "loss": 0.3095,
+      "step": 1936
+    },
+    {
+      "epoch": 2.1489391207876856,
+      "grad_norm": 0.3063139021396637,
+      "learning_rate": 9.449013157894736e-05,
+      "loss": 0.2183,
+      "step": 1937
+    },
+    {
+      "epoch": 2.150048536957426,
+      "grad_norm": 0.4679979979991913,
+      "learning_rate": 9.436677631578947e-05,
+      "loss": 0.514,
+      "step": 1938
+    },
+    {
+      "epoch": 2.151157953127167,
+      "grad_norm": 0.31606557965278625,
+      "learning_rate": 9.424342105263156e-05,
+      "loss": 0.4071,
+      "step": 1939
+    },
+    {
+      "epoch": 2.1522673692969074,
+      "grad_norm": 0.3824010193347931,
+      "learning_rate": 9.412006578947367e-05,
+      "loss": 0.3155,
+      "step": 1940
+    },
+    {
+      "epoch": 2.1533767854666483,
+      "grad_norm": 0.28109651803970337,
+      "learning_rate": 9.399671052631579e-05,
+      "loss": 0.344,
+      "step": 1941
+    },
+    {
+      "epoch": 2.154486201636389,
+      "grad_norm": 0.3306637704372406,
+      "learning_rate": 9.38733552631579e-05,
+      "loss": 0.1947,
+      "step": 1942
+    },
+    {
+      "epoch": 2.1555956178061297,
+      "grad_norm": 0.32935014367103577,
+      "learning_rate": 9.374999999999999e-05,
+      "loss": 0.2445,
+      "step": 1943
+    },
+    {
+      "epoch": 2.15670503397587,
+      "grad_norm": 0.33338576555252075,
+      "learning_rate": 9.36266447368421e-05,
+      "loss": 0.3996,
+      "step": 1944
+    },
+    {
+      "epoch": 2.1578144501456107,
+      "grad_norm": 0.25957128405570984,
+      "learning_rate": 9.35032894736842e-05,
+      "loss": 0.3473,
+      "step": 1945
+    },
+    {
+      "epoch": 2.1589238663153516,
+      "grad_norm": 0.48169735074043274,
+      "learning_rate": 9.337993421052631e-05,
+      "loss": 0.329,
+      "step": 1946
+    },
+    {
+      "epoch": 2.160033282485092,
+      "grad_norm": 0.4141751527786255,
+      "learning_rate": 9.325657894736842e-05,
+      "loss": 0.2981,
+      "step": 1947
+    },
+    {
+      "epoch": 2.161142698654833,
+      "grad_norm": 0.4750854969024658,
+      "learning_rate": 9.313322368421052e-05,
+      "loss": 0.5787,
+      "step": 1948
+    },
+    {
+      "epoch": 2.1622521148245735,
+      "grad_norm": 0.4501727819442749,
+      "learning_rate": 9.300986842105262e-05,
+      "loss": 0.2925,
+      "step": 1949
+    },
+    {
+      "epoch": 2.1633615309943144,
+      "grad_norm": 0.5955410599708557,
+      "learning_rate": 9.288651315789472e-05,
+      "loss": 0.3785,
+      "step": 1950
+    },
+    {
+      "epoch": 2.164470947164055,
+      "grad_norm": 0.36002209782600403,
+      "learning_rate": 9.276315789473684e-05,
+      "loss": 0.4342,
+      "step": 1951
+    },
+    {
+      "epoch": 2.1655803633337958,
+      "grad_norm": 0.4657028615474701,
+      "learning_rate": 9.263980263157894e-05,
+      "loss": 0.3671,
+      "step": 1952
+    },
+    {
+      "epoch": 2.1666897795035363,
+      "grad_norm": 0.7127841711044312,
+      "learning_rate": 9.251644736842104e-05,
+      "loss": 0.3303,
+      "step": 1953
+    },
+    {
+      "epoch": 2.167799195673277,
+      "grad_norm": 0.4676034152507782,
+      "learning_rate": 9.239309210526315e-05,
+      "loss": 0.507,
+      "step": 1954
+    },
+    {
+      "epoch": 2.1689086118430176,
+      "grad_norm": 0.6187905073165894,
+      "learning_rate": 9.226973684210527e-05,
+      "loss": 0.3498,
+      "step": 1955
+    },
+    {
+      "epoch": 2.170018028012758,
+      "grad_norm": 0.4146270751953125,
+      "learning_rate": 9.214638157894736e-05,
+      "loss": 0.4633,
+      "step": 1956
+    },
+    {
+      "epoch": 2.171127444182499,
+      "grad_norm": 0.3486730456352234,
+      "learning_rate": 9.202302631578947e-05,
+      "loss": 0.4834,
+      "step": 1957
+    },
+    {
+      "epoch": 2.1722368603522395,
+      "grad_norm": 0.30160388350486755,
+      "learning_rate": 9.189967105263156e-05,
+      "loss": 0.4886,
+      "step": 1958
+    },
+    {
+      "epoch": 2.1733462765219804,
+      "grad_norm": 0.571941077709198,
+      "learning_rate": 9.177631578947367e-05,
+      "loss": 0.6495,
+      "step": 1959
+    },
+    {
+      "epoch": 2.174455692691721,
+      "grad_norm": 0.2943151295185089,
+      "learning_rate": 9.165296052631579e-05,
+      "loss": 0.4508,
+      "step": 1960
+    },
+    {
+      "epoch": 2.175565108861462,
+      "grad_norm": 0.3195703625679016,
+      "learning_rate": 9.15296052631579e-05,
+      "loss": 0.4173,
+      "step": 1961
+    },
+    {
+      "epoch": 2.1766745250312023,
+      "grad_norm": 0.3255450427532196,
+      "learning_rate": 9.140624999999999e-05,
+      "loss": 0.2631,
+      "step": 1962
+    },
+    {
+      "epoch": 2.177783941200943,
+      "grad_norm": 0.34725460410118103,
+      "learning_rate": 9.12828947368421e-05,
+      "loss": 0.2789,
+      "step": 1963
+    },
+    {
+      "epoch": 2.1788933573706837,
+      "grad_norm": 0.284347265958786,
+      "learning_rate": 9.115953947368419e-05,
+      "loss": 0.3208,
+      "step": 1964
+    },
+    {
+      "epoch": 2.180002773540424,
+      "grad_norm": 0.3235912322998047,
+      "learning_rate": 9.103618421052631e-05,
+      "loss": 0.5076,
+      "step": 1965
+    },
+    {
+      "epoch": 2.181112189710165,
+      "grad_norm": 0.37382572889328003,
+      "learning_rate": 9.091282894736842e-05,
+      "loss": 0.4009,
+      "step": 1966
+    },
+    {
+      "epoch": 2.1822216058799055,
+      "grad_norm": 0.3380100727081299,
+      "learning_rate": 9.078947368421052e-05,
+      "loss": 0.3682,
+      "step": 1967
+    },
+    {
+      "epoch": 2.1833310220496465,
+      "grad_norm": 0.36504095792770386,
+      "learning_rate": 9.066611842105262e-05,
+      "loss": 0.4735,
+      "step": 1968
+    },
+    {
+      "epoch": 2.184440438219387,
+      "grad_norm": 0.44653430581092834,
+      "learning_rate": 9.054276315789472e-05,
+      "loss": 0.8669,
+      "step": 1969
+    },
+    {
+      "epoch": 2.185549854389128,
+      "grad_norm": 0.38333860039711,
+      "learning_rate": 9.041940789473684e-05,
+      "loss": 0.2508,
+      "step": 1970
+    },
+    {
+      "epoch": 2.1866592705588683,
+      "grad_norm": 0.413216233253479,
+      "learning_rate": 9.029605263157894e-05,
+      "loss": 0.5268,
+      "step": 1971
+    },
+    {
+      "epoch": 2.1877686867286092,
+      "grad_norm": 0.4521336257457733,
+      "learning_rate": 9.017269736842104e-05,
+      "loss": 0.62,
+      "step": 1972
+    },
+    {
+      "epoch": 2.1888781028983497,
+      "grad_norm": 0.39028438925743103,
+      "learning_rate": 9.004934210526315e-05,
+      "loss": 0.4885,
+      "step": 1973
+    },
+    {
+      "epoch": 2.18998751906809,
+      "grad_norm": 0.4401836693286896,
+      "learning_rate": 8.992598684210527e-05,
+      "loss": 0.3722,
+      "step": 1974
+    },
+    {
+      "epoch": 2.191096935237831,
+      "grad_norm": 0.4004587233066559,
+      "learning_rate": 8.980263157894736e-05,
+      "loss": 0.2697,
+      "step": 1975
+    },
+    {
+      "epoch": 2.1922063514075716,
+      "grad_norm": 0.5189459919929504,
+      "learning_rate": 8.967927631578947e-05,
+      "loss": 0.4762,
+      "step": 1976
+    },
+    {
+      "epoch": 2.1933157675773125,
+      "grad_norm": 0.5282573103904724,
+      "learning_rate": 8.955592105263156e-05,
+      "loss": 0.4813,
+      "step": 1977
+    },
+    {
+      "epoch": 2.194425183747053,
+      "grad_norm": 0.3748975098133087,
+      "learning_rate": 8.943256578947367e-05,
+      "loss": 0.4131,
+      "step": 1978
+    },
+    {
+      "epoch": 2.195534599916794,
+      "grad_norm": 0.3851288855075836,
+      "learning_rate": 8.930921052631579e-05,
+      "loss": 0.2714,
+      "step": 1979
+    },
+    {
+      "epoch": 2.1966440160865344,
+      "grad_norm": 0.2807680368423462,
+      "learning_rate": 8.91858552631579e-05,
+      "loss": 0.4121,
+      "step": 1980
+    },
+    {
+      "epoch": 2.1977534322562753,
+      "grad_norm": 0.42031747102737427,
+      "learning_rate": 8.906249999999999e-05,
+      "loss": 0.3362,
+      "step": 1981
+    },
+    {
+      "epoch": 2.1988628484260158,
+      "grad_norm": 0.33740946650505066,
+      "learning_rate": 8.89391447368421e-05,
+      "loss": 0.2981,
+      "step": 1982
+    },
+    {
+      "epoch": 2.1999722645957567,
+      "grad_norm": 0.27107852697372437,
+      "learning_rate": 8.881578947368419e-05,
+      "loss": 0.3691,
+      "step": 1983
+    },
+    {
+      "epoch": 2.201081680765497,
+      "grad_norm": 0.36126452684402466,
+      "learning_rate": 8.869243421052631e-05,
+      "loss": 0.4095,
+      "step": 1984
+    },
+    {
+      "epoch": 2.2021910969352376,
+      "grad_norm": 0.517387330532074,
+      "learning_rate": 8.856907894736842e-05,
+      "loss": 0.3078,
+      "step": 1985
+    },
+    {
+      "epoch": 2.2033005131049785,
+      "grad_norm": 0.5038511157035828,
+      "learning_rate": 8.844572368421052e-05,
+      "loss": 0.2611,
+      "step": 1986
+    },
+    {
+      "epoch": 2.204409929274719,
+      "grad_norm": 0.35804232954978943,
+      "learning_rate": 8.832236842105262e-05,
+      "loss": 0.37,
+      "step": 1987
+    },
+    {
+      "epoch": 2.20551934544446,
+      "grad_norm": 0.47403684258461,
+      "learning_rate": 8.819901315789472e-05,
+      "loss": 0.4773,
+      "step": 1988
+    },
+    {
+      "epoch": 2.2066287616142004,
+      "grad_norm": 0.38205355405807495,
+      "learning_rate": 8.807565789473684e-05,
+      "loss": 0.3112,
+      "step": 1989
+    },
+    {
+      "epoch": 2.2077381777839413,
+      "grad_norm": 0.39398112893104553,
+      "learning_rate": 8.795230263157894e-05,
+      "loss": 0.4293,
+      "step": 1990
+    },
+    {
+      "epoch": 2.208847593953682,
+      "grad_norm": 0.39993610978126526,
+      "learning_rate": 8.782894736842104e-05,
+      "loss": 0.5623,
+      "step": 1991
+    },
+    {
+      "epoch": 2.2099570101234227,
+      "grad_norm": 0.4437258243560791,
+      "learning_rate": 8.770559210526315e-05,
+      "loss": 0.5582,
+      "step": 1992
+    },
+    {
+      "epoch": 2.211066426293163,
+      "grad_norm": 0.3316510021686554,
+      "learning_rate": 8.758223684210526e-05,
+      "loss": 0.1147,
+      "step": 1993
+    },
+    {
+      "epoch": 2.212175842462904,
+      "grad_norm": 0.4257866442203522,
+      "learning_rate": 8.745888157894736e-05,
+      "loss": 0.2057,
+      "step": 1994
+    },
+    {
+      "epoch": 2.2132852586326446,
+      "grad_norm": 0.41767770051956177,
+      "learning_rate": 8.733552631578947e-05,
+      "loss": 0.5289,
+      "step": 1995
+    },
+    {
+      "epoch": 2.214394674802385,
+      "grad_norm": 0.3871503472328186,
+      "learning_rate": 8.721217105263156e-05,
+      "loss": 0.4098,
+      "step": 1996
+    },
+    {
+      "epoch": 2.215504090972126,
+      "grad_norm": 0.3213239014148712,
+      "learning_rate": 8.708881578947367e-05,
+      "loss": 0.3835,
+      "step": 1997
+    },
+    {
+      "epoch": 2.2166135071418664,
+      "grad_norm": 0.5189967155456543,
+      "learning_rate": 8.696546052631579e-05,
+      "loss": 0.473,
+      "step": 1998
+    },
+    {
+      "epoch": 2.2177229233116074,
+      "grad_norm": 0.36058536171913147,
+      "learning_rate": 8.68421052631579e-05,
+      "loss": 0.3896,
+      "step": 1999
+    },
+    {
+      "epoch": 2.218832339481348,
+      "grad_norm": 0.5485463738441467,
+      "learning_rate": 8.671874999999999e-05,
+      "loss": 0.5556,
+      "step": 2000
+    },
+    {
+      "epoch": 2.2199417556510888,
+      "grad_norm": 0.3734520971775055,
+      "learning_rate": 8.65953947368421e-05,
+      "loss": 0.42,
+      "step": 2001
+    },
+    {
+      "epoch": 2.2210511718208292,
+      "grad_norm": 0.30071988701820374,
+      "learning_rate": 8.647203947368419e-05,
+      "loss": 0.4252,
+      "step": 2002
+    },
+    {
+      "epoch": 2.22216058799057,
+      "grad_norm": 0.4292794167995453,
+      "learning_rate": 8.634868421052631e-05,
+      "loss": 0.2889,
+      "step": 2003
+    },
+    {
+      "epoch": 2.2232700041603106,
+      "grad_norm": 0.31171557307243347,
+      "learning_rate": 8.622532894736842e-05,
+      "loss": 0.4064,
+      "step": 2004
+    },
+    {
+      "epoch": 2.2243794203300515,
+      "grad_norm": 0.4342403709888458,
+      "learning_rate": 8.610197368421052e-05,
+      "loss": 0.3495,
+      "step": 2005
+    },
+    {
+      "epoch": 2.225488836499792,
+      "grad_norm": 0.3274979591369629,
+      "learning_rate": 8.597861842105262e-05,
+      "loss": 0.2234,
+      "step": 2006
+    },
+    {
+      "epoch": 2.2265982526695325,
+      "grad_norm": 0.3761701285839081,
+      "learning_rate": 8.585526315789472e-05,
+      "loss": 0.3346,
+      "step": 2007
+    },
+    {
+      "epoch": 2.2277076688392734,
+      "grad_norm": 0.3312693238258362,
+      "learning_rate": 8.573190789473684e-05,
+      "loss": 0.4378,
+      "step": 2008
+    },
+    {
+      "epoch": 2.228817085009014,
+      "grad_norm": 0.4094376862049103,
+      "learning_rate": 8.560855263157894e-05,
+      "loss": 0.3139,
+      "step": 2009
+    },
+    {
+      "epoch": 2.229926501178755,
+      "grad_norm": 0.44111502170562744,
+      "learning_rate": 8.548519736842104e-05,
+      "loss": 0.5182,
+      "step": 2010
+    },
+    {
+      "epoch": 2.2310359173484953,
+      "grad_norm": 0.40903040766716003,
+      "learning_rate": 8.536184210526315e-05,
+      "loss": 0.2903,
+      "step": 2011
+    },
+    {
+      "epoch": 2.232145333518236,
+      "grad_norm": 0.38946643471717834,
+      "learning_rate": 8.523848684210526e-05,
+      "loss": 0.5439,
+      "step": 2012
+    },
+    {
+      "epoch": 2.2332547496879767,
+      "grad_norm": 0.3774378001689911,
+      "learning_rate": 8.511513157894736e-05,
+      "loss": 0.4863,
+      "step": 2013
+    },
+    {
+      "epoch": 2.2343641658577176,
+      "grad_norm": 0.2982938289642334,
+      "learning_rate": 8.499177631578947e-05,
+      "loss": 0.3545,
+      "step": 2014
+    },
+    {
+      "epoch": 2.235473582027458,
+      "grad_norm": 0.24514225125312805,
+      "learning_rate": 8.486842105263156e-05,
+      "loss": 0.2306,
+      "step": 2015
+    },
+    {
+      "epoch": 2.2365829981971985,
+      "grad_norm": 0.4611694812774658,
+      "learning_rate": 8.474506578947367e-05,
+      "loss": 0.3808,
+      "step": 2016
+    },
+    {
+      "epoch": 2.2376924143669394,
+      "grad_norm": 0.43070465326309204,
+      "learning_rate": 8.462171052631579e-05,
+      "loss": 0.2849,
+      "step": 2017
+    },
+    {
+      "epoch": 2.23880183053668,
+      "grad_norm": 0.45175713300704956,
+      "learning_rate": 8.449835526315788e-05,
+      "loss": 0.2813,
+      "step": 2018
+    },
+    {
+      "epoch": 2.239911246706421,
+      "grad_norm": 0.4600198268890381,
+      "learning_rate": 8.437499999999999e-05,
+      "loss": 0.374,
+      "step": 2019
+    },
+    {
+      "epoch": 2.2410206628761613,
+      "grad_norm": 0.8133832216262817,
+      "learning_rate": 8.42516447368421e-05,
+      "loss": 0.4012,
+      "step": 2020
+    },
+    {
+      "epoch": 2.2421300790459022,
+      "grad_norm": 0.3792482316493988,
+      "learning_rate": 8.412828947368419e-05,
+      "loss": 0.3851,
+      "step": 2021
+    },
+    {
+      "epoch": 2.2432394952156427,
+      "grad_norm": 0.38575461506843567,
+      "learning_rate": 8.400493421052631e-05,
+      "loss": 0.3673,
+      "step": 2022
+    },
+    {
+      "epoch": 2.2443489113853836,
+      "grad_norm": 0.4712158143520355,
+      "learning_rate": 8.388157894736842e-05,
+      "loss": 0.5148,
+      "step": 2023
+    },
+    {
+      "epoch": 2.245458327555124,
+      "grad_norm": 0.4250771105289459,
+      "learning_rate": 8.375822368421052e-05,
+      "loss": 0.3809,
+      "step": 2024
+    },
+    {
+      "epoch": 2.2465677437248646,
+      "grad_norm": 0.3241025507450104,
+      "learning_rate": 8.363486842105262e-05,
+      "loss": 0.4474,
+      "step": 2025
+    },
+    {
+      "epoch": 2.2476771598946055,
+      "grad_norm": 0.2825429141521454,
+      "learning_rate": 8.351151315789472e-05,
+      "loss": 0.3693,
+      "step": 2026
+    },
+    {
+      "epoch": 2.248786576064346,
+      "grad_norm": 0.3415563106536865,
+      "learning_rate": 8.338815789473684e-05,
+      "loss": 0.3112,
+      "step": 2027
+    },
+    {
+      "epoch": 2.249895992234087,
+      "grad_norm": 0.4205566346645355,
+      "learning_rate": 8.326480263157894e-05,
+      "loss": 0.4012,
+      "step": 2028
+    },
+    {
+      "epoch": 2.2510054084038273,
+      "grad_norm": 0.39186304807662964,
+      "learning_rate": 8.314144736842104e-05,
+      "loss": 0.406,
+      "step": 2029
+    },
+    {
+      "epoch": 2.2521148245735683,
+      "grad_norm": 0.44930022954940796,
+      "learning_rate": 8.301809210526315e-05,
+      "loss": 0.3568,
+      "step": 2030
+    },
+    {
+      "epoch": 2.2532242407433087,
+      "grad_norm": 0.37317511439323425,
+      "learning_rate": 8.289473684210526e-05,
+      "loss": 0.3248,
+      "step": 2031
+    },
+    {
+      "epoch": 2.2543336569130497,
+      "grad_norm": 0.4311521053314209,
+      "learning_rate": 8.277138157894736e-05,
+      "loss": 0.4463,
+      "step": 2032
+    },
+    {
+      "epoch": 2.25544307308279,
+      "grad_norm": 0.29921552538871765,
+      "learning_rate": 8.264802631578947e-05,
+      "loss": 0.2457,
+      "step": 2033
+    },
+    {
+      "epoch": 2.256552489252531,
+      "grad_norm": 0.3454459011554718,
+      "learning_rate": 8.252467105263156e-05,
+      "loss": 0.505,
+      "step": 2034
+    },
+    {
+      "epoch": 2.2576619054222715,
+      "grad_norm": 0.38671955466270447,
+      "learning_rate": 8.240131578947367e-05,
+      "loss": 0.3366,
+      "step": 2035
+    },
+    {
+      "epoch": 2.258771321592012,
+      "grad_norm": 0.3811526596546173,
+      "learning_rate": 8.227796052631579e-05,
+      "loss": 0.4974,
+      "step": 2036
+    },
+    {
+      "epoch": 2.259880737761753,
+      "grad_norm": 0.3260783851146698,
+      "learning_rate": 8.215460526315788e-05,
+      "loss": 0.2986,
+      "step": 2037
+    },
+    {
+      "epoch": 2.2609901539314934,
+      "grad_norm": 0.3846670091152191,
+      "learning_rate": 8.203124999999999e-05,
+      "loss": 0.3801,
+      "step": 2038
+    },
+    {
+      "epoch": 2.2620995701012343,
+      "grad_norm": 0.38990718126296997,
+      "learning_rate": 8.19078947368421e-05,
+      "loss": 0.4151,
+      "step": 2039
+    },
+    {
+      "epoch": 2.2632089862709748,
+      "grad_norm": 0.36695122718811035,
+      "learning_rate": 8.178453947368419e-05,
+      "loss": 0.4518,
+      "step": 2040
+    },
+    {
+      "epoch": 2.2643184024407157,
+      "grad_norm": 0.33559417724609375,
+      "learning_rate": 8.166118421052631e-05,
+      "loss": 0.3286,
+      "step": 2041
+    },
+    {
+      "epoch": 2.265427818610456,
+      "grad_norm": 0.3156730830669403,
+      "learning_rate": 8.153782894736842e-05,
+      "loss": 0.3759,
+      "step": 2042
+    },
+    {
+      "epoch": 2.266537234780197,
+      "grad_norm": 0.4046294391155243,
+      "learning_rate": 8.141447368421051e-05,
+      "loss": 0.3196,
+      "step": 2043
+    },
+    {
+      "epoch": 2.2676466509499376,
+      "grad_norm": 0.24606218934059143,
+      "learning_rate": 8.129111842105262e-05,
+      "loss": 0.2994,
+      "step": 2044
+    },
+    {
+      "epoch": 2.2687560671196785,
+      "grad_norm": 0.2781525254249573,
+      "learning_rate": 8.116776315789473e-05,
+      "loss": 0.4763,
+      "step": 2045
+    },
+    {
+      "epoch": 2.269865483289419,
+      "grad_norm": 0.3566399812698364,
+      "learning_rate": 8.104440789473685e-05,
+      "loss": 0.2819,
+      "step": 2046
+    },
+    {
+      "epoch": 2.2709748994591594,
+      "grad_norm": 0.36124187707901,
+      "learning_rate": 8.092105263157894e-05,
+      "loss": 0.4515,
+      "step": 2047
+    },
+    {
+      "epoch": 2.2720843156289003,
+      "grad_norm": 0.42289498448371887,
+      "learning_rate": 8.079769736842105e-05,
+      "loss": 0.2499,
+      "step": 2048
+    },
+    {
+      "epoch": 2.273193731798641,
+      "grad_norm": 0.38013771176338196,
+      "learning_rate": 8.067434210526315e-05,
+      "loss": 0.2947,
+      "step": 2049
+    },
+    {
+      "epoch": 2.2743031479683817,
+      "grad_norm": 0.44941648840904236,
+      "learning_rate": 8.055098684210526e-05,
+      "loss": 0.5397,
+      "step": 2050
+    },
+    {
+      "epoch": 2.275412564138122,
+      "grad_norm": 0.4005190432071686,
+      "learning_rate": 8.042763157894737e-05,
+      "loss": 0.3519,
+      "step": 2051
+    },
+    {
+      "epoch": 2.276521980307863,
+      "grad_norm": 0.4730212986469269,
+      "learning_rate": 8.030427631578947e-05,
+      "loss": 0.3143,
+      "step": 2052
+    },
+    {
+      "epoch": 2.2776313964776036,
+      "grad_norm": 0.249730184674263,
+      "learning_rate": 8.018092105263157e-05,
+      "loss": 0.288,
+      "step": 2053
+    },
+    {
+      "epoch": 2.2787408126473445,
+      "grad_norm": 0.4012918174266815,
+      "learning_rate": 8.005756578947367e-05,
+      "loss": 0.3855,
+      "step": 2054
+    },
+    {
+      "epoch": 2.279850228817085,
+      "grad_norm": 0.40905871987342834,
+      "learning_rate": 7.993421052631579e-05,
+      "loss": 0.3615,
+      "step": 2055
+    },
+    {
+      "epoch": 2.280959644986826,
+      "grad_norm": 0.34130680561065674,
+      "learning_rate": 7.981085526315789e-05,
+      "loss": 0.4656,
+      "step": 2056
+    },
+    {
+      "epoch": 2.2820690611565664,
+      "grad_norm": 0.5201243162155151,
+      "learning_rate": 7.968749999999999e-05,
+      "loss": 0.5578,
+      "step": 2057
+    },
+    {
+      "epoch": 2.283178477326307,
+      "grad_norm": 0.3649352490901947,
+      "learning_rate": 7.95641447368421e-05,
+      "loss": 0.5485,
+      "step": 2058
+    },
+    {
+      "epoch": 2.2842878934960478,
+      "grad_norm": 0.3599608242511749,
+      "learning_rate": 7.944078947368419e-05,
+      "loss": 0.384,
+      "step": 2059
+    },
+    {
+      "epoch": 2.2853973096657882,
+      "grad_norm": 0.35313576459884644,
+      "learning_rate": 7.931743421052631e-05,
+      "loss": 0.3998,
+      "step": 2060
+    },
+    {
+      "epoch": 2.286506725835529,
+      "grad_norm": 0.3734196424484253,
+      "learning_rate": 7.919407894736842e-05,
+      "loss": 0.3514,
+      "step": 2061
+    },
+    {
+      "epoch": 2.2876161420052696,
+      "grad_norm": 0.4057319462299347,
+      "learning_rate": 7.907072368421051e-05,
+      "loss": 0.5574,
+      "step": 2062
+    },
+    {
+      "epoch": 2.2887255581750106,
+      "grad_norm": 0.3745683431625366,
+      "learning_rate": 7.894736842105262e-05,
+      "loss": 0.4718,
+      "step": 2063
+    },
+    {
+      "epoch": 2.289834974344751,
+      "grad_norm": 0.2819893956184387,
+      "learning_rate": 7.882401315789473e-05,
+      "loss": 0.3068,
+      "step": 2064
+    },
+    {
+      "epoch": 2.2909443905144915,
+      "grad_norm": 0.35861438512802124,
+      "learning_rate": 7.870065789473685e-05,
+      "loss": 0.5004,
+      "step": 2065
+    },
+    {
+      "epoch": 2.2920538066842324,
+      "grad_norm": 0.3276369273662567,
+      "learning_rate": 7.857730263157894e-05,
+      "loss": 0.4357,
+      "step": 2066
+    },
+    {
+      "epoch": 2.2931632228539733,
+      "grad_norm": 0.6049783229827881,
+      "learning_rate": 7.845394736842105e-05,
+      "loss": 0.2697,
+      "step": 2067
+    },
+    {
+      "epoch": 2.294272639023714,
+      "grad_norm": 0.3910931348800659,
+      "learning_rate": 7.833059210526314e-05,
+      "loss": 0.4302,
+      "step": 2068
+    },
+    {
+      "epoch": 2.2953820551934543,
+      "grad_norm": 0.5576995611190796,
+      "learning_rate": 7.820723684210526e-05,
+      "loss": 0.3253,
+      "step": 2069
+    },
+    {
+      "epoch": 2.296491471363195,
+      "grad_norm": 0.43437716364860535,
+      "learning_rate": 7.808388157894737e-05,
+      "loss": 0.5236,
+      "step": 2070
+    },
+    {
+      "epoch": 2.2976008875329357,
+      "grad_norm": 0.3469353914260864,
+      "learning_rate": 7.796052631578947e-05,
+      "loss": 0.391,
+      "step": 2071
+    },
+    {
+      "epoch": 2.2987103037026766,
+      "grad_norm": 0.5261640548706055,
+      "learning_rate": 7.783717105263157e-05,
+      "loss": 0.3683,
+      "step": 2072
+    },
+    {
+      "epoch": 2.299819719872417,
+      "grad_norm": 0.4726692736148834,
+      "learning_rate": 7.771381578947367e-05,
+      "loss": 0.4087,
+      "step": 2073
+    },
+    {
+      "epoch": 2.300929136042158,
+      "grad_norm": 0.3760102391242981,
+      "learning_rate": 7.759046052631579e-05,
+      "loss": 0.424,
+      "step": 2074
+    },
+    {
+      "epoch": 2.3020385522118985,
+      "grad_norm": 0.38098374009132385,
+      "learning_rate": 7.746710526315789e-05,
+      "loss": 0.3709,
+      "step": 2075
+    },
+    {
+      "epoch": 2.303147968381639,
+      "grad_norm": 0.42381149530410767,
+      "learning_rate": 7.734374999999999e-05,
+      "loss": 0.5073,
+      "step": 2076
+    },
+    {
+      "epoch": 2.30425738455138,
+      "grad_norm": 0.27536872029304504,
+      "learning_rate": 7.72203947368421e-05,
+      "loss": 0.339,
+      "step": 2077
+    },
+    {
+      "epoch": 2.3053668007211203,
+      "grad_norm": 0.7497661113739014,
+      "learning_rate": 7.709703947368419e-05,
+      "loss": 0.4369,
+      "step": 2078
+    },
+    {
+      "epoch": 2.3064762168908612,
+      "grad_norm": 0.3368166387081146,
+      "learning_rate": 7.697368421052631e-05,
+      "loss": 0.4125,
+      "step": 2079
+    },
+    {
+      "epoch": 2.3075856330606017,
+      "grad_norm": 0.43875083327293396,
+      "learning_rate": 7.685032894736842e-05,
+      "loss": 0.2639,
+      "step": 2080
+    },
+    {
+      "epoch": 2.3086950492303426,
+      "grad_norm": 0.44184234738349915,
+      "learning_rate": 7.672697368421051e-05,
+      "loss": 0.3158,
+      "step": 2081
+    },
+    {
+      "epoch": 2.309804465400083,
+      "grad_norm": 0.3724750280380249,
+      "learning_rate": 7.660361842105262e-05,
+      "loss": 0.4184,
+      "step": 2082
+    },
+    {
+      "epoch": 2.310913881569824,
+      "grad_norm": 0.37243518233299255,
+      "learning_rate": 7.648026315789473e-05,
+      "loss": 0.3311,
+      "step": 2083
+    },
+    {
+      "epoch": 2.3120232977395645,
+      "grad_norm": 0.31451210379600525,
+      "learning_rate": 7.635690789473685e-05,
+      "loss": 0.4138,
+      "step": 2084
+    },
+    {
+      "epoch": 2.3131327139093054,
+      "grad_norm": 0.5280860066413879,
+      "learning_rate": 7.623355263157894e-05,
+      "loss": 0.3653,
+      "step": 2085
+    },
+    {
+      "epoch": 2.314242130079046,
+      "grad_norm": 0.4472182095050812,
+      "learning_rate": 7.611019736842105e-05,
+      "loss": 0.3394,
+      "step": 2086
+    },
+    {
+      "epoch": 2.3153515462487864,
+      "grad_norm": 0.39410504698753357,
+      "learning_rate": 7.598684210526314e-05,
+      "loss": 0.4175,
+      "step": 2087
+    },
+    {
+      "epoch": 2.3164609624185273,
+      "grad_norm": 0.30498364567756653,
+      "learning_rate": 7.586348684210526e-05,
+      "loss": 0.2771,
+      "step": 2088
+    },
+    {
+      "epoch": 2.3175703785882678,
+      "grad_norm": 0.3658243417739868,
+      "learning_rate": 7.574013157894737e-05,
+      "loss": 0.211,
+      "step": 2089
+    },
+    {
+      "epoch": 2.3186797947580087,
+      "grad_norm": 0.27675336599349976,
+      "learning_rate": 7.561677631578947e-05,
+      "loss": 0.4383,
+      "step": 2090
+    },
+    {
+      "epoch": 2.319789210927749,
+      "grad_norm": 0.47002074122428894,
+      "learning_rate": 7.549342105263157e-05,
+      "loss": 0.2742,
+      "step": 2091
+    },
+    {
+      "epoch": 2.32089862709749,
+      "grad_norm": 0.39124634861946106,
+      "learning_rate": 7.537006578947367e-05,
+      "loss": 0.3856,
+      "step": 2092
+    },
+    {
+      "epoch": 2.3220080432672305,
+      "grad_norm": 0.467118501663208,
+      "learning_rate": 7.52467105263158e-05,
+      "loss": 0.4138,
+      "step": 2093
+    },
+    {
+      "epoch": 2.3231174594369715,
+      "grad_norm": 0.29437050223350525,
+      "learning_rate": 7.512335526315789e-05,
+      "loss": 0.2429,
+      "step": 2094
+    },
+    {
+      "epoch": 2.324226875606712,
+      "grad_norm": 0.29561078548431396,
+      "learning_rate": 7.5e-05,
+      "loss": 0.3395,
+      "step": 2095
+    },
+    {
+      "epoch": 2.325336291776453,
+      "grad_norm": 0.5432490706443787,
+      "learning_rate": 7.48766447368421e-05,
+      "loss": 0.6463,
+      "step": 2096
+    },
+    {
+      "epoch": 2.3264457079461933,
+      "grad_norm": 0.32837289571762085,
+      "learning_rate": 7.475328947368421e-05,
+      "loss": 0.4072,
+      "step": 2097
+    },
+    {
+      "epoch": 2.327555124115934,
+      "grad_norm": 0.3328750729560852,
+      "learning_rate": 7.462993421052631e-05,
+      "loss": 0.6708,
+      "step": 2098
+    },
+    {
+      "epoch": 2.3286645402856747,
+      "grad_norm": 0.44809266924858093,
+      "learning_rate": 7.450657894736842e-05,
+      "loss": 0.2939,
+      "step": 2099
+    },
+    {
+      "epoch": 2.329773956455415,
+      "grad_norm": 0.4526784121990204,
+      "learning_rate": 7.438322368421051e-05,
+      "loss": 0.4913,
+      "step": 2100
+    },
+    {
+      "epoch": 2.330883372625156,
+      "grad_norm": 0.3075268864631653,
+      "learning_rate": 7.425986842105263e-05,
+      "loss": 0.3334,
+      "step": 2101
+    },
+    {
+      "epoch": 2.3319927887948966,
+      "grad_norm": 0.33832165598869324,
+      "learning_rate": 7.413651315789473e-05,
+      "loss": 0.3785,
+      "step": 2102
+    },
+    {
+      "epoch": 2.3331022049646375,
+      "grad_norm": 0.3870348632335663,
+      "learning_rate": 7.401315789473683e-05,
+      "loss": 0.4643,
+      "step": 2103
+    },
+    {
+      "epoch": 2.334211621134378,
+      "grad_norm": 0.39963188767433167,
+      "learning_rate": 7.388980263157894e-05,
+      "loss": 0.296,
+      "step": 2104
+    },
+    {
+      "epoch": 2.335321037304119,
+      "grad_norm": 0.40964365005493164,
+      "learning_rate": 7.376644736842105e-05,
+      "loss": 0.3673,
+      "step": 2105
+    },
+    {
+      "epoch": 2.3364304534738594,
+      "grad_norm": 0.31597647070884705,
+      "learning_rate": 7.364309210526315e-05,
+      "loss": 0.3001,
+      "step": 2106
+    },
+    {
+      "epoch": 2.3375398696436003,
+      "grad_norm": 0.364797443151474,
+      "learning_rate": 7.351973684210526e-05,
+      "loss": 0.3574,
+      "step": 2107
+    },
+    {
+      "epoch": 2.3386492858133408,
+      "grad_norm": 0.4309171438217163,
+      "learning_rate": 7.339638157894735e-05,
+      "loss": 0.5573,
+      "step": 2108
+    },
+    {
+      "epoch": 2.3397587019830812,
+      "grad_norm": 0.34616556763648987,
+      "learning_rate": 7.327302631578947e-05,
+      "loss": 0.4426,
+      "step": 2109
+    },
+    {
+      "epoch": 2.340868118152822,
+      "grad_norm": 0.6331126689910889,
+      "learning_rate": 7.314967105263157e-05,
+      "loss": 0.7939,
+      "step": 2110
+    },
+    {
+      "epoch": 2.3419775343225626,
+      "grad_norm": 0.43343913555145264,
+      "learning_rate": 7.302631578947367e-05,
+      "loss": 0.3552,
+      "step": 2111
+    },
+    {
+      "epoch": 2.3430869504923035,
+      "grad_norm": 0.42529061436653137,
+      "learning_rate": 7.290296052631578e-05,
+      "loss": 0.5603,
+      "step": 2112
+    },
+    {
+      "epoch": 2.344196366662044,
+      "grad_norm": 0.32142138481140137,
+      "learning_rate": 7.277960526315789e-05,
+      "loss": 0.2149,
+      "step": 2113
+    },
+    {
+      "epoch": 2.345305782831785,
+      "grad_norm": 0.29198598861694336,
+      "learning_rate": 7.265625e-05,
+      "loss": 0.3719,
+      "step": 2114
+    },
+    {
+      "epoch": 2.3464151990015254,
+      "grad_norm": 0.36399996280670166,
+      "learning_rate": 7.25328947368421e-05,
+      "loss": 0.369,
+      "step": 2115
+    },
+    {
+      "epoch": 2.347524615171266,
+      "grad_norm": 0.288010835647583,
+      "learning_rate": 7.240953947368421e-05,
+      "loss": 0.4388,
+      "step": 2116
+    },
+    {
+      "epoch": 2.348634031341007,
+      "grad_norm": 0.45767742395401,
+      "learning_rate": 7.22861842105263e-05,
+      "loss": 0.5151,
+      "step": 2117
+    },
+    {
+      "epoch": 2.3497434475107477,
+      "grad_norm": 0.4015630781650543,
+      "learning_rate": 7.216282894736842e-05,
+      "loss": 0.2633,
+      "step": 2118
+    },
+    {
+      "epoch": 2.350852863680488,
+      "grad_norm": 0.4112047553062439,
+      "learning_rate": 7.203947368421051e-05,
+      "loss": 0.4331,
+      "step": 2119
+    },
+    {
+      "epoch": 2.3519622798502287,
+      "grad_norm": 0.4492191672325134,
+      "learning_rate": 7.191611842105263e-05,
+      "loss": 0.3771,
+      "step": 2120
+    },
+    {
+      "epoch": 2.3530716960199696,
+      "grad_norm": 0.36163103580474854,
+      "learning_rate": 7.179276315789473e-05,
+      "loss": 0.4598,
+      "step": 2121
+    },
+    {
+      "epoch": 2.35418111218971,
+      "grad_norm": 0.3377210795879364,
+      "learning_rate": 7.166940789473683e-05,
+      "loss": 0.1919,
+      "step": 2122
+    },
+    {
+      "epoch": 2.355290528359451,
+      "grad_norm": 0.6023211479187012,
+      "learning_rate": 7.154605263157894e-05,
+      "loss": 0.4939,
+      "step": 2123
+    },
+    {
+      "epoch": 2.3563999445291914,
+      "grad_norm": 0.31167641282081604,
+      "learning_rate": 7.142269736842105e-05,
+      "loss": 0.4228,
+      "step": 2124
+    },
+    {
+      "epoch": 2.3575093606989324,
+      "grad_norm": 0.39771780371665955,
+      "learning_rate": 7.129934210526315e-05,
+      "loss": 0.4054,
+      "step": 2125
+    },
+    {
+      "epoch": 2.358618776868673,
+      "grad_norm": 0.45539480447769165,
+      "learning_rate": 7.117598684210526e-05,
+      "loss": 0.4532,
+      "step": 2126
+    },
+    {
+      "epoch": 2.3597281930384133,
+      "grad_norm": 0.3616165220737457,
+      "learning_rate": 7.105263157894735e-05,
+      "loss": 0.4083,
+      "step": 2127
+    },
+    {
+      "epoch": 2.360837609208154,
+      "grad_norm": 0.3542384207248688,
+      "learning_rate": 7.092927631578947e-05,
+      "loss": 0.4871,
+      "step": 2128
+    },
+    {
+      "epoch": 2.3619470253778947,
+      "grad_norm": 0.3284272253513336,
+      "learning_rate": 7.080592105263157e-05,
+      "loss": 0.2563,
+      "step": 2129
+    },
+    {
+      "epoch": 2.3630564415476356,
+      "grad_norm": 0.48142263293266296,
+      "learning_rate": 7.068256578947367e-05,
+      "loss": 0.3424,
+      "step": 2130
+    },
+    {
+      "epoch": 2.364165857717376,
+      "grad_norm": 0.39508047699928284,
+      "learning_rate": 7.055921052631578e-05,
+      "loss": 0.4946,
+      "step": 2131
+    },
+    {
+      "epoch": 2.365275273887117,
+      "grad_norm": 0.31679755449295044,
+      "learning_rate": 7.043585526315789e-05,
+      "loss": 0.494,
+      "step": 2132
+    },
+    {
+      "epoch": 2.3663846900568575,
+      "grad_norm": 0.43126723170280457,
+      "learning_rate": 7.03125e-05,
+      "loss": 0.3824,
+      "step": 2133
+    },
+    {
+      "epoch": 2.3674941062265984,
+      "grad_norm": 0.43794259428977966,
+      "learning_rate": 7.01891447368421e-05,
+      "loss": 0.3238,
+      "step": 2134
+    },
+    {
+      "epoch": 2.368603522396339,
+      "grad_norm": 0.3279634714126587,
+      "learning_rate": 7.006578947368421e-05,
+      "loss": 0.3239,
+      "step": 2135
+    },
+    {
+      "epoch": 2.36971293856608,
+      "grad_norm": 0.32456138730049133,
+      "learning_rate": 6.99424342105263e-05,
+      "loss": 0.316,
+      "step": 2136
+    },
+    {
+      "epoch": 2.3708223547358203,
+      "grad_norm": 0.4760146141052246,
+      "learning_rate": 6.981907894736842e-05,
+      "loss": 0.3338,
+      "step": 2137
+    },
+    {
+      "epoch": 2.3719317709055607,
+      "grad_norm": 0.5461307168006897,
+      "learning_rate": 6.969572368421051e-05,
+      "loss": 0.2907,
+      "step": 2138
+    },
+    {
+      "epoch": 2.3730411870753017,
+      "grad_norm": 0.3460582494735718,
+      "learning_rate": 6.957236842105264e-05,
+      "loss": 0.3996,
+      "step": 2139
+    },
+    {
+      "epoch": 2.374150603245042,
+      "grad_norm": 0.41096046566963196,
+      "learning_rate": 6.944901315789473e-05,
+      "loss": 0.2277,
+      "step": 2140
+    },
+    {
+      "epoch": 2.375260019414783,
+      "grad_norm": 0.48936372995376587,
+      "learning_rate": 6.932565789473683e-05,
+      "loss": 0.3666,
+      "step": 2141
+    },
+    {
+      "epoch": 2.3763694355845235,
+      "grad_norm": 0.36358651518821716,
+      "learning_rate": 6.920230263157894e-05,
+      "loss": 0.3881,
+      "step": 2142
+    },
+    {
+      "epoch": 2.3774788517542644,
+      "grad_norm": 0.45791712403297424,
+      "learning_rate": 6.907894736842105e-05,
+      "loss": 0.3655,
+      "step": 2143
+    },
+    {
+      "epoch": 2.378588267924005,
+      "grad_norm": 0.3084203898906708,
+      "learning_rate": 6.895559210526316e-05,
+      "loss": 0.3494,
+      "step": 2144
+    },
+    {
+      "epoch": 2.379697684093746,
+      "grad_norm": 0.5072697997093201,
+      "learning_rate": 6.883223684210526e-05,
+      "loss": 0.4923,
+      "step": 2145
+    },
+    {
+      "epoch": 2.3808071002634863,
+      "grad_norm": 0.5244908928871155,
+      "learning_rate": 6.870888157894735e-05,
+      "loss": 0.4565,
+      "step": 2146
+    },
+    {
+      "epoch": 2.381916516433227,
+      "grad_norm": 0.3606557548046112,
+      "learning_rate": 6.858552631578948e-05,
+      "loss": 0.2892,
+      "step": 2147
+    },
+    {
+      "epoch": 2.3830259326029677,
+      "grad_norm": 0.5143904089927673,
+      "learning_rate": 6.846217105263157e-05,
+      "loss": 0.3186,
+      "step": 2148
+    },
+    {
+      "epoch": 2.384135348772708,
+      "grad_norm": 0.42753830552101135,
+      "learning_rate": 6.833881578947368e-05,
+      "loss": 0.4308,
+      "step": 2149
+    },
+    {
+      "epoch": 2.385244764942449,
+      "grad_norm": 0.32439666986465454,
+      "learning_rate": 6.821546052631578e-05,
+      "loss": 0.3906,
+      "step": 2150
+    },
+    {
+      "epoch": 2.3863541811121896,
+      "grad_norm": 0.4810985028743744,
+      "learning_rate": 6.809210526315789e-05,
+      "loss": 0.3133,
+      "step": 2151
+    },
+    {
+      "epoch": 2.3874635972819305,
+      "grad_norm": 0.4014139473438263,
+      "learning_rate": 6.796875e-05,
+      "loss": 0.3086,
+      "step": 2152
+    },
+    {
+      "epoch": 2.388573013451671,
+      "grad_norm": 0.3334631621837616,
+      "learning_rate": 6.78453947368421e-05,
+      "loss": 0.3678,
+      "step": 2153
+    },
+    {
+      "epoch": 2.389682429621412,
+      "grad_norm": 0.6455233097076416,
+      "learning_rate": 6.772203947368421e-05,
+      "loss": 0.3285,
+      "step": 2154
+    },
+    {
+      "epoch": 2.3907918457911523,
+      "grad_norm": 0.3901275396347046,
+      "learning_rate": 6.75986842105263e-05,
+      "loss": 0.3597,
+      "step": 2155
+    },
+    {
+      "epoch": 2.3919012619608933,
+      "grad_norm": 0.4130385220050812,
+      "learning_rate": 6.747532894736842e-05,
+      "loss": 0.3987,
+      "step": 2156
+    },
+    {
+      "epoch": 2.3930106781306337,
+      "grad_norm": 0.4633021652698517,
+      "learning_rate": 6.735197368421052e-05,
+      "loss": 0.3555,
+      "step": 2157
+    },
+    {
+      "epoch": 2.3941200943003746,
+      "grad_norm": 0.3566206097602844,
+      "learning_rate": 6.722861842105264e-05,
+      "loss": 0.5682,
+      "step": 2158
+    },
+    {
+      "epoch": 2.395229510470115,
+      "grad_norm": 0.37410175800323486,
+      "learning_rate": 6.710526315789473e-05,
+      "loss": 0.5559,
+      "step": 2159
+    },
+    {
+      "epoch": 2.3963389266398556,
+      "grad_norm": 0.47398602962493896,
+      "learning_rate": 6.698190789473684e-05,
+      "loss": 0.3974,
+      "step": 2160
+    },
+    {
+      "epoch": 2.3974483428095965,
+      "grad_norm": 0.3041347563266754,
+      "learning_rate": 6.685855263157894e-05,
+      "loss": 0.306,
+      "step": 2161
+    },
+    {
+      "epoch": 2.398557758979337,
+      "grad_norm": 0.6512690782546997,
+      "learning_rate": 6.673519736842105e-05,
+      "loss": 0.3073,
+      "step": 2162
+    },
+    {
+      "epoch": 2.399667175149078,
+      "grad_norm": 0.3038191497325897,
+      "learning_rate": 6.661184210526316e-05,
+      "loss": 0.3592,
+      "step": 2163
+    },
+    {
+      "epoch": 2.4007765913188184,
+      "grad_norm": 0.30106794834136963,
+      "learning_rate": 6.648848684210526e-05,
+      "loss": 0.3012,
+      "step": 2164
+    },
+    {
+      "epoch": 2.4018860074885593,
+      "grad_norm": 0.33144044876098633,
+      "learning_rate": 6.636513157894736e-05,
+      "loss": 0.4192,
+      "step": 2165
+    },
+    {
+      "epoch": 2.4029954236582998,
+      "grad_norm": 0.46323060989379883,
+      "learning_rate": 6.624177631578946e-05,
+      "loss": 0.2883,
+      "step": 2166
+    },
+    {
+      "epoch": 2.4041048398280407,
+      "grad_norm": 0.39412403106689453,
+      "learning_rate": 6.611842105263157e-05,
+      "loss": 0.3982,
+      "step": 2167
+    },
+    {
+      "epoch": 2.405214255997781,
+      "grad_norm": 0.4328696131706238,
+      "learning_rate": 6.599506578947368e-05,
+      "loss": 0.2954,
+      "step": 2168
+    },
+    {
+      "epoch": 2.406323672167522,
+      "grad_norm": 0.4066632390022278,
+      "learning_rate": 6.587171052631578e-05,
+      "loss": 0.5436,
+      "step": 2169
+    },
+    {
+      "epoch": 2.4074330883372626,
+      "grad_norm": 0.41099557280540466,
+      "learning_rate": 6.574835526315789e-05,
+      "loss": 0.3294,
+      "step": 2170
+    },
+    {
+      "epoch": 2.408542504507003,
+      "grad_norm": 0.44337305426597595,
+      "learning_rate": 6.5625e-05,
+      "loss": 0.4187,
+      "step": 2171
+    },
+    {
+      "epoch": 2.409651920676744,
+      "grad_norm": 0.43825507164001465,
+      "learning_rate": 6.55016447368421e-05,
+      "loss": 0.4014,
+      "step": 2172
+    },
+    {
+      "epoch": 2.4107613368464844,
+      "grad_norm": 0.37434348464012146,
+      "learning_rate": 6.537828947368421e-05,
+      "loss": 0.3387,
+      "step": 2173
+    },
+    {
+      "epoch": 2.4118707530162253,
+      "grad_norm": 0.49352914094924927,
+      "learning_rate": 6.52549342105263e-05,
+      "loss": 0.538,
+      "step": 2174
+    },
+    {
+      "epoch": 2.412980169185966,
+      "grad_norm": 0.3878787159919739,
+      "learning_rate": 6.513157894736842e-05,
+      "loss": 0.355,
+      "step": 2175
+    },
+    {
+      "epoch": 2.4140895853557067,
+      "grad_norm": 0.4009630084037781,
+      "learning_rate": 6.500822368421052e-05,
+      "loss": 0.2911,
+      "step": 2176
+    },
+    {
+      "epoch": 2.415199001525447,
+      "grad_norm": 0.5621581077575684,
+      "learning_rate": 6.488486842105264e-05,
+      "loss": 0.3631,
+      "step": 2177
+    },
+    {
+      "epoch": 2.4163084176951877,
+      "grad_norm": 0.5268844366073608,
+      "learning_rate": 6.476151315789473e-05,
+      "loss": 0.1947,
+      "step": 2178
+    },
+    {
+      "epoch": 2.4174178338649286,
+      "grad_norm": 0.4071340560913086,
+      "learning_rate": 6.463815789473684e-05,
+      "loss": 0.4349,
+      "step": 2179
+    },
+    {
+      "epoch": 2.418527250034669,
+      "grad_norm": 0.5059108734130859,
+      "learning_rate": 6.451480263157894e-05,
+      "loss": 0.4169,
+      "step": 2180
+    },
+    {
+      "epoch": 2.41963666620441,
+      "grad_norm": 0.3529709279537201,
+      "learning_rate": 6.439144736842105e-05,
+      "loss": 0.3216,
+      "step": 2181
+    },
+    {
+      "epoch": 2.4207460823741505,
+      "grad_norm": 0.3386879563331604,
+      "learning_rate": 6.426809210526316e-05,
+      "loss": 0.3775,
+      "step": 2182
+    },
+    {
+      "epoch": 2.4218554985438914,
+      "grad_norm": 0.4721252918243408,
+      "learning_rate": 6.414473684210526e-05,
+      "loss": 0.3298,
+      "step": 2183
+    },
+    {
+      "epoch": 2.422964914713632,
+      "grad_norm": 0.58592289686203,
+      "learning_rate": 6.402138157894736e-05,
+      "loss": 0.3353,
+      "step": 2184
+    },
+    {
+      "epoch": 2.4240743308833728,
+      "grad_norm": 0.3879697918891907,
+      "learning_rate": 6.389802631578946e-05,
+      "loss": 0.3383,
+      "step": 2185
+    },
+    {
+      "epoch": 2.4251837470531132,
+      "grad_norm": 0.5064356327056885,
+      "learning_rate": 6.377467105263157e-05,
+      "loss": 0.2966,
+      "step": 2186
+    },
+    {
+      "epoch": 2.426293163222854,
+      "grad_norm": 0.4402575194835663,
+      "learning_rate": 6.365131578947368e-05,
+      "loss": 0.3114,
+      "step": 2187
+    },
+    {
+      "epoch": 2.4274025793925946,
+      "grad_norm": 0.5321472883224487,
+      "learning_rate": 6.352796052631578e-05,
+      "loss": 0.4891,
+      "step": 2188
+    },
+    {
+      "epoch": 2.428511995562335,
+      "grad_norm": 0.4424992799758911,
+      "learning_rate": 6.340460526315789e-05,
+      "loss": 0.5393,
+      "step": 2189
+    },
+    {
+      "epoch": 2.429621411732076,
+      "grad_norm": 0.37487635016441345,
+      "learning_rate": 6.328125e-05,
+      "loss": 0.2881,
+      "step": 2190
+    },
+    {
+      "epoch": 2.4307308279018165,
+      "grad_norm": 0.4092381000518799,
+      "learning_rate": 6.315789473684209e-05,
+      "loss": 0.3569,
+      "step": 2191
+    },
+    {
+      "epoch": 2.4318402440715574,
+      "grad_norm": 0.43235230445861816,
+      "learning_rate": 6.303453947368421e-05,
+      "loss": 0.3249,
+      "step": 2192
+    },
+    {
+      "epoch": 2.432949660241298,
+      "grad_norm": 0.4262683689594269,
+      "learning_rate": 6.29111842105263e-05,
+      "loss": 0.3653,
+      "step": 2193
+    },
+    {
+      "epoch": 2.434059076411039,
+      "grad_norm": 0.46941083669662476,
+      "learning_rate": 6.278782894736842e-05,
+      "loss": 0.3763,
+      "step": 2194
+    },
+    {
+      "epoch": 2.4351684925807793,
+      "grad_norm": 0.44827941060066223,
+      "learning_rate": 6.266447368421052e-05,
+      "loss": 0.5289,
+      "step": 2195
+    },
+    {
+      "epoch": 2.43627790875052,
+      "grad_norm": 0.4619573652744293,
+      "learning_rate": 6.254111842105262e-05,
+      "loss": 0.3311,
+      "step": 2196
+    },
+    {
+      "epoch": 2.4373873249202607,
+      "grad_norm": 0.4534587562084198,
+      "learning_rate": 6.241776315789473e-05,
+      "loss": 0.5748,
+      "step": 2197
+    },
+    {
+      "epoch": 2.4384967410900016,
+      "grad_norm": 0.5071384906768799,
+      "learning_rate": 6.229440789473684e-05,
+      "loss": 0.436,
+      "step": 2198
+    },
+    {
+      "epoch": 2.439606157259742,
+      "grad_norm": 0.31029054522514343,
+      "learning_rate": 6.217105263157894e-05,
+      "loss": 0.4601,
+      "step": 2199
+    },
+    {
+      "epoch": 2.4407155734294825,
+      "grad_norm": 0.6883265972137451,
+      "learning_rate": 6.204769736842105e-05,
+      "loss": 0.7862,
+      "step": 2200
+    },
+    {
+      "epoch": 2.4418249895992234,
+      "grad_norm": 0.5015659332275391,
+      "learning_rate": 6.192434210526316e-05,
+      "loss": 0.589,
+      "step": 2201
+    },
+    {
+      "epoch": 2.442934405768964,
+      "grad_norm": 0.5587977170944214,
+      "learning_rate": 6.180098684210526e-05,
+      "loss": 0.5756,
+      "step": 2202
+    },
+    {
+      "epoch": 2.444043821938705,
+      "grad_norm": 0.3763371706008911,
+      "learning_rate": 6.167763157894736e-05,
+      "loss": 0.3825,
+      "step": 2203
+    },
+    {
+      "epoch": 2.4451532381084453,
+      "grad_norm": 0.39078426361083984,
+      "learning_rate": 6.155427631578946e-05,
+      "loss": 0.2897,
+      "step": 2204
+    },
+    {
+      "epoch": 2.4462626542781862,
+      "grad_norm": 0.2745046615600586,
+      "learning_rate": 6.143092105263157e-05,
+      "loss": 0.3853,
+      "step": 2205
+    },
+    {
+      "epoch": 2.4473720704479267,
+      "grad_norm": 0.3368263840675354,
+      "learning_rate": 6.130756578947368e-05,
+      "loss": 0.4171,
+      "step": 2206
+    },
+    {
+      "epoch": 2.4484814866176676,
+      "grad_norm": 0.3768693208694458,
+      "learning_rate": 6.118421052631578e-05,
+      "loss": 0.3516,
+      "step": 2207
+    },
+    {
+      "epoch": 2.449590902787408,
+      "grad_norm": 0.7670299410820007,
+      "learning_rate": 6.106085526315789e-05,
+      "loss": 0.4042,
+      "step": 2208
+    },
+    {
+      "epoch": 2.450700318957149,
+      "grad_norm": 0.332269549369812,
+      "learning_rate": 6.09375e-05,
+      "loss": 0.4697,
+      "step": 2209
+    },
+    {
+      "epoch": 2.4518097351268895,
+      "grad_norm": 0.3532228171825409,
+      "learning_rate": 6.08141447368421e-05,
+      "loss": 0.3612,
+      "step": 2210
+    },
+    {
+      "epoch": 2.45291915129663,
+      "grad_norm": 0.39647355675697327,
+      "learning_rate": 6.0690789473684204e-05,
+      "loss": 0.3681,
+      "step": 2211
+    },
+    {
+      "epoch": 2.454028567466371,
+      "grad_norm": 0.3135978579521179,
+      "learning_rate": 6.056743421052631e-05,
+      "loss": 0.3349,
+      "step": 2212
+    },
+    {
+      "epoch": 2.4551379836361114,
+      "grad_norm": 0.3503597378730774,
+      "learning_rate": 6.044407894736842e-05,
+      "loss": 0.4885,
+      "step": 2213
+    },
+    {
+      "epoch": 2.4562473998058523,
+      "grad_norm": 0.3332939147949219,
+      "learning_rate": 6.032072368421052e-05,
+      "loss": 0.494,
+      "step": 2214
+    },
+    {
+      "epoch": 2.4573568159755927,
+      "grad_norm": 0.24943743646144867,
+      "learning_rate": 6.019736842105263e-05,
+      "loss": 0.2715,
+      "step": 2215
+    },
+    {
+      "epoch": 2.4584662321453337,
+      "grad_norm": 0.3770546615123749,
+      "learning_rate": 6.007401315789473e-05,
+      "loss": 0.4246,
+      "step": 2216
+    },
+    {
+      "epoch": 2.459575648315074,
+      "grad_norm": 0.474202960729599,
+      "learning_rate": 5.995065789473683e-05,
+      "loss": 0.4433,
+      "step": 2217
+    },
+    {
+      "epoch": 2.460685064484815,
+      "grad_norm": 0.5108838081359863,
+      "learning_rate": 5.9827302631578944e-05,
+      "loss": 0.3707,
+      "step": 2218
+    },
+    {
+      "epoch": 2.4617944806545555,
+      "grad_norm": 0.3123144507408142,
+      "learning_rate": 5.9703947368421044e-05,
+      "loss": 0.5036,
+      "step": 2219
+    },
+    {
+      "epoch": 2.4629038968242964,
+      "grad_norm": 0.5090911984443665,
+      "learning_rate": 5.958059210526316e-05,
+      "loss": 0.2868,
+      "step": 2220
+    },
+    {
+      "epoch": 2.464013312994037,
+      "grad_norm": 0.300322562456131,
+      "learning_rate": 5.945723684210526e-05,
+      "loss": 0.4943,
+      "step": 2221
+    },
+    {
+      "epoch": 2.4651227291637774,
+      "grad_norm": 0.5102697014808655,
+      "learning_rate": 5.933388157894736e-05,
+      "loss": 0.2345,
+      "step": 2222
+    },
+    {
+      "epoch": 2.4662321453335183,
+      "grad_norm": 0.28978249430656433,
+      "learning_rate": 5.921052631578947e-05,
+      "loss": 0.4279,
+      "step": 2223
+    },
+    {
+      "epoch": 2.467341561503259,
+      "grad_norm": 0.29028403759002686,
+      "learning_rate": 5.908717105263157e-05,
+      "loss": 0.4429,
+      "step": 2224
+    },
+    {
+      "epoch": 2.4684509776729997,
+      "grad_norm": 0.2963179349899292,
+      "learning_rate": 5.8963815789473684e-05,
+      "loss": 0.2816,
+      "step": 2225
+    },
+    {
+      "epoch": 2.46956039384274,
+      "grad_norm": 0.45071807503700256,
+      "learning_rate": 5.8840460526315784e-05,
+      "loss": 0.4315,
+      "step": 2226
+    },
+    {
+      "epoch": 2.470669810012481,
+      "grad_norm": 0.5235294699668884,
+      "learning_rate": 5.871710526315789e-05,
+      "loss": 0.4252,
+      "step": 2227
+    },
+    {
+      "epoch": 2.4717792261822216,
+      "grad_norm": 0.3098270893096924,
+      "learning_rate": 5.859375e-05,
+      "loss": 0.5957,
+      "step": 2228
+    },
+    {
+      "epoch": 2.472888642351962,
+      "grad_norm": 0.4504237473011017,
+      "learning_rate": 5.84703947368421e-05,
+      "loss": 0.4309,
+      "step": 2229
+    },
+    {
+      "epoch": 2.473998058521703,
+      "grad_norm": 0.3145286738872528,
+      "learning_rate": 5.8347039473684205e-05,
+      "loss": 0.4284,
+      "step": 2230
+    },
+    {
+      "epoch": 2.475107474691444,
+      "grad_norm": 0.39920830726623535,
+      "learning_rate": 5.822368421052631e-05,
+      "loss": 0.3452,
+      "step": 2231
+    },
+    {
+      "epoch": 2.4762168908611843,
+      "grad_norm": 0.5176841020584106,
+      "learning_rate": 5.810032894736842e-05,
+      "loss": 0.3673,
+      "step": 2232
+    },
+    {
+      "epoch": 2.477326307030925,
+      "grad_norm": 0.3197839558124542,
+      "learning_rate": 5.797697368421052e-05,
+      "loss": 0.3608,
+      "step": 2233
+    },
+    {
+      "epoch": 2.4784357232006657,
+      "grad_norm": 0.46264639496803284,
+      "learning_rate": 5.785361842105263e-05,
+      "loss": 0.3975,
+      "step": 2234
+    },
+    {
+      "epoch": 2.479545139370406,
+      "grad_norm": 0.6301188468933105,
+      "learning_rate": 5.773026315789473e-05,
+      "loss": 0.5299,
+      "step": 2235
+    },
+    {
+      "epoch": 2.480654555540147,
+      "grad_norm": 0.34785377979278564,
+      "learning_rate": 5.760690789473683e-05,
+      "loss": 0.3115,
+      "step": 2236
+    },
+    {
+      "epoch": 2.4817639717098876,
+      "grad_norm": 0.658812403678894,
+      "learning_rate": 5.7483552631578945e-05,
+      "loss": 0.363,
+      "step": 2237
+    },
+    {
+      "epoch": 2.4828733878796285,
+      "grad_norm": 0.5214020013809204,
+      "learning_rate": 5.7360197368421045e-05,
+      "loss": 0.4809,
+      "step": 2238
+    },
+    {
+      "epoch": 2.483982804049369,
+      "grad_norm": 0.5607793927192688,
+      "learning_rate": 5.723684210526316e-05,
+      "loss": 0.2884,
+      "step": 2239
+    },
+    {
+      "epoch": 2.4850922202191095,
+      "grad_norm": 0.39174753427505493,
+      "learning_rate": 5.711348684210526e-05,
+      "loss": 0.353,
+      "step": 2240
+    },
+    {
+      "epoch": 2.4862016363888504,
+      "grad_norm": 0.3475854694843292,
+      "learning_rate": 5.699013157894736e-05,
+      "loss": 0.4008,
+      "step": 2241
+    },
+    {
+      "epoch": 2.487311052558591,
+      "grad_norm": 0.5239514708518982,
+      "learning_rate": 5.686677631578947e-05,
+      "loss": 0.4618,
+      "step": 2242
+    },
+    {
+      "epoch": 2.488420468728332,
+      "grad_norm": 0.358995646238327,
+      "learning_rate": 5.674342105263157e-05,
+      "loss": 0.4906,
+      "step": 2243
+    },
+    {
+      "epoch": 2.4895298848980723,
+      "grad_norm": 0.3828662633895874,
+      "learning_rate": 5.662006578947368e-05,
+      "loss": 0.2941,
+      "step": 2244
+    },
+    {
+      "epoch": 2.490639301067813,
+      "grad_norm": 0.43704545497894287,
+      "learning_rate": 5.6496710526315785e-05,
+      "loss": 0.3913,
+      "step": 2245
+    },
+    {
+      "epoch": 2.4917487172375536,
+      "grad_norm": 0.29927560687065125,
+      "learning_rate": 5.637335526315789e-05,
+      "loss": 0.2661,
+      "step": 2246
+    },
+    {
+      "epoch": 2.4928581334072946,
+      "grad_norm": 0.36267396807670593,
+      "learning_rate": 5.625e-05,
+      "loss": 0.2605,
+      "step": 2247
+    },
+    {
+      "epoch": 2.493967549577035,
+      "grad_norm": 0.34938421845436096,
+      "learning_rate": 5.61266447368421e-05,
+      "loss": 0.3623,
+      "step": 2248
+    },
+    {
+      "epoch": 2.495076965746776,
+      "grad_norm": 0.4633193016052246,
+      "learning_rate": 5.6003289473684205e-05,
+      "loss": 0.3735,
+      "step": 2249
+    },
+    {
+      "epoch": 2.4961863819165164,
+      "grad_norm": 0.3852117359638214,
+      "learning_rate": 5.587993421052631e-05,
+      "loss": 0.5006,
+      "step": 2250
+    },
+    {
+      "epoch": 2.497295798086257,
+      "grad_norm": 0.528650164604187,
+      "learning_rate": 5.575657894736842e-05,
+      "loss": 0.4337,
+      "step": 2251
+    },
+    {
+      "epoch": 2.498405214255998,
+      "grad_norm": 0.3810504972934723,
+      "learning_rate": 5.563322368421052e-05,
+      "loss": 0.3518,
+      "step": 2252
+    },
+    {
+      "epoch": 2.4995146304257383,
+      "grad_norm": 0.35557541251182556,
+      "learning_rate": 5.550986842105263e-05,
+      "loss": 0.3932,
+      "step": 2253
+    },
+    {
+      "epoch": 2.500624046595479,
+      "grad_norm": 0.2806094288825989,
+      "learning_rate": 5.538651315789473e-05,
+      "loss": 0.2386,
+      "step": 2254
+    },
+    {
+      "epoch": 2.5017334627652197,
+      "grad_norm": 0.42694249749183655,
+      "learning_rate": 5.526315789473683e-05,
+      "loss": 0.4441,
+      "step": 2255
+    },
+    {
+      "epoch": 2.5028428789349606,
+      "grad_norm": 0.32019782066345215,
+      "learning_rate": 5.5139802631578945e-05,
+      "loss": 0.3024,
+      "step": 2256
+    },
+    {
+      "epoch": 2.503952295104701,
+      "grad_norm": 0.43957844376564026,
+      "learning_rate": 5.5016447368421045e-05,
+      "loss": 0.3916,
+      "step": 2257
+    },
+    {
+      "epoch": 2.5050617112744415,
+      "grad_norm": 0.39406758546829224,
+      "learning_rate": 5.489309210526316e-05,
+      "loss": 0.3808,
+      "step": 2258
+    },
+    {
+      "epoch": 2.5061711274441825,
+      "grad_norm": 0.36642491817474365,
+      "learning_rate": 5.476973684210526e-05,
+      "loss": 0.4663,
+      "step": 2259
+    },
+    {
+      "epoch": 2.5072805436139234,
+      "grad_norm": 0.3601360619068146,
+      "learning_rate": 5.464638157894736e-05,
+      "loss": 0.4061,
+      "step": 2260
+    },
+    {
+      "epoch": 2.508389959783664,
+      "grad_norm": 0.406791627407074,
+      "learning_rate": 5.452302631578947e-05,
+      "loss": 0.3205,
+      "step": 2261
+    },
+    {
+      "epoch": 2.5094993759534043,
+      "grad_norm": 0.3350258469581604,
+      "learning_rate": 5.439967105263157e-05,
+      "loss": 0.2646,
+      "step": 2262
+    },
+    {
+      "epoch": 2.5106087921231452,
+      "grad_norm": 0.48063212633132935,
+      "learning_rate": 5.427631578947368e-05,
+      "loss": 0.4569,
+      "step": 2263
+    },
+    {
+      "epoch": 2.5117182082928857,
+      "grad_norm": 0.4414843022823334,
+      "learning_rate": 5.4152960526315786e-05,
+      "loss": 0.3192,
+      "step": 2264
+    },
+    {
+      "epoch": 2.5128276244626266,
+      "grad_norm": 0.4843035042285919,
+      "learning_rate": 5.402960526315789e-05,
+      "loss": 0.3547,
+      "step": 2265
+    },
+    {
+      "epoch": 2.513937040632367,
+      "grad_norm": 0.30912336707115173,
+      "learning_rate": 5.390624999999999e-05,
+      "loss": 0.4092,
+      "step": 2266
+    },
+    {
+      "epoch": 2.515046456802108,
+      "grad_norm": 0.3699786365032196,
+      "learning_rate": 5.37828947368421e-05,
+      "loss": 0.3192,
+      "step": 2267
+    },
+    {
+      "epoch": 2.5161558729718485,
+      "grad_norm": 0.6730918884277344,
+      "learning_rate": 5.3659539473684206e-05,
+      "loss": 0.4083,
+      "step": 2268
+    },
+    {
+      "epoch": 2.517265289141589,
+      "grad_norm": 0.44793230295181274,
+      "learning_rate": 5.3536184210526306e-05,
+      "loss": 0.3056,
+      "step": 2269
+    },
+    {
+      "epoch": 2.51837470531133,
+      "grad_norm": 0.48553967475891113,
+      "learning_rate": 5.341282894736842e-05,
+      "loss": 0.2853,
+      "step": 2270
+    },
+    {
+      "epoch": 2.519484121481071,
+      "grad_norm": 0.3934726417064667,
+      "learning_rate": 5.328947368421052e-05,
+      "loss": 0.252,
+      "step": 2271
+    },
+    {
+      "epoch": 2.5205935376508113,
+      "grad_norm": 0.44591614603996277,
+      "learning_rate": 5.316611842105263e-05,
+      "loss": 0.5068,
+      "step": 2272
+    },
+    {
+      "epoch": 2.5217029538205518,
+      "grad_norm": 0.40541309118270874,
+      "learning_rate": 5.304276315789473e-05,
+      "loss": 0.427,
+      "step": 2273
+    },
+    {
+      "epoch": 2.5228123699902927,
+      "grad_norm": 0.4913922846317291,
+      "learning_rate": 5.291940789473683e-05,
+      "loss": 0.3514,
+      "step": 2274
+    },
+    {
+      "epoch": 2.523921786160033,
+      "grad_norm": 0.3643013834953308,
+      "learning_rate": 5.2796052631578946e-05,
+      "loss": 0.2301,
+      "step": 2275
+    },
+    {
+      "epoch": 2.525031202329774,
+      "grad_norm": 0.389931321144104,
+      "learning_rate": 5.2672697368421046e-05,
+      "loss": 0.5792,
+      "step": 2276
+    },
+    {
+      "epoch": 2.5261406184995145,
+      "grad_norm": 0.353927344083786,
+      "learning_rate": 5.254934210526316e-05,
+      "loss": 0.2489,
+      "step": 2277
+    },
+    {
+      "epoch": 2.5272500346692555,
+      "grad_norm": 0.6424699425697327,
+      "learning_rate": 5.242598684210526e-05,
+      "loss": 0.4858,
+      "step": 2278
+    },
+    {
+      "epoch": 2.528359450838996,
+      "grad_norm": 0.35985687375068665,
+      "learning_rate": 5.230263157894736e-05,
+      "loss": 0.3414,
+      "step": 2279
+    },
+    {
+      "epoch": 2.5294688670087364,
+      "grad_norm": 0.4226178228855133,
+      "learning_rate": 5.217927631578947e-05,
+      "loss": 0.4188,
+      "step": 2280
+    },
+    {
+      "epoch": 2.5305782831784773,
+      "grad_norm": 0.3257390558719635,
+      "learning_rate": 5.205592105263157e-05,
+      "loss": 0.4046,
+      "step": 2281
+    },
+    {
+      "epoch": 2.5316876993482182,
+      "grad_norm": 0.31944599747657776,
+      "learning_rate": 5.193256578947368e-05,
+      "loss": 0.3429,
+      "step": 2282
+    },
+    {
+      "epoch": 2.5327971155179587,
+      "grad_norm": 0.3145938515663147,
+      "learning_rate": 5.1809210526315786e-05,
+      "loss": 0.4005,
+      "step": 2283
+    },
+    {
+      "epoch": 2.533906531687699,
+      "grad_norm": 0.4402817189693451,
+      "learning_rate": 5.168585526315789e-05,
+      "loss": 0.4823,
+      "step": 2284
+    },
+    {
+      "epoch": 2.53501594785744,
+      "grad_norm": 0.43508538603782654,
+      "learning_rate": 5.156249999999999e-05,
+      "loss": 0.3428,
+      "step": 2285
+    },
+    {
+      "epoch": 2.5361253640271806,
+      "grad_norm": 0.4584806561470032,
+      "learning_rate": 5.14391447368421e-05,
+      "loss": 0.396,
+      "step": 2286
+    },
+    {
+      "epoch": 2.5372347801969215,
+      "grad_norm": 0.454815149307251,
+      "learning_rate": 5.1315789473684206e-05,
+      "loss": 0.4953,
+      "step": 2287
+    },
+    {
+      "epoch": 2.538344196366662,
+      "grad_norm": 0.35620346665382385,
+      "learning_rate": 5.1192434210526306e-05,
+      "loss": 0.4026,
+      "step": 2288
+    },
+    {
+      "epoch": 2.539453612536403,
+      "grad_norm": 0.34608957171440125,
+      "learning_rate": 5.106907894736842e-05,
+      "loss": 0.5012,
+      "step": 2289
+    },
+    {
+      "epoch": 2.5405630287061434,
+      "grad_norm": 0.3685770630836487,
+      "learning_rate": 5.094572368421052e-05,
+      "loss": 0.4459,
+      "step": 2290
+    },
+    {
+      "epoch": 2.541672444875884,
+      "grad_norm": 0.34064555168151855,
+      "learning_rate": 5.082236842105263e-05,
+      "loss": 0.206,
+      "step": 2291
+    },
+    {
+      "epoch": 2.5427818610456248,
+      "grad_norm": 0.5084649324417114,
+      "learning_rate": 5.069901315789473e-05,
+      "loss": 0.4602,
+      "step": 2292
+    },
+    {
+      "epoch": 2.5438912772153657,
+      "grad_norm": 0.4061320126056671,
+      "learning_rate": 5.057565789473683e-05,
+      "loss": 0.5329,
+      "step": 2293
+    },
+    {
+      "epoch": 2.545000693385106,
+      "grad_norm": 0.2851752042770386,
+      "learning_rate": 5.045230263157895e-05,
+      "loss": 0.2781,
+      "step": 2294
+    },
+    {
+      "epoch": 2.5461101095548466,
+      "grad_norm": 0.2984931170940399,
+      "learning_rate": 5.032894736842105e-05,
+      "loss": 0.3925,
+      "step": 2295
+    },
+    {
+      "epoch": 2.5472195257245875,
+      "grad_norm": 0.3785136342048645,
+      "learning_rate": 5.020559210526315e-05,
+      "loss": 0.2643,
+      "step": 2296
+    },
+    {
+      "epoch": 2.548328941894328,
+      "grad_norm": 0.4343215823173523,
+      "learning_rate": 5.008223684210526e-05,
+      "loss": 0.4111,
+      "step": 2297
+    },
+    {
+      "epoch": 2.549438358064069,
+      "grad_norm": 0.3174603581428528,
+      "learning_rate": 4.995888157894736e-05,
+      "loss": 0.3798,
+      "step": 2298
+    },
+    {
+      "epoch": 2.5505477742338094,
+      "grad_norm": 0.30842000246047974,
+      "learning_rate": 4.9835526315789474e-05,
+      "loss": 0.2997,
+      "step": 2299
+    },
+    {
+      "epoch": 2.5516571904035503,
+      "grad_norm": 0.42980441451072693,
+      "learning_rate": 4.9712171052631573e-05,
+      "loss": 0.3467,
+      "step": 2300
+    },
+    {
+      "epoch": 2.552766606573291,
+      "grad_norm": 0.3155680000782013,
+      "learning_rate": 4.958881578947368e-05,
+      "loss": 0.3337,
+      "step": 2301
+    },
+    {
+      "epoch": 2.5538760227430313,
+      "grad_norm": 0.38970938324928284,
+      "learning_rate": 4.946546052631579e-05,
+      "loss": 0.327,
+      "step": 2302
+    },
+    {
+      "epoch": 2.554985438912772,
+      "grad_norm": 0.43559134006500244,
+      "learning_rate": 4.9342105263157894e-05,
+      "loss": 0.5547,
+      "step": 2303
+    },
+    {
+      "epoch": 2.5560948550825127,
+      "grad_norm": 0.4793894290924072,
+      "learning_rate": 4.9218749999999994e-05,
+      "loss": 0.4121,
+      "step": 2304
+    },
+    {
+      "epoch": 2.5572042712522536,
+      "grad_norm": 0.5885838270187378,
+      "learning_rate": 4.90953947368421e-05,
+      "loss": 0.5028,
+      "step": 2305
+    },
+    {
+      "epoch": 2.558313687421994,
+      "grad_norm": 0.4064948260784149,
+      "learning_rate": 4.897203947368421e-05,
+      "loss": 0.257,
+      "step": 2306
+    },
+    {
+      "epoch": 2.559423103591735,
+      "grad_norm": 0.6676486730575562,
+      "learning_rate": 4.884868421052631e-05,
+      "loss": 0.6808,
+      "step": 2307
+    },
+    {
+      "epoch": 2.5605325197614754,
+      "grad_norm": 0.3958938419818878,
+      "learning_rate": 4.872532894736842e-05,
+      "loss": 0.4467,
+      "step": 2308
+    },
+    {
+      "epoch": 2.5616419359312164,
+      "grad_norm": 0.5410012602806091,
+      "learning_rate": 4.860197368421052e-05,
+      "loss": 0.4545,
+      "step": 2309
+    },
+    {
+      "epoch": 2.562751352100957,
+      "grad_norm": 0.31039872765541077,
+      "learning_rate": 4.8478618421052634e-05,
+      "loss": 0.4389,
+      "step": 2310
+    },
+    {
+      "epoch": 2.5638607682706978,
+      "grad_norm": 0.4607661962509155,
+      "learning_rate": 4.8355263157894734e-05,
+      "loss": 0.3213,
+      "step": 2311
+    },
+    {
+      "epoch": 2.5649701844404382,
+      "grad_norm": 0.33510398864746094,
+      "learning_rate": 4.8231907894736834e-05,
+      "loss": 0.3981,
+      "step": 2312
+    },
+    {
+      "epoch": 2.5660796006101787,
+      "grad_norm": 0.42955508828163147,
+      "learning_rate": 4.810855263157895e-05,
+      "loss": 0.2282,
+      "step": 2313
+    },
+    {
+      "epoch": 2.5671890167799196,
+      "grad_norm": 0.5348376035690308,
+      "learning_rate": 4.798519736842105e-05,
+      "loss": 0.4386,
+      "step": 2314
+    },
+    {
+      "epoch": 2.56829843294966,
+      "grad_norm": 0.4563250243663788,
+      "learning_rate": 4.7861842105263154e-05,
+      "loss": 0.3593,
+      "step": 2315
+    },
+    {
+      "epoch": 2.569407849119401,
+      "grad_norm": 0.38774073123931885,
+      "learning_rate": 4.773848684210526e-05,
+      "loss": 0.3159,
+      "step": 2316
+    },
+    {
+      "epoch": 2.5705172652891415,
+      "grad_norm": 0.33762556314468384,
+      "learning_rate": 4.761513157894736e-05,
+      "loss": 0.2853,
+      "step": 2317
+    },
+    {
+      "epoch": 2.5716266814588824,
+      "grad_norm": 0.5899233222007751,
+      "learning_rate": 4.749177631578947e-05,
+      "loss": 0.2734,
+      "step": 2318
+    },
+    {
+      "epoch": 2.572736097628623,
+      "grad_norm": 0.4161059558391571,
+      "learning_rate": 4.7368421052631574e-05,
+      "loss": 0.442,
+      "step": 2319
+    },
+    {
+      "epoch": 2.5738455137983633,
+      "grad_norm": 0.366187185049057,
+      "learning_rate": 4.724506578947368e-05,
+      "loss": 0.2767,
+      "step": 2320
+    },
+    {
+      "epoch": 2.5749549299681043,
+      "grad_norm": 0.325467586517334,
+      "learning_rate": 4.712171052631578e-05,
+      "loss": 0.4227,
+      "step": 2321
+    },
+    {
+      "epoch": 2.576064346137845,
+      "grad_norm": 0.31031447649002075,
+      "learning_rate": 4.6998355263157894e-05,
+      "loss": 0.3563,
+      "step": 2322
+    },
+    {
+      "epoch": 2.5771737623075857,
+      "grad_norm": 0.4601028263568878,
+      "learning_rate": 4.6874999999999994e-05,
+      "loss": 0.3945,
+      "step": 2323
+    },
+    {
+      "epoch": 2.578283178477326,
+      "grad_norm": 0.5773911476135254,
+      "learning_rate": 4.67516447368421e-05,
+      "loss": 0.5002,
+      "step": 2324
+    },
+    {
+      "epoch": 2.579392594647067,
+      "grad_norm": 0.3071858584880829,
+      "learning_rate": 4.662828947368421e-05,
+      "loss": 0.3305,
+      "step": 2325
+    },
+    {
+      "epoch": 2.5805020108168075,
+      "grad_norm": 0.42758241295814514,
+      "learning_rate": 4.650493421052631e-05,
+      "loss": 0.3915,
+      "step": 2326
+    },
+    {
+      "epoch": 2.5816114269865484,
+      "grad_norm": 0.4389335811138153,
+      "learning_rate": 4.638157894736842e-05,
+      "loss": 0.41,
+      "step": 2327
+    },
+    {
+      "epoch": 2.582720843156289,
+      "grad_norm": 0.4536753296852112,
+      "learning_rate": 4.625822368421052e-05,
+      "loss": 0.552,
+      "step": 2328
+    },
+    {
+      "epoch": 2.58383025932603,
+      "grad_norm": 0.4767647981643677,
+      "learning_rate": 4.6134868421052635e-05,
+      "loss": 0.2961,
+      "step": 2329
+    },
+    {
+      "epoch": 2.5849396754957703,
+      "grad_norm": 0.33526867628097534,
+      "learning_rate": 4.6011513157894734e-05,
+      "loss": 0.4272,
+      "step": 2330
+    },
+    {
+      "epoch": 2.586049091665511,
+      "grad_norm": 0.3391248285770416,
+      "learning_rate": 4.5888157894736834e-05,
+      "loss": 0.4728,
+      "step": 2331
+    },
+    {
+      "epoch": 2.5871585078352517,
+      "grad_norm": 0.44053715467453003,
+      "learning_rate": 4.576480263157895e-05,
+      "loss": 0.4216,
+      "step": 2332
+    },
+    {
+      "epoch": 2.5882679240049926,
+      "grad_norm": 0.3894062340259552,
+      "learning_rate": 4.564144736842105e-05,
+      "loss": 0.163,
+      "step": 2333
+    },
+    {
+      "epoch": 2.589377340174733,
+      "grad_norm": 0.37204012274742126,
+      "learning_rate": 4.5518092105263155e-05,
+      "loss": 0.388,
+      "step": 2334
+    },
+    {
+      "epoch": 2.5904867563444736,
+      "grad_norm": 0.3864336907863617,
+      "learning_rate": 4.539473684210526e-05,
+      "loss": 0.2925,
+      "step": 2335
+    },
+    {
+      "epoch": 2.5915961725142145,
+      "grad_norm": 0.2916948199272156,
+      "learning_rate": 4.527138157894736e-05,
+      "loss": 0.2073,
+      "step": 2336
+    },
+    {
+      "epoch": 2.592705588683955,
+      "grad_norm": 0.8837294578552246,
+      "learning_rate": 4.514802631578947e-05,
+      "loss": 0.2756,
+      "step": 2337
+    },
+    {
+      "epoch": 2.593815004853696,
+      "grad_norm": 0.41746076941490173,
+      "learning_rate": 4.5024671052631575e-05,
+      "loss": 0.4903,
+      "step": 2338
+    },
+    {
+      "epoch": 2.5949244210234363,
+      "grad_norm": 0.31884750723838806,
+      "learning_rate": 4.490131578947368e-05,
+      "loss": 0.4432,
+      "step": 2339
+    },
+    {
+      "epoch": 2.5960338371931773,
+      "grad_norm": 0.30209752917289734,
+      "learning_rate": 4.477796052631578e-05,
+      "loss": 0.3105,
+      "step": 2340
+    },
+    {
+      "epoch": 2.5971432533629177,
+      "grad_norm": 0.42073366045951843,
+      "learning_rate": 4.4654605263157895e-05,
+      "loss": 0.2862,
+      "step": 2341
+    },
+    {
+      "epoch": 2.598252669532658,
+      "grad_norm": 1.1312992572784424,
+      "learning_rate": 4.4531249999999995e-05,
+      "loss": 0.3452,
+      "step": 2342
+    },
+    {
+      "epoch": 2.599362085702399,
+      "grad_norm": 0.35556334257125854,
+      "learning_rate": 4.4407894736842095e-05,
+      "loss": 0.3874,
+      "step": 2343
+    },
+    {
+      "epoch": 2.60047150187214,
+      "grad_norm": 0.43012315034866333,
+      "learning_rate": 4.428453947368421e-05,
+      "loss": 0.3205,
+      "step": 2344
+    },
+    {
+      "epoch": 2.6015809180418805,
+      "grad_norm": 0.3909721076488495,
+      "learning_rate": 4.416118421052631e-05,
+      "loss": 0.3442,
+      "step": 2345
+    },
+    {
+      "epoch": 2.602690334211621,
+      "grad_norm": 0.4908634424209595,
+      "learning_rate": 4.403782894736842e-05,
+      "loss": 0.5702,
+      "step": 2346
+    },
+    {
+      "epoch": 2.603799750381362,
+      "grad_norm": 0.6141018867492676,
+      "learning_rate": 4.391447368421052e-05,
+      "loss": 0.4149,
+      "step": 2347
+    },
+    {
+      "epoch": 2.6049091665511024,
+      "grad_norm": 0.4573987126350403,
+      "learning_rate": 4.379111842105263e-05,
+      "loss": 0.4007,
+      "step": 2348
+    },
+    {
+      "epoch": 2.6060185827208433,
+      "grad_norm": 0.4271261692047119,
+      "learning_rate": 4.3667763157894735e-05,
+      "loss": 0.4075,
+      "step": 2349
+    },
+    {
+      "epoch": 2.6071279988905838,
+      "grad_norm": 0.39978379011154175,
+      "learning_rate": 4.3544407894736835e-05,
+      "loss": 0.4654,
+      "step": 2350
+    },
+    {
+      "epoch": 2.6082374150603247,
+      "grad_norm": 0.33403047919273376,
+      "learning_rate": 4.342105263157895e-05,
+      "loss": 0.4196,
+      "step": 2351
+    },
+    {
+      "epoch": 2.609346831230065,
+      "grad_norm": 0.5256022214889526,
+      "learning_rate": 4.329769736842105e-05,
+      "loss": 0.2752,
+      "step": 2352
+    },
+    {
+      "epoch": 2.6104562473998056,
+      "grad_norm": 0.4153258204460144,
+      "learning_rate": 4.3174342105263155e-05,
+      "loss": 0.3468,
+      "step": 2353
+    },
+    {
+      "epoch": 2.6115656635695466,
+      "grad_norm": 0.6061992645263672,
+      "learning_rate": 4.305098684210526e-05,
+      "loss": 0.351,
+      "step": 2354
+    },
+    {
+      "epoch": 2.6126750797392875,
+      "grad_norm": 0.295296847820282,
+      "learning_rate": 4.292763157894736e-05,
+      "loss": 0.2476,
+      "step": 2355
+    },
+    {
+      "epoch": 2.613784495909028,
+      "grad_norm": 0.3813928961753845,
+      "learning_rate": 4.280427631578947e-05,
+      "loss": 0.3138,
+      "step": 2356
+    },
+    {
+      "epoch": 2.6148939120787684,
+      "grad_norm": 0.3294810652732849,
+      "learning_rate": 4.2680921052631575e-05,
+      "loss": 0.3273,
+      "step": 2357
+    },
+    {
+      "epoch": 2.6160033282485093,
+      "grad_norm": 0.4694172739982605,
+      "learning_rate": 4.255756578947368e-05,
+      "loss": 0.3346,
+      "step": 2358
+    },
+    {
+      "epoch": 2.61711274441825,
+      "grad_norm": 0.5706562399864197,
+      "learning_rate": 4.243421052631578e-05,
+      "loss": 0.2241,
+      "step": 2359
+    },
+    {
+      "epoch": 2.6182221605879907,
+      "grad_norm": 0.6200342774391174,
+      "learning_rate": 4.2310855263157896e-05,
+      "loss": 0.4129,
+      "step": 2360
+    },
+    {
+      "epoch": 2.619331576757731,
+      "grad_norm": 0.4055767059326172,
+      "learning_rate": 4.2187499999999995e-05,
+      "loss": 0.4006,
+      "step": 2361
+    },
+    {
+      "epoch": 2.620440992927472,
+      "grad_norm": 0.33380985260009766,
+      "learning_rate": 4.2064144736842095e-05,
+      "loss": 0.3732,
+      "step": 2362
+    },
+    {
+      "epoch": 2.6215504090972126,
+      "grad_norm": 0.41517943143844604,
+      "learning_rate": 4.194078947368421e-05,
+      "loss": 0.2458,
+      "step": 2363
+    },
+    {
+      "epoch": 2.622659825266953,
+      "grad_norm": 0.3231419324874878,
+      "learning_rate": 4.181743421052631e-05,
+      "loss": 0.5769,
+      "step": 2364
+    },
+    {
+      "epoch": 2.623769241436694,
+      "grad_norm": 0.4171946942806244,
+      "learning_rate": 4.169407894736842e-05,
+      "loss": 0.4702,
+      "step": 2365
+    },
+    {
+      "epoch": 2.6248786576064345,
+      "grad_norm": 0.3811667263507843,
+      "learning_rate": 4.157072368421052e-05,
+      "loss": 0.3849,
+      "step": 2366
+    },
+    {
+      "epoch": 2.6259880737761754,
+      "grad_norm": 0.41539669036865234,
+      "learning_rate": 4.144736842105263e-05,
+      "loss": 0.3845,
+      "step": 2367
+    },
+    {
+      "epoch": 2.627097489945916,
+      "grad_norm": 0.39716070890426636,
+      "learning_rate": 4.1324013157894736e-05,
+      "loss": 0.3744,
+      "step": 2368
+    },
+    {
+      "epoch": 2.6282069061156568,
+      "grad_norm": 0.38069993257522583,
+      "learning_rate": 4.1200657894736836e-05,
+      "loss": 0.2432,
+      "step": 2369
+    },
+    {
+      "epoch": 2.6293163222853972,
+      "grad_norm": 0.3581462502479553,
+      "learning_rate": 4.107730263157894e-05,
+      "loss": 0.6516,
+      "step": 2370
+    },
+    {
+      "epoch": 2.6304257384551377,
+      "grad_norm": 0.34660494327545166,
+      "learning_rate": 4.095394736842105e-05,
+      "loss": 0.3599,
+      "step": 2371
+    },
+    {
+      "epoch": 2.6315351546248786,
+      "grad_norm": 0.34078100323677063,
+      "learning_rate": 4.0830592105263156e-05,
+      "loss": 0.3611,
+      "step": 2372
+    },
+    {
+      "epoch": 2.6326445707946196,
+      "grad_norm": 0.3904211223125458,
+      "learning_rate": 4.0707236842105256e-05,
+      "loss": 0.301,
+      "step": 2373
+    },
+    {
+      "epoch": 2.63375398696436,
+      "grad_norm": 0.3696337342262268,
+      "learning_rate": 4.058388157894736e-05,
+      "loss": 0.3051,
+      "step": 2374
+    },
+    {
+      "epoch": 2.6348634031341005,
+      "grad_norm": 0.43636754155158997,
+      "learning_rate": 4.046052631578947e-05,
+      "loss": 0.4971,
+      "step": 2375
+    },
+    {
+      "epoch": 2.6359728193038414,
+      "grad_norm": 0.5181596279144287,
+      "learning_rate": 4.0337171052631576e-05,
+      "loss": 0.3127,
+      "step": 2376
+    },
+    {
+      "epoch": 2.637082235473582,
+      "grad_norm": 0.5476127862930298,
+      "learning_rate": 4.021381578947368e-05,
+      "loss": 0.5049,
+      "step": 2377
+    },
+    {
+      "epoch": 2.638191651643323,
+      "grad_norm": 0.5447264909744263,
+      "learning_rate": 4.009046052631578e-05,
+      "loss": 0.5255,
+      "step": 2378
+    },
+    {
+      "epoch": 2.6393010678130633,
+      "grad_norm": 0.5430484414100647,
+      "learning_rate": 3.9967105263157896e-05,
+      "loss": 0.3371,
+      "step": 2379
+    },
+    {
+      "epoch": 2.640410483982804,
+      "grad_norm": 0.4831867516040802,
+      "learning_rate": 3.9843749999999996e-05,
+      "loss": 0.4452,
+      "step": 2380
+    },
+    {
+      "epoch": 2.6415199001525447,
+      "grad_norm": 0.6055355072021484,
+      "learning_rate": 3.9720394736842096e-05,
+      "loss": 0.5317,
+      "step": 2381
+    },
+    {
+      "epoch": 2.642629316322285,
+      "grad_norm": 0.4254622757434845,
+      "learning_rate": 3.959703947368421e-05,
+      "loss": 0.3434,
+      "step": 2382
+    },
+    {
+      "epoch": 2.643738732492026,
+      "grad_norm": 0.4942837655544281,
+      "learning_rate": 3.947368421052631e-05,
+      "loss": 0.4468,
+      "step": 2383
+    },
+    {
+      "epoch": 2.644848148661767,
+      "grad_norm": 0.3226402997970581,
+      "learning_rate": 3.935032894736842e-05,
+      "loss": 0.4929,
+      "step": 2384
+    },
+    {
+      "epoch": 2.6459575648315075,
+      "grad_norm": 0.4470663368701935,
+      "learning_rate": 3.922697368421052e-05,
+      "loss": 0.33,
+      "step": 2385
+    },
+    {
+      "epoch": 2.647066981001248,
+      "grad_norm": 0.5086075663566589,
+      "learning_rate": 3.910361842105263e-05,
+      "loss": 0.6575,
+      "step": 2386
+    },
+    {
+      "epoch": 2.648176397170989,
+      "grad_norm": 0.39487743377685547,
+      "learning_rate": 3.8980263157894736e-05,
+      "loss": 0.3375,
+      "step": 2387
+    },
+    {
+      "epoch": 2.6492858133407293,
+      "grad_norm": 0.6295050978660583,
+      "learning_rate": 3.8856907894736836e-05,
+      "loss": 0.398,
+      "step": 2388
+    },
+    {
+      "epoch": 2.6503952295104702,
+      "grad_norm": 0.5763193368911743,
+      "learning_rate": 3.873355263157894e-05,
+      "loss": 0.3891,
+      "step": 2389
+    },
+    {
+      "epoch": 2.6515046456802107,
+      "grad_norm": 0.33462652564048767,
+      "learning_rate": 3.861019736842105e-05,
+      "loss": 0.428,
+      "step": 2390
+    },
+    {
+      "epoch": 2.6526140618499516,
+      "grad_norm": 0.6703096628189087,
+      "learning_rate": 3.8486842105263156e-05,
+      "loss": 0.3148,
+      "step": 2391
+    },
+    {
+      "epoch": 2.653723478019692,
+      "grad_norm": 0.3808022141456604,
+      "learning_rate": 3.8363486842105256e-05,
+      "loss": 0.4022,
+      "step": 2392
+    },
+    {
+      "epoch": 2.6548328941894326,
+      "grad_norm": 0.42382562160491943,
+      "learning_rate": 3.824013157894736e-05,
+      "loss": 0.2387,
+      "step": 2393
+    },
+    {
+      "epoch": 2.6559423103591735,
+      "grad_norm": 0.4965885579586029,
+      "learning_rate": 3.811677631578947e-05,
+      "loss": 0.5139,
+      "step": 2394
+    },
+    {
+      "epoch": 2.6570517265289144,
+      "grad_norm": 0.4375743269920349,
+      "learning_rate": 3.799342105263157e-05,
+      "loss": 0.3922,
+      "step": 2395
+    },
+    {
+      "epoch": 2.658161142698655,
+      "grad_norm": 0.44008010625839233,
+      "learning_rate": 3.787006578947368e-05,
+      "loss": 0.8074,
+      "step": 2396
+    },
+    {
+      "epoch": 2.6592705588683954,
+      "grad_norm": 0.40815237164497375,
+      "learning_rate": 3.774671052631578e-05,
+      "loss": 0.4673,
+      "step": 2397
+    },
+    {
+      "epoch": 2.6603799750381363,
+      "grad_norm": 0.3250466287136078,
+      "learning_rate": 3.76233552631579e-05,
+      "loss": 0.3603,
+      "step": 2398
+    },
+    {
+      "epoch": 2.6614893912078768,
+      "grad_norm": 0.4003657400608063,
+      "learning_rate": 3.75e-05,
+      "loss": 0.5028,
+      "step": 2399
+    },
+    {
+      "epoch": 2.6625988073776177,
+      "grad_norm": 0.475201278924942,
+      "learning_rate": 3.7376644736842103e-05,
+      "loss": 0.2309,
+      "step": 2400
+    },
+    {
+      "epoch": 2.663708223547358,
+      "grad_norm": 0.4348124563694,
+      "learning_rate": 3.725328947368421e-05,
+      "loss": 0.3997,
+      "step": 2401
+    },
+    {
+      "epoch": 2.664817639717099,
+      "grad_norm": 0.5189718008041382,
+      "learning_rate": 3.712993421052632e-05,
+      "loss": 0.4532,
+      "step": 2402
+    },
+    {
+      "epoch": 2.6659270558868395,
+      "grad_norm": 0.293647438287735,
+      "learning_rate": 3.700657894736842e-05,
+      "loss": 0.4344,
+      "step": 2403
+    },
+    {
+      "epoch": 2.66703647205658,
+      "grad_norm": 0.4177672564983368,
+      "learning_rate": 3.6883223684210524e-05,
+      "loss": 0.4263,
+      "step": 2404
+    },
+    {
+      "epoch": 2.668145888226321,
+      "grad_norm": 0.4328800141811371,
+      "learning_rate": 3.675986842105263e-05,
+      "loss": 0.3869,
+      "step": 2405
+    },
+    {
+      "epoch": 2.669255304396062,
+      "grad_norm": 0.3846050202846527,
+      "learning_rate": 3.663651315789474e-05,
+      "loss": 0.4011,
+      "step": 2406
+    },
+    {
+      "epoch": 2.6703647205658023,
+      "grad_norm": 0.3619375228881836,
+      "learning_rate": 3.651315789473684e-05,
+      "loss": 0.2928,
+      "step": 2407
+    },
+    {
+      "epoch": 2.671474136735543,
+      "grad_norm": 0.5226150155067444,
+      "learning_rate": 3.6389802631578944e-05,
+      "loss": 0.2546,
+      "step": 2408
+    },
+    {
+      "epoch": 2.6725835529052837,
+      "grad_norm": 0.3613680303096771,
+      "learning_rate": 3.626644736842105e-05,
+      "loss": 0.3706,
+      "step": 2409
+    },
+    {
+      "epoch": 2.673692969075024,
+      "grad_norm": 0.38295650482177734,
+      "learning_rate": 3.614309210526315e-05,
+      "loss": 0.3319,
+      "step": 2410
+    },
+    {
+      "epoch": 2.674802385244765,
+      "grad_norm": 0.3644557595252991,
+      "learning_rate": 3.601973684210526e-05,
+      "loss": 0.3344,
+      "step": 2411
+    },
+    {
+      "epoch": 2.6759118014145056,
+      "grad_norm": 0.4107860028743744,
+      "learning_rate": 3.5896381578947364e-05,
+      "loss": 0.2881,
+      "step": 2412
+    },
+    {
+      "epoch": 2.6770212175842465,
+      "grad_norm": 0.39361506700515747,
+      "learning_rate": 3.577302631578947e-05,
+      "loss": 0.3575,
+      "step": 2413
+    },
+    {
+      "epoch": 2.678130633753987,
+      "grad_norm": 0.30774053931236267,
+      "learning_rate": 3.564967105263158e-05,
+      "loss": 0.2718,
+      "step": 2414
+    },
+    {
+      "epoch": 2.6792400499237274,
+      "grad_norm": 0.40254542231559753,
+      "learning_rate": 3.552631578947368e-05,
+      "loss": 0.4364,
+      "step": 2415
+    },
+    {
+      "epoch": 2.6803494660934684,
+      "grad_norm": 0.39855626225471497,
+      "learning_rate": 3.5402960526315784e-05,
+      "loss": 0.2258,
+      "step": 2416
+    },
+    {
+      "epoch": 2.681458882263209,
+      "grad_norm": 0.3063789904117584,
+      "learning_rate": 3.527960526315789e-05,
+      "loss": 0.3248,
+      "step": 2417
+    },
+    {
+      "epoch": 2.6825682984329497,
+      "grad_norm": 0.48203665018081665,
+      "learning_rate": 3.515625e-05,
+      "loss": 0.5971,
+      "step": 2418
+    },
+    {
+      "epoch": 2.6836777146026902,
+      "grad_norm": 0.40560922026634216,
+      "learning_rate": 3.5032894736842104e-05,
+      "loss": 0.2832,
+      "step": 2419
+    },
+    {
+      "epoch": 2.684787130772431,
+      "grad_norm": 0.4571300745010376,
+      "learning_rate": 3.490953947368421e-05,
+      "loss": 0.414,
+      "step": 2420
+    },
+    {
+      "epoch": 2.6858965469421716,
+      "grad_norm": 0.29357972741127014,
+      "learning_rate": 3.478618421052632e-05,
+      "loss": 0.4382,
+      "step": 2421
+    },
+    {
+      "epoch": 2.687005963111912,
+      "grad_norm": 0.4692830741405487,
+      "learning_rate": 3.466282894736842e-05,
+      "loss": 0.4557,
+      "step": 2422
+    },
+    {
+      "epoch": 2.688115379281653,
+      "grad_norm": 0.5453107953071594,
+      "learning_rate": 3.4539473684210524e-05,
+      "loss": 0.456,
+      "step": 2423
+    },
+    {
+      "epoch": 2.689224795451394,
+      "grad_norm": 0.8587030172348022,
+      "learning_rate": 3.441611842105263e-05,
+      "loss": 0.5188,
+      "step": 2424
+    },
+    {
+      "epoch": 2.6903342116211344,
+      "grad_norm": 0.5260380506515503,
+      "learning_rate": 3.429276315789474e-05,
+      "loss": 0.4475,
+      "step": 2425
+    },
+    {
+      "epoch": 2.691443627790875,
+      "grad_norm": 0.2895418405532837,
+      "learning_rate": 3.416940789473684e-05,
+      "loss": 0.245,
+      "step": 2426
+    },
+    {
+      "epoch": 2.692553043960616,
+      "grad_norm": 0.44756054878234863,
+      "learning_rate": 3.4046052631578944e-05,
+      "loss": 0.4375,
+      "step": 2427
+    },
+    {
+      "epoch": 2.6936624601303563,
+      "grad_norm": 0.39095133543014526,
+      "learning_rate": 3.392269736842105e-05,
+      "loss": 0.405,
+      "step": 2428
+    },
+    {
+      "epoch": 2.694771876300097,
+      "grad_norm": 0.6433843374252319,
+      "learning_rate": 3.379934210526315e-05,
+      "loss": 0.2905,
+      "step": 2429
+    },
+    {
+      "epoch": 2.6958812924698377,
+      "grad_norm": 0.5514675974845886,
+      "learning_rate": 3.367598684210526e-05,
+      "loss": 0.3651,
+      "step": 2430
+    },
+    {
+      "epoch": 2.6969907086395786,
+      "grad_norm": 0.4037233293056488,
+      "learning_rate": 3.3552631578947364e-05,
+      "loss": 0.5534,
+      "step": 2431
+    },
+    {
+      "epoch": 2.698100124809319,
+      "grad_norm": 0.3961438238620758,
+      "learning_rate": 3.342927631578947e-05,
+      "loss": 0.4399,
+      "step": 2432
+    },
+    {
+      "epoch": 2.6992095409790595,
+      "grad_norm": 0.38111430406570435,
+      "learning_rate": 3.330592105263158e-05,
+      "loss": 0.3532,
+      "step": 2433
+    },
+    {
+      "epoch": 2.7003189571488004,
+      "grad_norm": 0.35400980710983276,
+      "learning_rate": 3.318256578947368e-05,
+      "loss": 0.5905,
+      "step": 2434
+    },
+    {
+      "epoch": 2.7014283733185414,
+      "grad_norm": 0.469614714384079,
+      "learning_rate": 3.3059210526315785e-05,
+      "loss": 0.4461,
+      "step": 2435
+    },
+    {
+      "epoch": 2.702537789488282,
+      "grad_norm": 0.2964523732662201,
+      "learning_rate": 3.293585526315789e-05,
+      "loss": 0.3752,
+      "step": 2436
+    },
+    {
+      "epoch": 2.7036472056580223,
+      "grad_norm": 0.4390257000923157,
+      "learning_rate": 3.28125e-05,
+      "loss": 0.384,
+      "step": 2437
+    },
+    {
+      "epoch": 2.704756621827763,
+      "grad_norm": 0.6685402393341064,
+      "learning_rate": 3.2689144736842105e-05,
+      "loss": 0.3686,
+      "step": 2438
+    },
+    {
+      "epoch": 2.7058660379975037,
+      "grad_norm": 0.42853912711143494,
+      "learning_rate": 3.256578947368421e-05,
+      "loss": 0.3165,
+      "step": 2439
+    },
+    {
+      "epoch": 2.7069754541672446,
+      "grad_norm": 0.382541686296463,
+      "learning_rate": 3.244243421052632e-05,
+      "loss": 0.4957,
+      "step": 2440
+    },
+    {
+      "epoch": 2.708084870336985,
+      "grad_norm": 0.5440720319747925,
+      "learning_rate": 3.231907894736842e-05,
+      "loss": 0.4417,
+      "step": 2441
+    },
+    {
+      "epoch": 2.709194286506726,
+      "grad_norm": 0.4604506492614746,
+      "learning_rate": 3.2195723684210525e-05,
+      "loss": 0.4257,
+      "step": 2442
+    },
+    {
+      "epoch": 2.7103037026764665,
+      "grad_norm": 0.45395079255104065,
+      "learning_rate": 3.207236842105263e-05,
+      "loss": 0.2593,
+      "step": 2443
+    },
+    {
+      "epoch": 2.711413118846207,
+      "grad_norm": 0.4349961578845978,
+      "learning_rate": 3.194901315789473e-05,
+      "loss": 0.3596,
+      "step": 2444
+    },
+    {
+      "epoch": 2.712522535015948,
+      "grad_norm": 0.46334758400917053,
+      "learning_rate": 3.182565789473684e-05,
+      "loss": 0.3347,
+      "step": 2445
+    },
+    {
+      "epoch": 2.713631951185689,
+      "grad_norm": 0.4385397136211395,
+      "learning_rate": 3.1702302631578945e-05,
+      "loss": 0.1988,
+      "step": 2446
+    },
+    {
+      "epoch": 2.7147413673554293,
+      "grad_norm": 0.6983950734138489,
+      "learning_rate": 3.1578947368421045e-05,
+      "loss": 0.3902,
+      "step": 2447
+    },
+    {
+      "epoch": 2.7158507835251697,
+      "grad_norm": 0.46519941091537476,
+      "learning_rate": 3.145559210526315e-05,
+      "loss": 0.9286,
+      "step": 2448
+    },
+    {
+      "epoch": 2.7169601996949106,
+      "grad_norm": 0.4739225506782532,
+      "learning_rate": 3.133223684210526e-05,
+      "loss": 0.4028,
+      "step": 2449
+    },
+    {
+      "epoch": 2.718069615864651,
+      "grad_norm": 0.45045381784439087,
+      "learning_rate": 3.1208881578947365e-05,
+      "loss": 0.2989,
+      "step": 2450
+    },
+    {
+      "epoch": 2.719179032034392,
+      "grad_norm": 0.5387830138206482,
+      "learning_rate": 3.108552631578947e-05,
+      "loss": 0.3876,
+      "step": 2451
+    },
+    {
+      "epoch": 2.7202884482041325,
+      "grad_norm": 0.4015057682991028,
+      "learning_rate": 3.096217105263158e-05,
+      "loss": 0.3363,
+      "step": 2452
+    },
+    {
+      "epoch": 2.7213978643738734,
+      "grad_norm": 0.9120666980743408,
+      "learning_rate": 3.083881578947368e-05,
+      "loss": 0.3437,
+      "step": 2453
+    },
+    {
+      "epoch": 2.722507280543614,
+      "grad_norm": 0.3784855902194977,
+      "learning_rate": 3.0715460526315785e-05,
+      "loss": 0.3015,
+      "step": 2454
+    },
+    {
+      "epoch": 2.7236166967133544,
+      "grad_norm": 0.3855441212654114,
+      "learning_rate": 3.059210526315789e-05,
+      "loss": 0.3813,
+      "step": 2455
+    },
+    {
+      "epoch": 2.7247261128830953,
+      "grad_norm": 0.8900882601737976,
+      "learning_rate": 3.046875e-05,
+      "loss": 0.6206,
+      "step": 2456
+    },
+    {
+      "epoch": 2.725835529052836,
+      "grad_norm": 0.43435898423194885,
+      "learning_rate": 3.0345394736842102e-05,
+      "loss": 0.2936,
+      "step": 2457
+    },
+    {
+      "epoch": 2.7269449452225767,
+      "grad_norm": 0.4391900300979614,
+      "learning_rate": 3.022203947368421e-05,
+      "loss": 0.6765,
+      "step": 2458
+    },
+    {
+      "epoch": 2.728054361392317,
+      "grad_norm": 0.566494882106781,
+      "learning_rate": 3.0098684210526315e-05,
+      "loss": 0.3304,
+      "step": 2459
+    },
+    {
+      "epoch": 2.729163777562058,
+      "grad_norm": 0.35425856709480286,
+      "learning_rate": 2.9975328947368415e-05,
+      "loss": 0.3772,
+      "step": 2460
+    },
+    {
+      "epoch": 2.7302731937317986,
+      "grad_norm": 0.3330824673175812,
+      "learning_rate": 2.9851973684210522e-05,
+      "loss": 0.3451,
+      "step": 2461
+    },
+    {
+      "epoch": 2.7313826099015395,
+      "grad_norm": 0.5032577514648438,
+      "learning_rate": 2.972861842105263e-05,
+      "loss": 0.4958,
+      "step": 2462
+    },
+    {
+      "epoch": 2.73249202607128,
+      "grad_norm": 0.46107593178749084,
+      "learning_rate": 2.9605263157894735e-05,
+      "loss": 0.5935,
+      "step": 2463
+    },
+    {
+      "epoch": 2.733601442241021,
+      "grad_norm": 0.4279358983039856,
+      "learning_rate": 2.9481907894736842e-05,
+      "loss": 0.352,
+      "step": 2464
+    },
+    {
+      "epoch": 2.7347108584107613,
+      "grad_norm": 0.35816699266433716,
+      "learning_rate": 2.9358552631578946e-05,
+      "loss": 0.4058,
+      "step": 2465
+    },
+    {
+      "epoch": 2.735820274580502,
+      "grad_norm": 0.3129696249961853,
+      "learning_rate": 2.923519736842105e-05,
+      "loss": 0.2704,
+      "step": 2466
+    },
+    {
+      "epoch": 2.7369296907502427,
+      "grad_norm": 0.5344390869140625,
+      "learning_rate": 2.9111842105263156e-05,
+      "loss": 0.3926,
+      "step": 2467
+    },
+    {
+      "epoch": 2.738039106919983,
+      "grad_norm": 0.3743567168712616,
+      "learning_rate": 2.898848684210526e-05,
+      "loss": 0.3455,
+      "step": 2468
+    },
+    {
+      "epoch": 2.739148523089724,
+      "grad_norm": 0.4825611412525177,
+      "learning_rate": 2.8865131578947366e-05,
+      "loss": 0.4537,
+      "step": 2469
+    },
+    {
+      "epoch": 2.7402579392594646,
+      "grad_norm": 0.47796013951301575,
+      "learning_rate": 2.8741776315789472e-05,
+      "loss": 0.4268,
+      "step": 2470
+    },
+    {
+      "epoch": 2.7413673554292055,
+      "grad_norm": 0.38644805550575256,
+      "learning_rate": 2.861842105263158e-05,
+      "loss": 0.3989,
+      "step": 2471
+    },
+    {
+      "epoch": 2.742476771598946,
+      "grad_norm": 0.41950738430023193,
+      "learning_rate": 2.849506578947368e-05,
+      "loss": 0.4983,
+      "step": 2472
+    },
+    {
+      "epoch": 2.7435861877686865,
+      "grad_norm": 0.34149548411369324,
+      "learning_rate": 2.8371710526315786e-05,
+      "loss": 0.4519,
+      "step": 2473
+    },
+    {
+      "epoch": 2.7446956039384274,
+      "grad_norm": 0.4213709831237793,
+      "learning_rate": 2.8248355263157893e-05,
+      "loss": 0.4212,
+      "step": 2474
+    },
+    {
+      "epoch": 2.7458050201081683,
+      "grad_norm": 0.3141988217830658,
+      "learning_rate": 2.8125e-05,
+      "loss": 0.3161,
+      "step": 2475
+    },
+    {
+      "epoch": 2.7469144362779088,
+      "grad_norm": 0.3500676155090332,
+      "learning_rate": 2.8001644736842103e-05,
+      "loss": 0.3603,
+      "step": 2476
+    },
+    {
+      "epoch": 2.7480238524476492,
+      "grad_norm": 0.4086303114891052,
+      "learning_rate": 2.787828947368421e-05,
+      "loss": 0.5926,
+      "step": 2477
+    },
+    {
+      "epoch": 2.74913326861739,
+      "grad_norm": 0.5967698097229004,
+      "learning_rate": 2.7754934210526316e-05,
+      "loss": 0.237,
+      "step": 2478
+    },
+    {
+      "epoch": 2.7502426847871306,
+      "grad_norm": 0.4627840220928192,
+      "learning_rate": 2.7631578947368416e-05,
+      "loss": 0.3304,
+      "step": 2479
+    },
+    {
+      "epoch": 2.7513521009568715,
+      "grad_norm": 0.37266799807548523,
+      "learning_rate": 2.7508223684210523e-05,
+      "loss": 0.5739,
+      "step": 2480
+    },
+    {
+      "epoch": 2.752461517126612,
+      "grad_norm": 0.42507204413414,
+      "learning_rate": 2.738486842105263e-05,
+      "loss": 0.4707,
+      "step": 2481
+    },
+    {
+      "epoch": 2.753570933296353,
+      "grad_norm": 0.4577075242996216,
+      "learning_rate": 2.7261513157894736e-05,
+      "loss": 0.5197,
+      "step": 2482
+    },
+    {
+      "epoch": 2.7546803494660934,
+      "grad_norm": 0.6312010884284973,
+      "learning_rate": 2.713815789473684e-05,
+      "loss": 0.5129,
+      "step": 2483
+    },
+    {
+      "epoch": 2.755789765635834,
+      "grad_norm": 0.5095051527023315,
+      "learning_rate": 2.7014802631578946e-05,
+      "loss": 0.329,
+      "step": 2484
+    },
+    {
+      "epoch": 2.756899181805575,
+      "grad_norm": 0.5552304983139038,
+      "learning_rate": 2.689144736842105e-05,
+      "loss": 0.516,
+      "step": 2485
+    },
+    {
+      "epoch": 2.7580085979753157,
+      "grad_norm": 0.4446256458759308,
+      "learning_rate": 2.6768092105263153e-05,
+      "loss": 0.4809,
+      "step": 2486
+    },
+    {
+      "epoch": 2.759118014145056,
+      "grad_norm": 0.45802775025367737,
+      "learning_rate": 2.664473684210526e-05,
+      "loss": 0.2507,
+      "step": 2487
+    },
+    {
+      "epoch": 2.7602274303147967,
+      "grad_norm": 1.002074956893921,
+      "learning_rate": 2.6521381578947366e-05,
+      "loss": 0.3177,
+      "step": 2488
+    },
+    {
+      "epoch": 2.7613368464845376,
+      "grad_norm": 0.37677767872810364,
+      "learning_rate": 2.6398026315789473e-05,
+      "loss": 0.3994,
+      "step": 2489
+    },
+    {
+      "epoch": 2.762446262654278,
+      "grad_norm": 0.5203359723091125,
+      "learning_rate": 2.627467105263158e-05,
+      "loss": 0.3421,
+      "step": 2490
+    },
+    {
+      "epoch": 2.763555678824019,
+      "grad_norm": 0.41637536883354187,
+      "learning_rate": 2.615131578947368e-05,
+      "loss": 0.4927,
+      "step": 2491
+    },
+    {
+      "epoch": 2.7646650949937595,
+      "grad_norm": 0.44756412506103516,
+      "learning_rate": 2.6027960526315786e-05,
+      "loss": 0.3734,
+      "step": 2492
+    },
+    {
+      "epoch": 2.7657745111635004,
+      "grad_norm": 0.3564557731151581,
+      "learning_rate": 2.5904605263157893e-05,
+      "loss": 0.3779,
+      "step": 2493
+    },
+    {
+      "epoch": 2.766883927333241,
+      "grad_norm": 0.3852544128894806,
+      "learning_rate": 2.5781249999999996e-05,
+      "loss": 0.6018,
+      "step": 2494
+    },
+    {
+      "epoch": 2.7679933435029813,
+      "grad_norm": 0.3930635452270508,
+      "learning_rate": 2.5657894736842103e-05,
+      "loss": 0.5023,
+      "step": 2495
+    },
+    {
+      "epoch": 2.7691027596727222,
+      "grad_norm": 0.34579432010650635,
+      "learning_rate": 2.553453947368421e-05,
+      "loss": 0.5322,
+      "step": 2496
+    },
+    {
+      "epoch": 2.770212175842463,
+      "grad_norm": 0.38263997435569763,
+      "learning_rate": 2.5411184210526317e-05,
+      "loss": 0.3458,
+      "step": 2497
+    },
+    {
+      "epoch": 2.7713215920122036,
+      "grad_norm": 0.34970882534980774,
+      "learning_rate": 2.5287828947368417e-05,
+      "loss": 0.4036,
+      "step": 2498
+    },
+    {
+      "epoch": 2.772431008181944,
+      "grad_norm": 0.36101511120796204,
+      "learning_rate": 2.5164473684210523e-05,
+      "loss": 0.2912,
+      "step": 2499
+    },
+    {
+      "epoch": 2.773540424351685,
+      "grad_norm": 0.39122408628463745,
+      "learning_rate": 2.504111842105263e-05,
+      "loss": 0.3174,
+      "step": 2500
+    },
+    {
+      "epoch": 2.7746498405214255,
+      "grad_norm": 0.3180815875530243,
+      "learning_rate": 2.4917763157894737e-05,
+      "loss": 0.4215,
+      "step": 2501
+    },
+    {
+      "epoch": 2.7757592566911664,
+      "grad_norm": 0.4079410433769226,
+      "learning_rate": 2.479440789473684e-05,
+      "loss": 0.5662,
+      "step": 2502
+    },
+    {
+      "epoch": 2.776868672860907,
+      "grad_norm": 0.35038089752197266,
+      "learning_rate": 2.4671052631578947e-05,
+      "loss": 0.4216,
+      "step": 2503
+    },
+    {
+      "epoch": 2.777978089030648,
+      "grad_norm": 0.34112969040870667,
+      "learning_rate": 2.454769736842105e-05,
+      "loss": 0.3627,
+      "step": 2504
+    },
+    {
+      "epoch": 2.7790875052003883,
+      "grad_norm": 0.38411781191825867,
+      "learning_rate": 2.4424342105263153e-05,
+      "loss": 0.3333,
+      "step": 2505
+    },
+    {
+      "epoch": 2.7801969213701287,
+      "grad_norm": 0.4743475615978241,
+      "learning_rate": 2.430098684210526e-05,
+      "loss": 0.4561,
+      "step": 2506
+    },
+    {
+      "epoch": 2.7813063375398697,
+      "grad_norm": 0.3875446915626526,
+      "learning_rate": 2.4177631578947367e-05,
+      "loss": 0.5246,
+      "step": 2507
+    },
+    {
+      "epoch": 2.7824157537096106,
+      "grad_norm": 0.3834385573863983,
+      "learning_rate": 2.4054276315789474e-05,
+      "loss": 0.4338,
+      "step": 2508
+    },
+    {
+      "epoch": 2.783525169879351,
+      "grad_norm": 0.4199206233024597,
+      "learning_rate": 2.3930921052631577e-05,
+      "loss": 0.3568,
+      "step": 2509
+    },
+    {
+      "epoch": 2.7846345860490915,
+      "grad_norm": 0.3700524568557739,
+      "learning_rate": 2.380756578947368e-05,
+      "loss": 0.4707,
+      "step": 2510
+    },
+    {
+      "epoch": 2.7857440022188324,
+      "grad_norm": 0.3416059911251068,
+      "learning_rate": 2.3684210526315787e-05,
+      "loss": 0.3664,
+      "step": 2511
+    },
+    {
+      "epoch": 2.786853418388573,
+      "grad_norm": 0.39632654190063477,
+      "learning_rate": 2.356085526315789e-05,
+      "loss": 0.5128,
+      "step": 2512
+    },
+    {
+      "epoch": 2.787962834558314,
+      "grad_norm": 0.5743038654327393,
+      "learning_rate": 2.3437499999999997e-05,
+      "loss": 0.4739,
+      "step": 2513
+    },
+    {
+      "epoch": 2.7890722507280543,
+      "grad_norm": 0.3902910649776459,
+      "learning_rate": 2.3314144736842104e-05,
+      "loss": 0.4639,
+      "step": 2514
+    },
+    {
+      "epoch": 2.7901816668977952,
+      "grad_norm": 0.3890102803707123,
+      "learning_rate": 2.319078947368421e-05,
+      "loss": 0.4134,
+      "step": 2515
+    },
+    {
+      "epoch": 2.7912910830675357,
+      "grad_norm": 0.5704060196876526,
+      "learning_rate": 2.3067434210526317e-05,
+      "loss": 0.4693,
+      "step": 2516
+    },
+    {
+      "epoch": 2.792400499237276,
+      "grad_norm": 0.37423625588417053,
+      "learning_rate": 2.2944078947368417e-05,
+      "loss": 0.3357,
+      "step": 2517
+    },
+    {
+      "epoch": 2.793509915407017,
+      "grad_norm": 0.33532601594924927,
+      "learning_rate": 2.2820723684210524e-05,
+      "loss": 0.4696,
+      "step": 2518
+    },
+    {
+      "epoch": 2.7946193315767576,
+      "grad_norm": 0.3831970989704132,
+      "learning_rate": 2.269736842105263e-05,
+      "loss": 0.3145,
+      "step": 2519
+    },
+    {
+      "epoch": 2.7957287477464985,
+      "grad_norm": 0.312404602766037,
+      "learning_rate": 2.2574013157894734e-05,
+      "loss": 0.408,
+      "step": 2520
+    },
+    {
+      "epoch": 2.796838163916239,
+      "grad_norm": 0.396106094121933,
+      "learning_rate": 2.245065789473684e-05,
+      "loss": 0.33,
+      "step": 2521
+    },
+    {
+      "epoch": 2.79794758008598,
+      "grad_norm": 0.3440202474594116,
+      "learning_rate": 2.2327302631578947e-05,
+      "loss": 0.3087,
+      "step": 2522
+    },
+    {
+      "epoch": 2.7990569962557204,
+      "grad_norm": 0.3287065625190735,
+      "learning_rate": 2.2203947368421047e-05,
+      "loss": 0.2821,
+      "step": 2523
+    },
+    {
+      "epoch": 2.800166412425461,
+      "grad_norm": 0.5490508675575256,
+      "learning_rate": 2.2080592105263154e-05,
+      "loss": 0.4684,
+      "step": 2524
+    },
+    {
+      "epoch": 2.8012758285952017,
+      "grad_norm": 0.560986340045929,
+      "learning_rate": 2.195723684210526e-05,
+      "loss": 0.2953,
+      "step": 2525
+    },
+    {
+      "epoch": 2.8023852447649427,
+      "grad_norm": 0.4068681001663208,
+      "learning_rate": 2.1833881578947368e-05,
+      "loss": 0.4226,
+      "step": 2526
+    },
+    {
+      "epoch": 2.803494660934683,
+      "grad_norm": 0.3921424448490143,
+      "learning_rate": 2.1710526315789474e-05,
+      "loss": 0.3098,
+      "step": 2527
+    },
+    {
+      "epoch": 2.8046040771044236,
+      "grad_norm": 0.40971678495407104,
+      "learning_rate": 2.1587171052631578e-05,
+      "loss": 0.3498,
+      "step": 2528
+    },
+    {
+      "epoch": 2.8057134932741645,
+      "grad_norm": 0.2933211326599121,
+      "learning_rate": 2.146381578947368e-05,
+      "loss": 0.4814,
+      "step": 2529
+    },
+    {
+      "epoch": 2.806822909443905,
+      "grad_norm": 0.5020395517349243,
+      "learning_rate": 2.1340460526315788e-05,
+      "loss": 0.4159,
+      "step": 2530
+    },
+    {
+      "epoch": 2.807932325613646,
+      "grad_norm": 0.3937400281429291,
+      "learning_rate": 2.121710526315789e-05,
+      "loss": 0.3239,
+      "step": 2531
+    },
+    {
+      "epoch": 2.8090417417833864,
+      "grad_norm": 0.47247281670570374,
+      "learning_rate": 2.1093749999999998e-05,
+      "loss": 0.4367,
+      "step": 2532
+    },
+    {
+      "epoch": 2.8101511579531273,
+      "grad_norm": 0.44965869188308716,
+      "learning_rate": 2.0970394736842104e-05,
+      "loss": 0.3051,
+      "step": 2533
+    },
+    {
+      "epoch": 2.811260574122868,
+      "grad_norm": 0.5918648838996887,
+      "learning_rate": 2.084703947368421e-05,
+      "loss": 0.3252,
+      "step": 2534
+    },
+    {
+      "epoch": 2.8123699902926083,
+      "grad_norm": 0.4035295844078064,
+      "learning_rate": 2.0723684210526315e-05,
+      "loss": 0.3419,
+      "step": 2535
+    },
+    {
+      "epoch": 2.813479406462349,
+      "grad_norm": 0.43882495164871216,
+      "learning_rate": 2.0600328947368418e-05,
+      "loss": 0.3963,
+      "step": 2536
+    },
+    {
+      "epoch": 2.81458882263209,
+      "grad_norm": 0.3037029206752777,
+      "learning_rate": 2.0476973684210525e-05,
+      "loss": 0.2939,
+      "step": 2537
+    },
+    {
+      "epoch": 2.8156982388018306,
+      "grad_norm": 0.42646703124046326,
+      "learning_rate": 2.0353618421052628e-05,
+      "loss": 0.376,
+      "step": 2538
+    },
+    {
+      "epoch": 2.816807654971571,
+      "grad_norm": 0.3901432156562805,
+      "learning_rate": 2.0230263157894735e-05,
+      "loss": 0.3588,
+      "step": 2539
+    },
+    {
+      "epoch": 2.817917071141312,
+      "grad_norm": 0.4185813367366791,
+      "learning_rate": 2.010690789473684e-05,
+      "loss": 0.4512,
+      "step": 2540
+    },
+    {
+      "epoch": 2.8190264873110524,
+      "grad_norm": 0.46253278851509094,
+      "learning_rate": 1.9983552631578948e-05,
+      "loss": 0.3559,
+      "step": 2541
+    },
+    {
+      "epoch": 2.8201359034807933,
+      "grad_norm": 0.49084243178367615,
+      "learning_rate": 1.9860197368421048e-05,
+      "loss": 0.4087,
+      "step": 2542
+    },
+    {
+      "epoch": 2.821245319650534,
+      "grad_norm": 0.4082900285720825,
+      "learning_rate": 1.9736842105263155e-05,
+      "loss": 0.3204,
+      "step": 2543
+    },
+    {
+      "epoch": 2.8223547358202747,
+      "grad_norm": 0.4727761149406433,
+      "learning_rate": 1.961348684210526e-05,
+      "loss": 0.5809,
+      "step": 2544
+    },
+    {
+      "epoch": 2.823464151990015,
+      "grad_norm": 0.37872323393821716,
+      "learning_rate": 1.9490131578947368e-05,
+      "loss": 0.2695,
+      "step": 2545
+    },
+    {
+      "epoch": 2.8245735681597557,
+      "grad_norm": 0.4297656714916229,
+      "learning_rate": 1.936677631578947e-05,
+      "loss": 0.2949,
+      "step": 2546
+    },
+    {
+      "epoch": 2.8256829843294966,
+      "grad_norm": 0.5141124725341797,
+      "learning_rate": 1.9243421052631578e-05,
+      "loss": 0.475,
+      "step": 2547
+    },
+    {
+      "epoch": 2.8267924004992375,
+      "grad_norm": 0.4390595853328705,
+      "learning_rate": 1.912006578947368e-05,
+      "loss": 0.4907,
+      "step": 2548
+    },
+    {
+      "epoch": 2.827901816668978,
+      "grad_norm": 0.5395392179489136,
+      "learning_rate": 1.8996710526315785e-05,
+      "loss": 0.3074,
+      "step": 2549
+    },
+    {
+      "epoch": 2.8290112328387185,
+      "grad_norm": 0.5195388197898865,
+      "learning_rate": 1.887335526315789e-05,
+      "loss": 0.3901,
+      "step": 2550
+    },
+    {
+      "epoch": 2.8301206490084594,
+      "grad_norm": 0.5643588304519653,
+      "learning_rate": 1.875e-05,
+      "loss": 0.528,
+      "step": 2551
+    },
+    {
+      "epoch": 2.8312300651782,
+      "grad_norm": 0.6140674352645874,
+      "learning_rate": 1.8626644736842105e-05,
+      "loss": 0.404,
+      "step": 2552
+    },
+    {
+      "epoch": 2.832339481347941,
+      "grad_norm": 1.3323380947113037,
+      "learning_rate": 1.850328947368421e-05,
+      "loss": 0.1839,
+      "step": 2553
+    },
+    {
+      "epoch": 2.8334488975176813,
+      "grad_norm": 0.3624992072582245,
+      "learning_rate": 1.8379934210526315e-05,
+      "loss": 0.5518,
+      "step": 2554
+    },
+    {
+      "epoch": 2.834558313687422,
+      "grad_norm": 0.493804395198822,
+      "learning_rate": 1.825657894736842e-05,
+      "loss": 0.2108,
+      "step": 2555
+    },
+    {
+      "epoch": 2.8356677298571626,
+      "grad_norm": 0.41342535614967346,
+      "learning_rate": 1.8133223684210525e-05,
+      "loss": 0.5485,
+      "step": 2556
+    },
+    {
+      "epoch": 2.836777146026903,
+      "grad_norm": 0.37890028953552246,
+      "learning_rate": 1.800986842105263e-05,
+      "loss": 0.2336,
+      "step": 2557
+    },
+    {
+      "epoch": 2.837886562196644,
+      "grad_norm": 0.5276500582695007,
+      "learning_rate": 1.7886513157894735e-05,
+      "loss": 0.3295,
+      "step": 2558
+    },
+    {
+      "epoch": 2.838995978366385,
+      "grad_norm": 0.8836992979049683,
+      "learning_rate": 1.776315789473684e-05,
+      "loss": 0.4195,
+      "step": 2559
+    },
+    {
+      "epoch": 2.8401053945361254,
+      "grad_norm": 0.36474689841270447,
+      "learning_rate": 1.7639802631578945e-05,
+      "loss": 0.5335,
+      "step": 2560
+    },
+    {
+      "epoch": 2.841214810705866,
+      "grad_norm": 0.3125799298286438,
+      "learning_rate": 1.7516447368421052e-05,
+      "loss": 0.3293,
+      "step": 2561
+    },
+    {
+      "epoch": 2.842324226875607,
+      "grad_norm": 0.4086284041404724,
+      "learning_rate": 1.739309210526316e-05,
+      "loss": 0.8532,
+      "step": 2562
+    },
+    {
+      "epoch": 2.8434336430453473,
+      "grad_norm": 0.5705865025520325,
+      "learning_rate": 1.7269736842105262e-05,
+      "loss": 0.7003,
+      "step": 2563
+    },
+    {
+      "epoch": 2.844543059215088,
+      "grad_norm": 0.3946170508861542,
+      "learning_rate": 1.714638157894737e-05,
+      "loss": 0.3741,
+      "step": 2564
+    },
+    {
+      "epoch": 2.8456524753848287,
+      "grad_norm": 0.5471513271331787,
+      "learning_rate": 1.7023026315789472e-05,
+      "loss": 0.3288,
+      "step": 2565
+    },
+    {
+      "epoch": 2.8467618915545696,
+      "grad_norm": 0.43899303674697876,
+      "learning_rate": 1.6899671052631575e-05,
+      "loss": 0.5189,
+      "step": 2566
+    },
+    {
+      "epoch": 2.84787130772431,
+      "grad_norm": 0.6397573351860046,
+      "learning_rate": 1.6776315789473682e-05,
+      "loss": 0.3055,
+      "step": 2567
+    },
+    {
+      "epoch": 2.8489807238940505,
+      "grad_norm": 0.3993845582008362,
+      "learning_rate": 1.665296052631579e-05,
+      "loss": 0.3922,
+      "step": 2568
+    },
+    {
+      "epoch": 2.8500901400637915,
+      "grad_norm": 0.27621200680732727,
+      "learning_rate": 1.6529605263157892e-05,
+      "loss": 0.3626,
+      "step": 2569
+    },
+    {
+      "epoch": 2.851199556233532,
+      "grad_norm": 0.721236526966095,
+      "learning_rate": 1.640625e-05,
+      "loss": 0.5003,
+      "step": 2570
+    },
+    {
+      "epoch": 2.852308972403273,
+      "grad_norm": 0.46138882637023926,
+      "learning_rate": 1.6282894736842106e-05,
+      "loss": 0.3559,
+      "step": 2571
+    },
+    {
+      "epoch": 2.8534183885730133,
+      "grad_norm": 0.3132235109806061,
+      "learning_rate": 1.615953947368421e-05,
+      "loss": 0.3119,
+      "step": 2572
+    },
+    {
+      "epoch": 2.8545278047427542,
+      "grad_norm": 0.4404788911342621,
+      "learning_rate": 1.6036184210526316e-05,
+      "loss": 0.3208,
+      "step": 2573
+    },
+    {
+      "epoch": 2.8556372209124947,
+      "grad_norm": 0.5489991307258606,
+      "learning_rate": 1.591282894736842e-05,
+      "loss": 0.3501,
+      "step": 2574
+    },
+    {
+      "epoch": 2.856746637082235,
+      "grad_norm": 0.5425270199775696,
+      "learning_rate": 1.5789473684210522e-05,
+      "loss": 0.3652,
+      "step": 2575
+    },
+    {
+      "epoch": 2.857856053251976,
+      "grad_norm": 0.3924858570098877,
+      "learning_rate": 1.566611842105263e-05,
+      "loss": 0.3799,
+      "step": 2576
+    },
+    {
+      "epoch": 2.858965469421717,
+      "grad_norm": 0.381849080324173,
+      "learning_rate": 1.5542763157894736e-05,
+      "loss": 0.4581,
+      "step": 2577
+    },
+    {
+      "epoch": 2.8600748855914575,
+      "grad_norm": 0.38210979104042053,
+      "learning_rate": 1.541940789473684e-05,
+      "loss": 0.4339,
+      "step": 2578
+    },
+    {
+      "epoch": 2.861184301761198,
+      "grad_norm": 0.49273112416267395,
+      "learning_rate": 1.5296052631578946e-05,
+      "loss": 0.3847,
+      "step": 2579
+    },
+    {
+      "epoch": 2.862293717930939,
+      "grad_norm": 0.45563986897468567,
+      "learning_rate": 1.5172697368421051e-05,
+      "loss": 0.3684,
+      "step": 2580
+    },
+    {
+      "epoch": 2.8634031341006794,
+      "grad_norm": 0.5015951991081238,
+      "learning_rate": 1.5049342105263158e-05,
+      "loss": 0.5179,
+      "step": 2581
+    },
+    {
+      "epoch": 2.8645125502704203,
+      "grad_norm": 0.6307615041732788,
+      "learning_rate": 1.4925986842105261e-05,
+      "loss": 0.486,
+      "step": 2582
+    },
+    {
+      "epoch": 2.8656219664401608,
+      "grad_norm": 0.36506178975105286,
+      "learning_rate": 1.4802631578947368e-05,
+      "loss": 0.5584,
+      "step": 2583
+    },
+    {
+      "epoch": 2.8667313826099017,
+      "grad_norm": 0.39168620109558105,
+      "learning_rate": 1.4679276315789473e-05,
+      "loss": 0.3033,
+      "step": 2584
+    },
+    {
+      "epoch": 2.867840798779642,
+      "grad_norm": 0.5694892406463623,
+      "learning_rate": 1.4555921052631578e-05,
+      "loss": 0.4762,
+      "step": 2585
+    },
+    {
+      "epoch": 2.8689502149493826,
+      "grad_norm": 0.3743266463279724,
+      "learning_rate": 1.4432565789473683e-05,
+      "loss": 0.2169,
+      "step": 2586
+    },
+    {
+      "epoch": 2.8700596311191235,
+      "grad_norm": 0.32552504539489746,
+      "learning_rate": 1.430921052631579e-05,
+      "loss": 0.5623,
+      "step": 2587
+    },
+    {
+      "epoch": 2.8711690472888645,
+      "grad_norm": 0.4949817359447479,
+      "learning_rate": 1.4185855263157893e-05,
+      "loss": 0.5808,
+      "step": 2588
+    },
+    {
+      "epoch": 2.872278463458605,
+      "grad_norm": 0.5175449848175049,
+      "learning_rate": 1.40625e-05,
+      "loss": 0.4294,
+      "step": 2589
+    },
+    {
+      "epoch": 2.8733878796283454,
+      "grad_norm": 0.4763440191745758,
+      "learning_rate": 1.3939144736842105e-05,
+      "loss": 0.3579,
+      "step": 2590
+    },
+    {
+      "epoch": 2.8744972957980863,
+      "grad_norm": 0.3480813205242157,
+      "learning_rate": 1.3815789473684208e-05,
+      "loss": 0.3258,
+      "step": 2591
+    },
+    {
+      "epoch": 2.875606711967827,
+      "grad_norm": 0.33044132590293884,
+      "learning_rate": 1.3692434210526315e-05,
+      "loss": 0.4225,
+      "step": 2592
+    },
+    {
+      "epoch": 2.8767161281375677,
+      "grad_norm": 0.4161834418773651,
+      "learning_rate": 1.356907894736842e-05,
+      "loss": 0.3972,
+      "step": 2593
+    },
+    {
+      "epoch": 2.877825544307308,
+      "grad_norm": 0.42129024863243103,
+      "learning_rate": 1.3445723684210525e-05,
+      "loss": 0.4956,
+      "step": 2594
+    },
+    {
+      "epoch": 2.878934960477049,
+      "grad_norm": 0.3572520613670349,
+      "learning_rate": 1.332236842105263e-05,
+      "loss": 0.7549,
+      "step": 2595
+    },
+    {
+      "epoch": 2.8800443766467896,
+      "grad_norm": 0.3236415684223175,
+      "learning_rate": 1.3199013157894737e-05,
+      "loss": 0.7084,
+      "step": 2596
+    },
+    {
+      "epoch": 2.88115379281653,
+      "grad_norm": 0.42282864451408386,
+      "learning_rate": 1.307565789473684e-05,
+      "loss": 0.439,
+      "step": 2597
+    },
+    {
+      "epoch": 2.882263208986271,
+      "grad_norm": 0.3626263737678528,
+      "learning_rate": 1.2952302631578947e-05,
+      "loss": 0.36,
+      "step": 2598
+    },
+    {
+      "epoch": 2.883372625156012,
+      "grad_norm": 0.4625903367996216,
+      "learning_rate": 1.2828947368421052e-05,
+      "loss": 0.3951,
+      "step": 2599
+    },
+    {
+      "epoch": 2.8844820413257524,
+      "grad_norm": 0.30453893542289734,
+      "learning_rate": 1.2705592105263158e-05,
+      "loss": 0.3333,
+      "step": 2600
+    },
+    {
+      "epoch": 2.885591457495493,
+      "grad_norm": 0.43295082449913025,
+      "learning_rate": 1.2582236842105262e-05,
+      "loss": 0.4729,
+      "step": 2601
+    },
+    {
+      "epoch": 2.8867008736652338,
+      "grad_norm": 0.3720959424972534,
+      "learning_rate": 1.2458881578947368e-05,
+      "loss": 0.3683,
+      "step": 2602
+    },
+    {
+      "epoch": 2.8878102898349742,
+      "grad_norm": 0.2515357434749603,
+      "learning_rate": 1.2335526315789473e-05,
+      "loss": 0.2588,
+      "step": 2603
+    },
+    {
+      "epoch": 2.888919706004715,
+      "grad_norm": 0.41880831122398376,
+      "learning_rate": 1.2212171052631577e-05,
+      "loss": 0.4133,
+      "step": 2604
+    },
+    {
+      "epoch": 2.8900291221744556,
+      "grad_norm": 0.4307885766029358,
+      "learning_rate": 1.2088815789473683e-05,
+      "loss": 0.3466,
+      "step": 2605
+    },
+    {
+      "epoch": 2.8911385383441965,
+      "grad_norm": 0.5042724609375,
+      "learning_rate": 1.1965460526315788e-05,
+      "loss": 0.3241,
+      "step": 2606
+    },
+    {
+      "epoch": 2.892247954513937,
+      "grad_norm": 0.4066576063632965,
+      "learning_rate": 1.1842105263157894e-05,
+      "loss": 0.4199,
+      "step": 2607
+    },
+    {
+      "epoch": 2.8933573706836775,
+      "grad_norm": 0.43350422382354736,
+      "learning_rate": 1.1718749999999999e-05,
+      "loss": 0.4903,
+      "step": 2608
+    },
+    {
+      "epoch": 2.8944667868534184,
+      "grad_norm": 0.5719215273857117,
+      "learning_rate": 1.1595394736842105e-05,
+      "loss": 0.4011,
+      "step": 2609
+    },
+    {
+      "epoch": 2.8955762030231593,
+      "grad_norm": 0.44598355889320374,
+      "learning_rate": 1.1472039473684209e-05,
+      "loss": 0.3502,
+      "step": 2610
+    },
+    {
+      "epoch": 2.8966856191929,
+      "grad_norm": 0.3669666051864624,
+      "learning_rate": 1.1348684210526315e-05,
+      "loss": 0.5565,
+      "step": 2611
+    },
+    {
+      "epoch": 2.8977950353626403,
+      "grad_norm": 0.3861677348613739,
+      "learning_rate": 1.122532894736842e-05,
+      "loss": 0.4093,
+      "step": 2612
+    },
+    {
+      "epoch": 2.898904451532381,
+      "grad_norm": 0.3836744427680969,
+      "learning_rate": 1.1101973684210524e-05,
+      "loss": 0.5552,
+      "step": 2613
+    },
+    {
+      "epoch": 2.9000138677021217,
+      "grad_norm": 0.5049236416816711,
+      "learning_rate": 1.097861842105263e-05,
+      "loss": 0.3947,
+      "step": 2614
+    },
+    {
+      "epoch": 2.9011232838718626,
+      "grad_norm": 0.3340136706829071,
+      "learning_rate": 1.0855263157894737e-05,
+      "loss": 0.3379,
+      "step": 2615
+    },
+    {
+      "epoch": 2.902232700041603,
+      "grad_norm": 0.41469404101371765,
+      "learning_rate": 1.073190789473684e-05,
+      "loss": 0.4554,
+      "step": 2616
+    },
+    {
+      "epoch": 2.903342116211344,
+      "grad_norm": 0.34765610098838806,
+      "learning_rate": 1.0608552631578946e-05,
+      "loss": 0.287,
+      "step": 2617
+    },
+    {
+      "epoch": 2.9044515323810844,
+      "grad_norm": 0.3712421655654907,
+      "learning_rate": 1.0485197368421052e-05,
+      "loss": 0.3694,
+      "step": 2618
+    },
+    {
+      "epoch": 2.905560948550825,
+      "grad_norm": 0.37911391258239746,
+      "learning_rate": 1.0361842105263157e-05,
+      "loss": 0.3963,
+      "step": 2619
+    },
+    {
+      "epoch": 2.906670364720566,
+      "grad_norm": 0.3563274145126343,
+      "learning_rate": 1.0238486842105262e-05,
+      "loss": 0.4382,
+      "step": 2620
+    },
+    {
+      "epoch": 2.9077797808903068,
+      "grad_norm": 0.535892128944397,
+      "learning_rate": 1.0115131578947367e-05,
+      "loss": 0.4363,
+      "step": 2621
+    },
+    {
+      "epoch": 2.9088891970600472,
+      "grad_norm": 0.4795096218585968,
+      "learning_rate": 9.991776315789474e-06,
+      "loss": 0.5396,
+      "step": 2622
+    },
+    {
+      "epoch": 2.9099986132297877,
+      "grad_norm": 0.5538339614868164,
+      "learning_rate": 9.868421052631577e-06,
+      "loss": 0.3635,
+      "step": 2623
+    },
+    {
+      "epoch": 2.9111080293995286,
+      "grad_norm": 0.3773338198661804,
+      "learning_rate": 9.745065789473684e-06,
+      "loss": 0.4025,
+      "step": 2624
+    },
+    {
+      "epoch": 2.912217445569269,
+      "grad_norm": 0.6125039458274841,
+      "learning_rate": 9.621710526315789e-06,
+      "loss": 0.4946,
+      "step": 2625
+    },
+    {
+      "epoch": 2.91332686173901,
+      "grad_norm": 0.3754761219024658,
+      "learning_rate": 9.498355263157892e-06,
+      "loss": 0.3312,
+      "step": 2626
+    },
+    {
+      "epoch": 2.9144362779087505,
+      "grad_norm": 0.5250300168991089,
+      "learning_rate": 9.375e-06,
+      "loss": 0.4738,
+      "step": 2627
+    },
+    {
+      "epoch": 2.9155456940784914,
+      "grad_norm": 0.42315995693206787,
+      "learning_rate": 9.251644736842104e-06,
+      "loss": 0.3803,
+      "step": 2628
+    },
+    {
+      "epoch": 2.916655110248232,
+      "grad_norm": 0.5445030927658081,
+      "learning_rate": 9.12828947368421e-06,
+      "loss": 0.5786,
+      "step": 2629
+    },
+    {
+      "epoch": 2.9177645264179723,
+      "grad_norm": 0.6609618067741394,
+      "learning_rate": 9.004934210526314e-06,
+      "loss": 0.386,
+      "step": 2630
+    },
+    {
+      "epoch": 2.9188739425877133,
+      "grad_norm": 0.36153900623321533,
+      "learning_rate": 8.88157894736842e-06,
+      "loss": 0.3705,
+      "step": 2631
+    },
+    {
+      "epoch": 2.9199833587574537,
+      "grad_norm": 0.39644429087638855,
+      "learning_rate": 8.758223684210526e-06,
+      "loss": 0.4203,
+      "step": 2632
+    },
+    {
+      "epoch": 2.9210927749271947,
+      "grad_norm": 0.44478243589401245,
+      "learning_rate": 8.634868421052631e-06,
+      "loss": 0.5676,
+      "step": 2633
+    },
+    {
+      "epoch": 2.922202191096935,
+      "grad_norm": 0.3410852551460266,
+      "learning_rate": 8.511513157894736e-06,
+      "loss": 0.29,
+      "step": 2634
+    },
+    {
+      "epoch": 2.923311607266676,
+      "grad_norm": 0.38531428575515747,
+      "learning_rate": 8.388157894736841e-06,
+      "loss": 0.4727,
+      "step": 2635
+    },
+    {
+      "epoch": 2.9244210234364165,
+      "grad_norm": 0.40856409072875977,
+      "learning_rate": 8.264802631578946e-06,
+      "loss": 0.2994,
+      "step": 2636
+    },
+    {
+      "epoch": 2.925530439606157,
+      "grad_norm": 0.3600262999534607,
+      "learning_rate": 8.141447368421053e-06,
+      "loss": 0.3595,
+      "step": 2637
+    },
+    {
+      "epoch": 2.926639855775898,
+      "grad_norm": 0.4171159863471985,
+      "learning_rate": 8.018092105263158e-06,
+      "loss": 0.2841,
+      "step": 2638
+    },
+    {
+      "epoch": 2.927749271945639,
+      "grad_norm": 0.35529500246047974,
+      "learning_rate": 7.894736842105261e-06,
+      "loss": 0.278,
+      "step": 2639
+    },
+    {
+      "epoch": 2.9288586881153793,
+      "grad_norm": 0.3548045754432678,
+      "learning_rate": 7.771381578947368e-06,
+      "loss": 0.6584,
+      "step": 2640
+    },
+    {
+      "epoch": 2.92996810428512,
+      "grad_norm": 0.8075834512710571,
+      "learning_rate": 7.648026315789473e-06,
+      "loss": 0.2551,
+      "step": 2641
+    },
+    {
+      "epoch": 2.9310775204548607,
+      "grad_norm": 0.4155935049057007,
+      "learning_rate": 7.524671052631579e-06,
+      "loss": 0.3003,
+      "step": 2642
+    },
+    {
+      "epoch": 2.932186936624601,
+      "grad_norm": 0.4321109652519226,
+      "learning_rate": 7.401315789473684e-06,
+      "loss": 0.343,
+      "step": 2643
+    },
+    {
+      "epoch": 2.933296352794342,
+      "grad_norm": 0.3877204358577728,
+      "learning_rate": 7.277960526315789e-06,
+      "loss": 0.3259,
+      "step": 2644
+    },
+    {
+      "epoch": 2.9344057689640826,
+      "grad_norm": 0.30750367045402527,
+      "learning_rate": 7.154605263157895e-06,
+      "loss": 0.4062,
+      "step": 2645
+    },
+    {
+      "epoch": 2.9355151851338235,
+      "grad_norm": 0.6012676954269409,
+      "learning_rate": 7.03125e-06,
+      "loss": 0.4992,
+      "step": 2646
+    },
+    {
+      "epoch": 2.936624601303564,
+      "grad_norm": 0.3799775242805481,
+      "learning_rate": 6.907894736842104e-06,
+      "loss": 0.4411,
+      "step": 2647
+    },
+    {
+      "epoch": 2.9377340174733044,
+      "grad_norm": 0.49381592869758606,
+      "learning_rate": 6.78453947368421e-06,
+      "loss": 0.3694,
+      "step": 2648
+    },
+    {
+      "epoch": 2.9388434336430453,
+      "grad_norm": 0.7122082114219666,
+      "learning_rate": 6.661184210526315e-06,
+      "loss": 0.4398,
+      "step": 2649
+    },
+    {
+      "epoch": 2.9399528498127863,
+      "grad_norm": 0.48016512393951416,
+      "learning_rate": 6.53782894736842e-06,
+      "loss": 0.3111,
+      "step": 2650
+    },
+    {
+      "epoch": 2.9410622659825267,
+      "grad_norm": 0.449733167886734,
+      "learning_rate": 6.414473684210526e-06,
+      "loss": 0.3113,
+      "step": 2651
+    },
+    {
+      "epoch": 2.942171682152267,
+      "grad_norm": 0.33274686336517334,
+      "learning_rate": 6.291118421052631e-06,
+      "loss": 0.4131,
+      "step": 2652
+    },
+    {
+      "epoch": 2.943281098322008,
+      "grad_norm": 0.37493860721588135,
+      "learning_rate": 6.167763157894737e-06,
+      "loss": 0.2557,
+      "step": 2653
+    },
+    {
+      "epoch": 2.9443905144917486,
+      "grad_norm": 0.37679123878479004,
+      "learning_rate": 6.044407894736842e-06,
+      "loss": 0.4252,
+      "step": 2654
+    },
+    {
+      "epoch": 2.9454999306614895,
+      "grad_norm": 0.40936678647994995,
+      "learning_rate": 5.921052631578947e-06,
+      "loss": 0.4795,
+      "step": 2655
+    },
+    {
+      "epoch": 2.94660934683123,
+      "grad_norm": 0.4506048560142517,
+      "learning_rate": 5.797697368421053e-06,
+      "loss": 0.5137,
+      "step": 2656
+    },
+    {
+      "epoch": 2.947718763000971,
+      "grad_norm": 0.39844781160354614,
+      "learning_rate": 5.674342105263158e-06,
+      "loss": 0.3271,
+      "step": 2657
+    },
+    {
+      "epoch": 2.9488281791707114,
+      "grad_norm": 0.39184948801994324,
+      "learning_rate": 5.550986842105262e-06,
+      "loss": 0.3885,
+      "step": 2658
+    },
+    {
+      "epoch": 2.949937595340452,
+      "grad_norm": 0.3266217112541199,
+      "learning_rate": 5.4276315789473686e-06,
+      "loss": 0.3644,
+      "step": 2659
+    },
+    {
+      "epoch": 2.9510470115101928,
+      "grad_norm": 0.4527483284473419,
+      "learning_rate": 5.304276315789473e-06,
+      "loss": 0.3945,
+      "step": 2660
+    },
+    {
+      "epoch": 2.9521564276799337,
+      "grad_norm": 0.5377230644226074,
+      "learning_rate": 5.180921052631579e-06,
+      "loss": 0.458,
+      "step": 2661
+    },
+    {
+      "epoch": 2.953265843849674,
+      "grad_norm": 0.39653611183166504,
+      "learning_rate": 5.057565789473684e-06,
+      "loss": 0.3541,
+      "step": 2662
+    },
+    {
+      "epoch": 2.9543752600194146,
+      "grad_norm": 0.2904629409313202,
+      "learning_rate": 4.934210526315789e-06,
+      "loss": 0.1772,
+      "step": 2663
+    },
+    {
+      "epoch": 2.9554846761891556,
+      "grad_norm": 0.3323127329349518,
+      "learning_rate": 4.8108552631578946e-06,
+      "loss": 0.5255,
+      "step": 2664
+    },
+    {
+      "epoch": 2.956594092358896,
+      "grad_norm": 0.4561123847961426,
+      "learning_rate": 4.6875e-06,
+      "loss": 0.3503,
+      "step": 2665
+    },
+    {
+      "epoch": 2.957703508528637,
+      "grad_norm": 0.37283483147621155,
+      "learning_rate": 4.564144736842105e-06,
+      "loss": 0.4867,
+      "step": 2666
+    },
+    {
+      "epoch": 2.9588129246983774,
+      "grad_norm": 0.5536327362060547,
+      "learning_rate": 4.44078947368421e-06,
+      "loss": 0.3599,
+      "step": 2667
+    },
+    {
+      "epoch": 2.9599223408681183,
+      "grad_norm": 0.42849066853523254,
+      "learning_rate": 4.3174342105263155e-06,
+      "loss": 0.4078,
+      "step": 2668
+    },
+    {
+      "epoch": 2.961031757037859,
+      "grad_norm": 0.4770295023918152,
+      "learning_rate": 4.1940789473684206e-06,
+      "loss": 0.5948,
+      "step": 2669
+    },
+    {
+      "epoch": 2.9621411732075993,
+      "grad_norm": 0.4041629433631897,
+      "learning_rate": 4.0707236842105264e-06,
+      "loss": 0.3036,
+      "step": 2670
+    },
+    {
+      "epoch": 2.96325058937734,
+      "grad_norm": 0.4390498697757721,
+      "learning_rate": 3.947368421052631e-06,
+      "loss": 0.373,
+      "step": 2671
+    },
+    {
+      "epoch": 2.964360005547081,
+      "grad_norm": 0.31613367795944214,
+      "learning_rate": 3.8240131578947365e-06,
+      "loss": 0.3119,
+      "step": 2672
+    },
+    {
+      "epoch": 2.9654694217168216,
+      "grad_norm": 0.4405744969844818,
+      "learning_rate": 3.700657894736842e-06,
+      "loss": 0.4571,
+      "step": 2673
+    },
+    {
+      "epoch": 2.966578837886562,
+      "grad_norm": 0.4392789900302887,
+      "learning_rate": 3.5773026315789474e-06,
+      "loss": 0.4928,
+      "step": 2674
+    },
+    {
+      "epoch": 2.967688254056303,
+      "grad_norm": 0.4602966606616974,
+      "learning_rate": 3.453947368421052e-06,
+      "loss": 0.345,
+      "step": 2675
+    },
+    {
+      "epoch": 2.9687976702260435,
+      "grad_norm": 0.447457492351532,
+      "learning_rate": 3.3305921052631574e-06,
+      "loss": 0.4556,
+      "step": 2676
+    },
+    {
+      "epoch": 2.9699070863957844,
+      "grad_norm": 0.38211768865585327,
+      "learning_rate": 3.207236842105263e-06,
+      "loss": 0.4022,
+      "step": 2677
+    },
+    {
+      "epoch": 2.971016502565525,
+      "grad_norm": 0.35372307896614075,
+      "learning_rate": 3.0838815789473684e-06,
+      "loss": 0.3703,
+      "step": 2678
+    },
+    {
+      "epoch": 2.9721259187352658,
+      "grad_norm": 0.44302472472190857,
+      "learning_rate": 2.9605263157894734e-06,
+      "loss": 0.4514,
+      "step": 2679
+    },
+    {
+      "epoch": 2.9732353349050062,
+      "grad_norm": 0.5012083053588867,
+      "learning_rate": 2.837171052631579e-06,
+      "loss": 0.4311,
+      "step": 2680
+    },
+    {
+      "epoch": 2.9743447510747467,
+      "grad_norm": 0.4370708763599396,
+      "learning_rate": 2.7138157894736843e-06,
+      "loss": 0.3342,
+      "step": 2681
+    },
+    {
+      "epoch": 2.9754541672444876,
+      "grad_norm": 0.37920454144477844,
+      "learning_rate": 2.5904605263157893e-06,
+      "loss": 0.5212,
+      "step": 2682
+    },
+    {
+      "epoch": 2.976563583414228,
+      "grad_norm": 0.2939812242984772,
+      "learning_rate": 2.4671052631578943e-06,
+      "loss": 0.3169,
+      "step": 2683
+    },
+    {
+      "epoch": 2.977672999583969,
+      "grad_norm": 0.354949414730072,
+      "learning_rate": 2.34375e-06,
+      "loss": 0.2588,
+      "step": 2684
+    },
+    {
+      "epoch": 2.9787824157537095,
+      "grad_norm": 0.334835022687912,
+      "learning_rate": 2.220394736842105e-06,
+      "loss": 0.4058,
+      "step": 2685
+    },
+    {
+      "epoch": 2.9798918319234504,
+      "grad_norm": 0.4206099212169647,
+      "learning_rate": 2.0970394736842103e-06,
+      "loss": 0.3528,
+      "step": 2686
+    },
+    {
+      "epoch": 2.981001248093191,
+      "grad_norm": 0.4646962583065033,
+      "learning_rate": 1.9736842105263153e-06,
+      "loss": 0.5455,
+      "step": 2687
+    },
+    {
+      "epoch": 2.9821106642629314,
+      "grad_norm": 0.4603005647659302,
+      "learning_rate": 1.850328947368421e-06,
+      "loss": 0.3451,
+      "step": 2688
+    },
+    {
+      "epoch": 2.9832200804326723,
+      "grad_norm": 0.4108537435531616,
+      "learning_rate": 1.726973684210526e-06,
+      "loss": 0.5365,
+      "step": 2689
+    },
+    {
+      "epoch": 2.984329496602413,
+      "grad_norm": 0.43144112825393677,
+      "learning_rate": 1.6036184210526314e-06,
+      "loss": 0.3908,
+      "step": 2690
+    },
+    {
+      "epoch": 2.9854389127721537,
+      "grad_norm": 0.5529698133468628,
+      "learning_rate": 1.4802631578947367e-06,
+      "loss": 0.3784,
+      "step": 2691
+    },
+    {
+      "epoch": 2.986548328941894,
+      "grad_norm": 0.37649285793304443,
+      "learning_rate": 1.3569078947368421e-06,
+      "loss": 0.4073,
+      "step": 2692
+    },
+    {
+      "epoch": 2.987657745111635,
+      "grad_norm": 0.3004691004753113,
+      "learning_rate": 1.2335526315789472e-06,
+      "loss": 0.4305,
+      "step": 2693
+    },
+    {
+      "epoch": 2.9887671612813755,
+      "grad_norm": 0.4551496207714081,
+      "learning_rate": 1.1101973684210524e-06,
+      "loss": 0.3474,
+      "step": 2694
+    },
+    {
+      "epoch": 2.9898765774511165,
+      "grad_norm": 0.3873107135295868,
+      "learning_rate": 9.868421052631577e-07,
+      "loss": 0.3239,
+      "step": 2695
+    },
+    {
+      "epoch": 2.990985993620857,
+      "grad_norm": 0.483542263507843,
+      "learning_rate": 8.63486842105263e-07,
+      "loss": 0.3812,
+      "step": 2696
+    },
+    {
+      "epoch": 2.992095409790598,
+      "grad_norm": 0.484693318605423,
+      "learning_rate": 7.401315789473683e-07,
+      "loss": 0.4768,
+      "step": 2697
+    },
+    {
+      "epoch": 2.9932048259603383,
+      "grad_norm": 0.3889079988002777,
+      "learning_rate": 6.167763157894736e-07,
+      "loss": 0.5391,
+      "step": 2698
+    },
+    {
+      "epoch": 2.994314242130079,
+      "grad_norm": 0.39534711837768555,
+      "learning_rate": 4.934210526315788e-07,
+      "loss": 0.2693,
+      "step": 2699
+    },
+    {
+      "epoch": 2.9954236582998197,
+      "grad_norm": 0.28255271911621094,
+      "learning_rate": 3.7006578947368417e-07,
+      "loss": 0.496,
+      "step": 2700
+    },
+    {
+      "epoch": 2.9965330744695606,
+      "grad_norm": 0.39558374881744385,
+      "learning_rate": 2.467105263157894e-07,
+      "loss": 0.3213,
+      "step": 2701
+    },
+    {
+      "epoch": 2.997642490639301,
+      "grad_norm": 0.39010322093963623,
+      "learning_rate": 1.233552631578947e-07,
+      "loss": 0.3245,
+      "step": 2702
+    },
+    {
+      "epoch": 2.9987519068090416,
+      "grad_norm": 0.33332499861717224,
+      "learning_rate": 0.0,
+      "loss": 0.3557,
+      "step": 2703
+    },
+    {
+      "epoch": 2.9987519068090416,
+      "step": 2703,
+      "total_flos": 1.203022932864172e+18,
+      "train_loss": 0.49336877679696933,
+      "train_runtime": 13859.7262,
+      "train_samples_per_second": 1.561,
+      "train_steps_per_second": 0.195
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2703,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.203022932864172e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}